Unverified Commit f26772ac authored by MissPenguin's avatar MissPenguin Committed by GitHub
Browse files

Merge pull request #1425 from Evezerest/dy3

Add PPOCRLabel and FAQ_dy
parents 762c5787 14df1b95
<?xml version="1.0" encoding="UTF-8" standalone="no"?>
<!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 20010904//EN"
"http://www.w3.org/TR/2001/REC-SVG-20010904/DTD/svg10.dtd">
<!-- Created with Sodipodi ("http://www.sodipodi.com/") -->
<svg
width="48pt"
height="48pt"
viewBox="0 0 48 48"
style="overflow:visible;enable-background:new 0 0 48 48"
xml:space="preserve"
id="svg589"
sodipodi:version="0.32"
sodipodi:docname="/home/david/Desktop/action/filesaveas.svg"
sodipodi:docbase="/home/david/Desktop/action"
xmlns="http://www.w3.org/2000/svg"
xmlns:xap="http://ns.adobe.com/xap/1.0/"
xmlns:xapGImg="http://ns.adobe.com/xap/1.0/g/img/"
xmlns:sodipodi="http://sodipodi.sourceforge.net/DTD/sodipodi-0.dtd"
xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
xmlns:xml="http://www.w3.org/XML/1998/namespace"
xmlns:xapMM="http://ns.adobe.com/xap/1.0/mm/"
xmlns:pdf="http://ns.adobe.com/pdf/1.3/"
xmlns:dc="http://purl.org/dc/elements/1.1/"
xmlns:a="http://ns.adobe.com/AdobeSVGViewerExtensions/3.0/"
xmlns:x="adobe:ns:meta/"
xmlns:xlink="http://www.w3.org/1999/xlink">
<defs
id="defs677">
<defs
id="defs796" />
<sodipodi:namedview
id="namedview726" />
<metadata
id="metadata711">
<sfw>
<slices>
<slice
x="0"
y="0"
width="256"
height="256"
sliceID="124333141" />
</slices>
<sliceSourceBounds
x="0"
y="0"
width="256"
height="256"
bottomLeftOrigin="true" />
<optimizationSettings>
<targetSettings
fileFormat="PNG24Format"
targetSettingsID="0">
<PNG24Format
transparency="true"
includeCaption="false"
interlaced="false"
noMatteColor="false"
matteColor="#FFFFFF"
filtered="false" />
</targetSettings>
</optimizationSettings>
</sfw>
<xpacket>
begin='' id='W5M0MpCehiHzreSzNTczkc9d'</xpacket>
<x:xmpmeta
x:xmptk="XMP toolkit 3.0-29, framework 1.6">
<rdf:RDF>
<rdf:Description
rdf:about="uuid:cbee75c6-82d1-45ba-8274-b89c6084675c">
<pdf:Producer>
Adobe PDF library 5.00</pdf:Producer>
</rdf:Description>
<rdf:Description
rdf:about="uuid:cbee75c6-82d1-45ba-8274-b89c6084675c" />
<rdf:Description
rdf:about="uuid:cbee75c6-82d1-45ba-8274-b89c6084675c" />
<rdf:Description
rdf:about="uuid:cbee75c6-82d1-45ba-8274-b89c6084675c">
<xap:CreateDate>
2004-01-26T11:58:28+02:00</xap:CreateDate>
<xap:ModifyDate>
2004-03-28T20:41:40Z</xap:ModifyDate>
<xap:CreatorTool>
Adobe Illustrator 10.0</xap:CreatorTool>
<xap:MetadataDate>
2004-02-16T23:58:32+01:00</xap:MetadataDate>
<xap:Thumbnails>
<rdf:Alt>
<rdf:li
rdf:parseType="Resource">
<xapGImg:format>
JPEG</xapGImg:format>
<xapGImg:width>
256</xapGImg:width>
<xapGImg:height>
256</xapGImg:height>
<xapGImg:image>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</xapGImg:image>
</rdf:li>
</rdf:Alt>
</xap:Thumbnails>
</rdf:Description>
<rdf:Description
rdf:about="uuid:cbee75c6-82d1-45ba-8274-b89c6084675c">
<xapMM:DocumentID>
uuid:4ee3f24b-6ed2-4a2e-8f7a-50b762c8da8b</xapMM:DocumentID>
</rdf:Description>
<rdf:Description
rdf:about="uuid:cbee75c6-82d1-45ba-8274-b89c6084675c">
<dc:format>
image/svg+xml</dc:format>
<dc:title>
<rdf:Alt>
<rdf:li
xml:lang="x-default">
mime.ai</rdf:li>
</rdf:Alt>
</dc:title>
</rdf:Description>
</rdf:RDF>
</x:xmpmeta>
<xpacket>
end='w'</xpacket>
</metadata>
<linearGradient
id="XMLID_9_"
gradientUnits="userSpaceOnUse"
x1="128.9995"
y1="11"
x2="128.9995"
y2="245.0005">
<stop
offset="0"
style="stop-color:#494949"
id="stop717" />
<stop
offset="1"
style="stop-color:#000000"
id="stop718" />
<a:midPointStop
offset="0"
style="stop-color:#494949"
id="midPointStop719" />
<a:midPointStop
offset="0.5"
style="stop-color:#494949"
id="midPointStop720" />
<a:midPointStop
offset="1"
style="stop-color:#000000"
id="midPointStop721" />
</linearGradient>
<linearGradient
id="XMLID_10_"
gradientUnits="userSpaceOnUse"
x1="29.0532"
y1="29.0532"
x2="226.9471"
y2="226.9471">
<stop
offset="0"
style="stop-color:#FFFFFF"
id="stop725" />
<stop
offset="1"
style="stop-color:#DADADA"
id="stop726" />
<a:midPointStop
offset="0"
style="stop-color:#FFFFFF"
id="midPointStop727" />
<a:midPointStop
offset="0.5"
style="stop-color:#FFFFFF"
id="midPointStop728" />
<a:midPointStop
offset="1"
style="stop-color:#DADADA"
id="midPointStop729" />
</linearGradient>
<linearGradient
id="XMLID_11_"
gradientUnits="userSpaceOnUse"
x1="-481.7007"
y1="-94.4194"
x2="-360.2456"
y2="-164.2214"
gradientTransform="matrix(0.1991 0.98 -0.98 0.1991 91.6944 573.5653)">
<stop
offset="0"
style="stop-color:#990000"
id="stop736" />
<stop
offset="1"
style="stop-color:#7C0000"
id="stop737" />
<a:midPointStop
offset="0"
style="stop-color:#990000"
id="midPointStop738" />
<a:midPointStop
offset="0.5"
style="stop-color:#990000"
id="midPointStop739" />
<a:midPointStop
offset="1"
style="stop-color:#7C0000"
id="midPointStop740" />
</linearGradient>
<linearGradient
id="XMLID_12_"
gradientUnits="userSpaceOnUse"
x1="-1375.9844"
y1="685.3809"
x2="-1355.0455"
y2="706.3217"
gradientTransform="matrix(-0.999 0.0435 0.0435 0.999 -1277.0056 -496.5172)">
<stop
offset="0"
style="stop-color:#F8F1DC"
id="stop743" />
<stop
offset="1"
style="stop-color:#D6A84A"
id="stop744" />
<a:midPointStop
offset="0"
style="stop-color:#F8F1DC"
id="midPointStop745" />
<a:midPointStop
offset="0.5"
style="stop-color:#F8F1DC"
id="midPointStop746" />
<a:midPointStop
offset="1"
style="stop-color:#D6A84A"
id="midPointStop747" />
</linearGradient>
<linearGradient
id="XMLID_13_"
gradientUnits="userSpaceOnUse"
x1="65.0947"
y1="-0.7954"
x2="137.6021"
y2="160.1823">
<stop
offset="0"
style="stop-color:#FFA700"
id="stop750" />
<stop
offset="0.7753"
style="stop-color:#FFD700"
id="stop751" />
<stop
offset="1"
style="stop-color:#FF794B"
id="stop752" />
<a:midPointStop
offset="0"
style="stop-color:#FFA700"
id="midPointStop753" />
<a:midPointStop
offset="0.5"
style="stop-color:#FFA700"
id="midPointStop754" />
<a:midPointStop
offset="0.7753"
style="stop-color:#FFD700"
id="midPointStop755" />
<a:midPointStop
offset="0.5"
style="stop-color:#FFD700"
id="midPointStop756" />
<a:midPointStop
offset="1"
style="stop-color:#FF794B"
id="midPointStop757" />
</linearGradient>
<linearGradient
id="XMLID_14_"
gradientUnits="userSpaceOnUse"
x1="-1336.4497"
y1="635.7949"
x2="-1325.3219"
y2="622.5333"
gradientTransform="matrix(-0.999 0.0435 0.0435 0.999 -1277.0056 -496.5172)">
<stop
offset="0"
style="stop-color:#FFC957"
id="stop763" />
<stop
offset="1"
style="stop-color:#FF6D00"
id="stop764" />
<a:midPointStop
offset="0"
style="stop-color:#FFC957"
id="midPointStop765" />
<a:midPointStop
offset="0.5"
style="stop-color:#FFC957"
id="midPointStop766" />
<a:midPointStop
offset="1"
style="stop-color:#FF6D00"
id="midPointStop767" />
</linearGradient>
<linearGradient
id="XMLID_15_"
gradientUnits="userSpaceOnUse"
x1="-1401.459"
y1="595.6309"
x2="-1354.6851"
y2="699.4763"
gradientTransform="matrix(-0.999 0.0435 0.0435 0.999 -1277.0056 -496.5172)">
<stop
offset="0"
style="stop-color:#FFA700"
id="stop770" />
<stop
offset="0.7753"
style="stop-color:#FFD700"
id="stop771" />
<stop
offset="1"
style="stop-color:#FF9200"
id="stop772" />
<a:midPointStop
offset="0"
style="stop-color:#FFA700"
id="midPointStop773" />
<a:midPointStop
offset="0.5"
style="stop-color:#FFA700"
id="midPointStop774" />
<a:midPointStop
offset="0.7753"
style="stop-color:#FFD700"
id="midPointStop775" />
<a:midPointStop
offset="0.5"
style="stop-color:#FFD700"
id="midPointStop776" />
<a:midPointStop
offset="1"
style="stop-color:#FF9200"
id="midPointStop777" />
</linearGradient>
<linearGradient
id="XMLID_16_"
gradientUnits="userSpaceOnUse"
x1="67.8452"
y1="115.5361"
x2="144.5898"
y2="115.5361">
<stop
offset="0"
style="stop-color:#7D7D99"
id="stop780" />
<stop
offset="0.1798"
style="stop-color:#B1B1C5"
id="stop781" />
<stop
offset="0.3727"
style="stop-color:#BCBCC8"
id="stop782" />
<stop
offset="0.6825"
style="stop-color:#C8C8CB"
id="stop783" />
<stop
offset="1"
style="stop-color:#CCCCCC"
id="stop784" />
<a:midPointStop
offset="0"
style="stop-color:#7D7D99"
id="midPointStop785" />
<a:midPointStop
offset="0.5"
style="stop-color:#7D7D99"
id="midPointStop786" />
<a:midPointStop
offset="0.1798"
style="stop-color:#B1B1C5"
id="midPointStop787" />
<a:midPointStop
offset="0.2881"
style="stop-color:#B1B1C5"
id="midPointStop788" />
<a:midPointStop
offset="1"
style="stop-color:#CCCCCC"
id="midPointStop789" />
</linearGradient>
</defs>
<sodipodi:namedview
id="base" />
<metadata
id="metadata590">
<xpacket>
begin='' id='W5M0MpCehiHzreSzNTczkc9d'</xpacket>
<x:xmpmeta
x:xmptk="XMP toolkit 3.0-29, framework 1.6">
<rdf:RDF>
<rdf:Description
rdf:about="uuid:9dfcc10e-f4e2-4cbf-91b0-8deea2f1a998">
<pdf:Producer>
Adobe PDF library 5.00</pdf:Producer>
</rdf:Description>
<rdf:Description
rdf:about="uuid:9dfcc10e-f4e2-4cbf-91b0-8deea2f1a998" />
<rdf:Description
rdf:about="uuid:9dfcc10e-f4e2-4cbf-91b0-8deea2f1a998" />
<rdf:Description
rdf:about="uuid:9dfcc10e-f4e2-4cbf-91b0-8deea2f1a998">
<xap:CreateDate>
2004-02-04T02:08:51+02:00</xap:CreateDate>
<xap:ModifyDate>
2004-03-29T09:20:16Z</xap:ModifyDate>
<xap:CreatorTool>
Adobe Illustrator 10.0</xap:CreatorTool>
<xap:MetadataDate>
2004-02-29T14:54:28+01:00</xap:MetadataDate>
<xap:Thumbnails>
<rdf:Alt>
<rdf:li
rdf:parseType="Resource">
<xapGImg:format>
JPEG</xapGImg:format>
<xapGImg:width>
256</xapGImg:width>
<xapGImg:height>
256</xapGImg:height>
<xapGImg:image>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</xapGImg:image>
</rdf:li>
</rdf:Alt>
</xap:Thumbnails>
</rdf:Description>
<rdf:Description
rdf:about="uuid:9dfcc10e-f4e2-4cbf-91b0-8deea2f1a998">
<xapMM:DocumentID>
uuid:f3c53255-be8a-4b04-817b-695bf2c54c8b</xapMM:DocumentID>
</rdf:Description>
<rdf:Description
rdf:about="uuid:9dfcc10e-f4e2-4cbf-91b0-8deea2f1a998">
<dc:format>
image/svg+xml</dc:format>
<dc:title>
<rdf:Alt>
<rdf:li
xml:lang="x-default">
filesave.ai</rdf:li>
</rdf:Alt>
</dc:title>
</rdf:Description>
</rdf:RDF>
</x:xmpmeta>
<xpacket>
end='w'</xpacket>
</metadata>
<g
id="Layer_1">
<path
style="opacity:0.2;"
d="M9.416,5.208c-2.047,0-3.712,1.693-3.712,3.775V39.15c0,2.082,1.666,3.775,3.712,3.775h29.401 c2.047,0,3.712-1.693,3.712-3.775V8.983c0-2.082-1.665-3.775-3.712-3.775H9.416z"
id="path592" />
<path
style="opacity:0.2;"
d="M9.041,4.833c-2.047,0-3.712,1.693-3.712,3.775v30.167c0,2.082,1.666,3.775,3.712,3.775h29.401 c2.047,0,3.712-1.693,3.712-3.775V8.608c0-2.082-1.665-3.775-3.712-3.775H9.041z"
id="path593" />
<path
style="fill:#00008D;"
d="M8.854,4.646c-2.047,0-3.712,1.693-3.712,3.775v30.167c0,2.082,1.666,3.775,3.712,3.775h29.401 c2.047,0,3.712-1.693,3.712-3.775V8.42c0-2.082-1.665-3.775-3.712-3.775H8.854z"
id="path594" />
<path
style="fill:#00008D;"
d="M8.854,5.021c-1.84,0-3.337,1.525-3.337,3.4v30.167c0,1.875,1.497,3.4,3.337,3.4h29.401 c1.84,0,3.337-1.525,3.337-3.4V8.42c0-1.875-1.497-3.4-3.337-3.4H8.854z"
id="path595" />
<path
id="path166_1_"
style="fill:#FFFFFF;"
d="M40.654,38.588c0,1.36-1.074,2.463-2.399,2.463H8.854c-1.326,0-2.4-1.103-2.4-2.463V8.42 c0-1.36,1.074-2.462,2.4-2.462h29.401c1.325,0,2.399,1.103,2.399,2.462V38.588z" />
<linearGradient
id="path166_2_"
gradientUnits="userSpaceOnUse"
x1="-149.0464"
y1="251.1436"
x2="-149.0464"
y2="436.303"
gradientTransform="matrix(0.1875 0 0 -0.1875 51.5 83.75)">
<stop
offset="0"
style="stop-color:#B4E2FF"
id="stop598" />
<stop
offset="1"
style="stop-color:#006DFF"
id="stop599" />
<a:midPointStop
offset="0"
style="stop-color:#B4E2FF"
id="midPointStop600" />
<a:midPointStop
offset="0.5"
style="stop-color:#B4E2FF"
id="midPointStop601" />
<a:midPointStop
offset="1"
style="stop-color:#006DFF"
id="midPointStop602" />
</linearGradient>
<path
id="path166"
style="fill:url(#path166_2_);"
d="M40.654,38.588c0,1.36-1.074,2.463-2.399,2.463H8.854c-1.326,0-2.4-1.103-2.4-2.463V8.42 c0-1.36,1.074-2.462,2.4-2.462h29.401c1.325,0,2.399,1.103,2.399,2.462V38.588z" />
<path
style="fill:#FFFFFF;"
d="M8.854,6.521c-1.013,0-1.837,0.852-1.837,1.9v30.167c0,1.048,0.824,1.9,1.837,1.9h29.401 c1.013,0,1.837-0.853,1.837-1.9V8.42c0-1.048-0.824-1.9-1.837-1.9H8.854z"
id="path604" />
<linearGradient
id="XMLID_1_"
gradientUnits="userSpaceOnUse"
x1="7.3057"
y1="7.2559"
x2="50.7728"
y2="50.7231">
<stop
offset="0"
style="stop-color:#94CAFF"
id="stop606" />
<stop
offset="1"
style="stop-color:#006DFF"
id="stop607" />
<a:midPointStop
offset="0"
style="stop-color:#94CAFF"
id="midPointStop608" />
<a:midPointStop
offset="0.5"
style="stop-color:#94CAFF"
id="midPointStop609" />
<a:midPointStop
offset="1"
style="stop-color:#006DFF"
id="midPointStop610" />
</linearGradient>
<path
style="fill:url(#XMLID_1_);"
d="M8.854,6.521c-1.013,0-1.837,0.852-1.837,1.9v30.167c0,1.048,0.824,1.9,1.837,1.9h29.401 c1.013,0,1.837-0.853,1.837-1.9V8.42c0-1.048-0.824-1.9-1.837-1.9H8.854z"
id="path611" />
<linearGradient
id="XMLID_2_"
gradientUnits="userSpaceOnUse"
x1="23.5039"
y1="2.187"
x2="23.5039"
y2="34.4368">
<stop
offset="0"
style="stop-color:#428AFF"
id="stop613" />
<stop
offset="1"
style="stop-color:#C9E6FF"
id="stop614" />
<a:midPointStop
offset="0"
style="stop-color:#428AFF"
id="midPointStop615" />
<a:midPointStop
offset="0.5"
style="stop-color:#428AFF"
id="midPointStop616" />
<a:midPointStop
offset="1"
style="stop-color:#C9E6FF"
id="midPointStop617" />
</linearGradient>
<path
style="fill:url(#XMLID_2_);"
d="M36.626,6.861c0,0-26.184,0-26.914,0c0,0.704,0,16.59,0,17.294c0.721,0,26.864,0,27.583,0 c0-0.704,0-16.59,0-17.294C36.988,6.861,36.626,6.861,36.626,6.861z"
id="path618" />
<polygon
id="path186_1_"
style="fill:#FFFFFF;"
points="35.809,6.486 10.221,6.486 10.221,23.405 36.788,23.405 36.788,6.486 " />
<linearGradient
id="path186_2_"
gradientUnits="userSpaceOnUse"
x1="-104.5933"
y1="411.6699"
x2="-206.815"
y2="309.4482"
gradientTransform="matrix(0.1875 0 0 -0.1875 51.5 83.75)">
<stop
offset="0"
style="stop-color:#CCCCCC"
id="stop621" />
<stop
offset="1"
style="stop-color:#F0F0F0"
id="stop622" />
<a:midPointStop
offset="0"
style="stop-color:#CCCCCC"
id="midPointStop623" />
<a:midPointStop
offset="0.5"
style="stop-color:#CCCCCC"
id="midPointStop624" />
<a:midPointStop
offset="1"
style="stop-color:#F0F0F0"
id="midPointStop625" />
</linearGradient>
<polygon
id="path186"
style="fill:url(#path186_2_);"
points="35.809,6.486 10.221,6.486 10.221,23.405 36.788,23.405 36.788,6.486 " />
<path
style="fill:#FFFFFF;stroke:#FFFFFF;stroke-width:0.1875;"
d="M11.488,7.019c0,0.698,0,14.542,0,15.239c0.716,0,23.417,0,24.133,0c0-0.698,0-14.541,0-15.239 C34.904,7.019,12.204,7.019,11.488,7.019z"
id="path627" />
<linearGradient
id="XMLID_3_"
gradientUnits="userSpaceOnUse"
x1="34.5967"
y1="3.5967"
x2="18.4087"
y2="19.7847">
<stop
offset="0"
style="stop-color:#FFFFFF"
id="stop629" />
<stop
offset="0.5506"
style="stop-color:#E6EDFF"
id="stop630" />
<stop
offset="1"
style="stop-color:#FFFFFF"
id="stop631" />
<a:midPointStop
offset="0"
style="stop-color:#FFFFFF"
id="midPointStop632" />
<a:midPointStop
offset="0.5"
style="stop-color:#FFFFFF"
id="midPointStop633" />
<a:midPointStop
offset="0.5506"
style="stop-color:#E6EDFF"
id="midPointStop634" />
<a:midPointStop
offset="0.5"
style="stop-color:#E6EDFF"
id="midPointStop635" />
<a:midPointStop
offset="1"
style="stop-color:#FFFFFF"
id="midPointStop636" />
</linearGradient>
<path
style="fill:url(#XMLID_3_);stroke:#FFFFFF;stroke-width:0.1875;"
d="M11.488,7.019c0,0.698,0,14.542,0,15.239c0.716,0,23.417,0,24.133,0c0-0.698,0-14.541,0-15.239 C34.904,7.019,12.204,7.019,11.488,7.019z"
id="path637" />
<linearGradient
id="path205_1_"
gradientUnits="userSpaceOnUse"
x1="-174.4409"
y1="300.0908"
x2="-108.8787"
y2="210.2074"
gradientTransform="matrix(0.1875 0 0 -0.1875 51.5 83.75)">
<stop
offset="0"
style="stop-color:#003399"
id="stop639" />
<stop
offset="0.2697"
style="stop-color:#0035ED"
id="stop640" />
<stop
offset="1"
style="stop-color:#57ADFF"
id="stop641" />
<a:midPointStop
offset="0"
style="stop-color:#003399"
id="midPointStop642" />
<a:midPointStop
offset="0.5"
style="stop-color:#003399"
id="midPointStop643" />
<a:midPointStop
offset="0.2697"
style="stop-color:#0035ED"
id="midPointStop644" />
<a:midPointStop
offset="0.5"
style="stop-color:#0035ED"
id="midPointStop645" />
<a:midPointStop
offset="1"
style="stop-color:#57ADFF"
id="midPointStop646" />
</linearGradient>
<rect
id="path205"
x="12.154"
y="26.479"
style="fill:url(#path205_1_);"
width="22.007"
height="13.978" />
<linearGradient
id="XMLID_4_"
gradientUnits="userSpaceOnUse"
x1="21.8687"
y1="25.1875"
x2="21.8687"
y2="44.6251">
<stop
offset="0"
style="stop-color:#DFDFDF"
id="stop649" />
<stop
offset="1"
style="stop-color:#7D7D99"
id="stop650" />
<a:midPointStop
offset="0"
style="stop-color:#DFDFDF"
id="midPointStop651" />
<a:midPointStop
offset="0.5"
style="stop-color:#DFDFDF"
id="midPointStop652" />
<a:midPointStop
offset="1"
style="stop-color:#7D7D99"
id="midPointStop653" />
</linearGradient>
<path
style="fill:url(#XMLID_4_);"
d="M13.244,27.021c-0.311,0-0.563,0.252-0.563,0.563v13.104c0,0.312,0.252,0.563,0.563,0.563h17.249 c0.311,0,0.563-0.251,0.563-0.563V27.583c0-0.311-0.252-0.563-0.563-0.563H13.244z M18.85,30.697c0,0.871,0,5.078,0,5.949 c-0.683,0-2.075,0-2.759,0c0-0.871,0-5.078,0-5.949C16.775,30.697,18.167,30.697,18.85,30.697z"
id="path654" />
<linearGradient
id="XMLID_5_"
gradientUnits="userSpaceOnUse"
x1="-158.0337"
y1="288.0684"
x2="-158.0337"
y2="231.3219"
gradientTransform="matrix(0.1875 0 0 -0.1875 51.5 83.75)">
<stop
offset="0"
style="stop-color:#F0F0F0"
id="stop656" />
<stop
offset="0.6348"
style="stop-color:#CECEDB"
id="stop657" />
<stop
offset="0.8595"
style="stop-color:#B1B1C5"
id="stop658" />
<stop
offset="1"
style="stop-color:#FFFFFF"
id="stop659" />
<a:midPointStop
offset="0"
style="stop-color:#F0F0F0"
id="midPointStop660" />
<a:midPointStop
offset="0.5"
style="stop-color:#F0F0F0"
id="midPointStop661" />
<a:midPointStop
offset="0.6348"
style="stop-color:#CECEDB"
id="midPointStop662" />
<a:midPointStop
offset="0.5"
style="stop-color:#CECEDB"
id="midPointStop663" />
<a:midPointStop
offset="0.8595"
style="stop-color:#B1B1C5"
id="midPointStop664" />
<a:midPointStop
offset="0.5"
style="stop-color:#B1B1C5"
id="midPointStop665" />
<a:midPointStop
offset="1"
style="stop-color:#FFFFFF"
id="midPointStop666" />
</linearGradient>
<path
style="fill:url(#XMLID_5_);"
d="M13.244,27.583v13.104h17.249V27.583H13.244z M19.413,37.209h-3.884v-7.074h3.884V37.209z"
id="path667" />
<linearGradient
id="path228_1_"
gradientUnits="userSpaceOnUse"
x1="-68.1494"
y1="388.4561"
x2="-68.1494"
y2="404.6693"
gradientTransform="matrix(0.1875 0 0 -0.1875 51.5 83.75)">
<stop
offset="0"
style="stop-color:#3399FF"
id="stop669" />
<stop
offset="1"
style="stop-color:#000000"
id="stop670" />
<a:midPointStop
offset="0"
style="stop-color:#3399FF"
id="midPointStop671" />
<a:midPointStop
offset="0.5"
style="stop-color:#3399FF"
id="midPointStop672" />
<a:midPointStop
offset="1"
style="stop-color:#000000"
id="midPointStop673" />
</linearGradient>
<rect
id="path228"
x="37.83"
y="9.031"
style="fill:url(#path228_1_);"
width="1.784"
height="1.785" />
<polyline
id="_x3C_Slice_x3E_"
style="fill:none;"
points="0,48 0,0 48,0 48,48 " />
</g>
<rect
id="rect810"
fill="none"
width="256"
height="256"
transform="matrix(0.125000,0.000000,0.000000,0.125000,-41.51768,12.75884)"
style="font-size:12;fill:none;" />
<g
id="g979"
transform="matrix(0.207200,1.691268,-1.691268,0.207200,86.28419,53.75496)">
<path
opacity="0.2"
d="M191.924,195.984c-11.613-36.127-13.717-42.67-14.859-44.064c0.119,0.076,0.289,0.178,0.289,0.178 l-78.55-87.455c-4.195-4.65-14.005,0.356-21.355,6.976c-7.283,6.542-13.32,15.773-9.37,20.564l78.944,87.543l0.533,0.094 l37.768,17.602l7.688,2.365L191.924,195.984z"
id="path731"
transform="matrix(0.125000,0.000000,0.000000,0.125000,-41.51768,12.75884)"
style="font-size:12;opacity:0.2;" />
<path
opacity="0.2"
d="M193.557,193.516c-11.611-36.125-13.713-42.67-14.855-44.064c0.117,0.072,0.287,0.178,0.287,0.178 l-78.545-87.455c-4.199-4.651-14.015,0.355-21.361,6.975c-7.281,6.545-13.32,15.773-9.368,20.566l78.945,87.539l0.533,0.1 l37.77,17.598l7.682,2.367L193.557,193.516z"
id="path732"
transform="matrix(0.125000,0.000000,0.000000,0.125000,-41.51768,12.75884)"
style="font-size:12;opacity:0.2;" />
<path
opacity="0.2"
d="M186.773,191.049c-11.613-36.127-13.713-42.672-14.863-44.068c0.121,0.074,0.295,0.18,0.295,0.18 L93.653,59.704c-4.192-4.65-14.009,0.359-21.354,6.978c-7.283,6.542-13.321,15.771-9.369,20.565l78.942,87.541l0.535,0.096 l37.768,17.598l7.686,2.367L186.773,191.049z"
id="path733"
transform="matrix(0.125000,0.000000,0.000000,0.125000,-41.51768,12.75884)"
style="font-size:12;opacity:0.2;" />
<path
fill="#FFFFFF"
d="M186.43,189.355c-11.613-36.125-13.713-42.666-14.863-44.061c0.123,0.072,0.293,0.18,0.293,0.18 L93.314,58.016c-4.199-4.651-14.015,0.357-21.359,6.977c-7.283,6.543-13.322,15.774-9.37,20.566l78.941,87.541l0.535,0.098 l37.771,17.598l7.686,2.363L186.43,189.355z"
id="path734"
transform="matrix(0.125000,0.000000,0.000000,0.125000,-41.51768,12.75884)"
style="font-size:12;fill:#ffffff;" />
<path
fill="url(#XMLID_11_)"
d="M186.43,189.355c-11.613-36.125-13.713-42.666-14.863-44.061c0.123,0.072,0.293,0.18,0.293,0.18 L93.314,58.016c-4.199-4.651-14.015,0.357-21.359,6.977c-7.283,6.543-13.322,15.774-9.37,20.566l78.941,87.541l0.535,0.098 l37.771,17.598l7.686,2.363L186.43,189.355z"
id="path741"
transform="matrix(0.125000,0.000000,0.000000,0.125000,-41.51768,12.75884)"
style="font-size:12;fill:url(#XMLID_11_);" />
<path
fill="url(#XMLID_12_)"
d="M166.969,147.762l13.723,38.129l-36.371-17.902l0.168-0.152c-0.25-0.08-0.496-0.178-0.701-0.316 l-0.125,0.121l-75.303-83.57l0.123-0.104c-2.246-2.49,1.032-9.094,7.308-14.752c6.28-5.652,13.18-8.219,15.425-5.733 l75.292,83.565L166.969,147.762z"
id="path748"
transform="matrix(0.125000,0.000000,0.000000,0.125000,-41.51768,12.75884)"
style="font-size:12;fill:url(#XMLID_12_);" />
<path
fill="url(#XMLID_13_)"
d="M148.652,170.121c2.076-0.369,4.635-1.479,7.252-3.139c1.617-1.018,3.279-2.283,4.898-3.744 c1.455-1.303,2.736-2.666,3.84-4.01c2.076-2.531,3.322-5.213,3.781-7.424l-1.455-4.043l-0.463-0.715L91.707,64.028 c0.608,2.24-0.962,5.938-4.063,9.74c-1.134,1.389-2.441,2.789-3.945,4.141c-1.574,1.419-3.195,2.652-4.767,3.654 c-4.493,2.871-8.628,3.928-10.548,2.486l-0.025,0.021l75.303,83.57l0.125-0.121c0.205,0.139,0.451,0.236,0.701,0.316 l-0.168,0.152L148.652,170.121z"
id="path758"
transform="matrix(0.125000,0.000000,0.000000,0.125000,-41.51768,12.75884)"
style="font-size:12;fill:url(#XMLID_13_);" />
<path
fill="#FFFFFF"
d="M68.083,83.41c1.732,1.772,5.994,0.776,10.643-2.194c1.541-0.982,3.132-2.193,4.677-3.586 c1.476-1.325,2.759-2.701,3.872-4.063c3.578-4.388,5.091-8.642,3.477-10.584l0.023-0.024l75.817,84.119 c0.635,2.262-0.588,6.498-3.754,10.357c-1.082,1.318-2.34,2.656-3.77,3.934c-1.588,1.434-3.219,2.676-4.807,3.676 c-4.74,3.006-9.303,4.199-11.016,2.301c-0.393-0.439-2.098-2.336-2.145-2.406L67.845,83.626L68.083,83.41z"
id="path759"
transform="matrix(0.125000,0.000000,0.000000,0.125000,-41.51768,12.75884)"
style="font-size:12;fill:#ffffff;" />
<path
fill="#FFFFFF"
d="M75.79,69.215c6.28-5.652,13.18-8.219,15.425-5.733l16.961,18.828l1.152,26.49l-17.973,0.784 L68.359,84.071l0.123-0.104C66.236,81.477,69.514,74.874,75.79,69.215z"
id="path760"
transform="matrix(0.125000,0.000000,0.000000,0.125000,-41.51768,12.75884)"
style="font-size:12;fill:#ffffff;" />
<path
fill="#FFFFFF"
d="M68.083,83.41c1.732,1.772,5.994,0.776,10.643-2.194c1.541-0.982,3.132-2.193,4.677-3.586 c1.476-1.325,2.759-2.701,3.872-4.063c3.578-4.388,5.091-8.642,3.477-10.584l0.023-0.024l75.817,84.119 c0.635,2.262-0.588,6.498-3.754,10.357c-1.082,1.318-2.34,2.656-3.77,3.934c-1.588,1.434-3.219,2.676-4.807,3.676 c-4.74,3.006-9.303,4.199-11.016,2.301c-0.393-0.439-2.098-2.336-2.145-2.406L67.845,83.626L68.083,83.41z"
id="path761"
transform="matrix(0.125000,0.000000,0.000000,0.125000,-41.51768,12.75884)"
style="font-size:12;fill:#ffffff;" />
<path
fill="url(#XMLID_14_)"
d="M75.79,69.215c6.28-5.652,13.18-8.219,15.425-5.733l16.961,18.828l1.152,26.49l-17.973,0.784 L68.359,84.071l0.123-0.104C66.236,81.477,69.514,74.874,75.79,69.215z"
id="path768"
transform="matrix(0.125000,0.000000,0.000000,0.125000,-41.51768,12.75884)"
style="font-size:12;fill:url(#XMLID_14_);" />
<path
fill="url(#XMLID_15_)"
d="M68.083,83.41c1.732,1.772,5.994,0.776,10.643-2.194c1.541-0.982,3.132-2.193,4.677-3.586 c1.476-1.325,2.759-2.701,3.872-4.063c3.578-4.388,5.091-8.642,3.477-10.584l0.023-0.024l75.817,84.119 c0.635,2.262-0.588,6.498-3.754,10.357c-1.082,1.318-2.34,2.656-3.77,3.934c-1.588,1.434-3.219,2.676-4.807,3.676 c-4.74,3.006-9.303,4.199-11.016,2.301c-0.393-0.439-2.098-2.336-2.145-2.406L67.845,83.626L68.083,83.41z"
id="path778"
transform="matrix(0.125000,0.000000,0.000000,0.125000,-41.51768,12.75884)"
style="font-size:12;fill:url(#XMLID_15_);" />
<path
fill="url(#XMLID_16_)"
d="M74.357,90.713c0,0,6.036-0.212,10.685-3.182c1.542-0.983,3.132-2.193,4.677-3.586 c1.477-1.326,2.76-2.701,3.873-4.064c2.928-3.589,4.469-7.088,4.049-9.307l-6.865-7.617l-0.023,0.024 c1.614,1.942,0.102,6.196-3.477,10.584c-1.113,1.362-2.396,2.738-3.872,4.063c-1.545,1.393-3.136,2.604-4.677,3.586 c-4.648,2.971-8.91,3.967-10.643,2.194l-0.238,0.217l73.256,81.311c0.047,0.07,1.752,1.967,2.145,2.406 c0.342,0.377,0.799,0.627,1.344,0.771L74.357,90.713z"
id="path790"
transform="matrix(0.125000,0.000000,0.000000,0.125000,-41.51768,12.75884)"
style="font-size:12;fill:url(#XMLID_16_);" />
<path
fill="#003333"
d="M172.035,175.354c-1.635,1.477-3.307,2.764-4.949,3.84l13.605,6.697l-5.096-14.156 C174.537,172.953,173.352,174.176,172.035,175.354z"
id="path791"
transform="matrix(0.125000,0.000000,0.000000,0.125000,-41.51768,12.75884)"
style="font-size:12;fill:#003333;" />
<path
opacity="0.5"
fill="#FFFFFF"
d="M163.121,157.053L86.968,73.93c0.1-0.12,0.213-0.242,0.307-0.364 c1.428-1.752,2.52-3.49,3.225-5.058l75.768,82.707C165.715,153.039,164.668,155.082,163.121,157.053z"
id="path792"
transform="matrix(0.125000,0.000000,0.000000,0.125000,-41.51768,12.75884)"
style="font-size:12;opacity:0.5;fill:#ffffff;" />
<path
opacity="0.5"
fill="#FFFFFF"
d="M87.275,73.566c0.634-0.774,1.189-1.548,1.694-2.3l76.015,82.974 c-0.578,1.063-1.283,2.146-2.146,3.193c-0.744,0.896-1.566,1.805-2.465,2.697L84.152,76.932 C85.316,75.824,86.361,74.692,87.275,73.566z"
id="path793"
transform="matrix(0.125000,0.000000,0.000000,0.125000,-41.51768,12.75884)"
style="font-size:12;opacity:0.5;fill:#ffffff;" />
</g>
</svg>
<?xml version="1.0" encoding="UTF-8" standalone="no"?>
<!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 20010904//EN"
"http://www.w3.org/TR/2001/REC-SVG-20010904/DTD/svg10.dtd">
<!-- Created with Sodipodi ("http://www.sodipodi.com/") -->
<svg
width="48pt"
height="48pt"
viewBox="0 0 48 48"
style="overflow:visible;enable-background:new 0 0 48 48"
xml:space="preserve"
xmlns="http://www.w3.org/2000/svg"
xmlns:xap="http://ns.adobe.com/xap/1.0/"
xmlns:xapGImg="http://ns.adobe.com/xap/1.0/g/img/"
xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
xmlns:xml="http://www.w3.org/XML/1998/namespace"
xmlns:xapMM="http://ns.adobe.com/xap/1.0/mm/"
xmlns:pdf="http://ns.adobe.com/pdf/1.3/"
xmlns:dc="http://purl.org/dc/elements/1.1/"
xmlns:a="http://ns.adobe.com/AdobeSVGViewerExtensions/3.0/"
xmlns:x="adobe:ns:meta/"
xmlns:sodipodi="http://sodipodi.sourceforge.net/DTD/sodipodi-0.dtd"
xmlns:xlink="http://www.w3.org/1999/xlink"
id="svg589"
sodipodi:version="0.32"
sodipodi:docname="/home/david/Desktop/temp/devices/gnome-dev-floppy.svg"
sodipodi:docbase="/home/david/Desktop/temp/devices/">
<defs
id="defs677" />
<sodipodi:namedview
id="base" />
<metadata
id="metadata590">
<xpacket>begin='' id='W5M0MpCehiHzreSzNTczkc9d' </xpacket>
<x:xmpmeta
x:xmptk="XMP toolkit 3.0-29, framework 1.6">
<rdf:RDF>
<rdf:Description
rdf:about="uuid:9dfcc10e-f4e2-4cbf-91b0-8deea2f1a998">
<pdf:Producer>
Adobe PDF library 5.00</pdf:Producer>
</rdf:Description>
<rdf:Description
rdf:about="uuid:9dfcc10e-f4e2-4cbf-91b0-8deea2f1a998" />
<rdf:Description
rdf:about="uuid:9dfcc10e-f4e2-4cbf-91b0-8deea2f1a998" />
<rdf:Description
rdf:about="uuid:9dfcc10e-f4e2-4cbf-91b0-8deea2f1a998">
<xap:CreateDate>
2004-02-04T02:08:51+02:00</xap:CreateDate>
<xap:ModifyDate>
2004-03-29T09:20:16Z</xap:ModifyDate>
<xap:CreatorTool>
Adobe Illustrator 10.0</xap:CreatorTool>
<xap:MetadataDate>
2004-02-29T14:54:28+01:00</xap:MetadataDate>
<xap:Thumbnails>
<rdf:Alt>
<rdf:li
rdf:parseType="Resource">
<xapGImg:format>
JPEG</xapGImg:format>
<xapGImg:width>
256</xapGImg:width>
<xapGImg:height>
256</xapGImg:height>
<xapGImg:image>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</xapGImg:image>
</rdf:li>
</rdf:Alt>
</xap:Thumbnails>
</rdf:Description>
<rdf:Description
rdf:about="uuid:9dfcc10e-f4e2-4cbf-91b0-8deea2f1a998">
<xapMM:DocumentID>
uuid:f3c53255-be8a-4b04-817b-695bf2c54c8b</xapMM:DocumentID>
</rdf:Description>
<rdf:Description
rdf:about="uuid:9dfcc10e-f4e2-4cbf-91b0-8deea2f1a998">
<dc:format>
image/svg+xml</dc:format>
<dc:title>
<rdf:Alt>
<rdf:li
xml:lang="x-default">
filesave.ai</rdf:li>
</rdf:Alt>
</dc:title>
</rdf:Description>
</rdf:RDF>
</x:xmpmeta>
<xpacket>end='w' </xpacket>
</metadata>
<g
id="Layer_1">
<path
style="opacity:0.2;"
d="M9.416,5.208c-2.047,0-3.712,1.693-3.712,3.775V39.15c0,2.082,1.666,3.775,3.712,3.775h29.401 c2.047,0,3.712-1.693,3.712-3.775V8.983c0-2.082-1.665-3.775-3.712-3.775H9.416z"
id="path592" />
<path
style="opacity:0.2;"
d="M9.041,4.833c-2.047,0-3.712,1.693-3.712,3.775v30.167c0,2.082,1.666,3.775,3.712,3.775h29.401 c2.047,0,3.712-1.693,3.712-3.775V8.608c0-2.082-1.665-3.775-3.712-3.775H9.041z"
id="path593" />
<path
style="fill:#00008D;"
d="M8.854,4.646c-2.047,0-3.712,1.693-3.712,3.775v30.167c0,2.082,1.666,3.775,3.712,3.775h29.401 c2.047,0,3.712-1.693,3.712-3.775V8.42c0-2.082-1.665-3.775-3.712-3.775H8.854z"
id="path594" />
<path
style="fill:#00008D;"
d="M8.854,5.021c-1.84,0-3.337,1.525-3.337,3.4v30.167c0,1.875,1.497,3.4,3.337,3.4h29.401 c1.84,0,3.337-1.525,3.337-3.4V8.42c0-1.875-1.497-3.4-3.337-3.4H8.854z"
id="path595" />
<path
id="path166_1_"
style="fill:#FFFFFF;"
d="M40.654,38.588c0,1.36-1.074,2.463-2.399,2.463H8.854c-1.326,0-2.4-1.103-2.4-2.463V8.42 c0-1.36,1.074-2.462,2.4-2.462h29.401c1.325,0,2.399,1.103,2.399,2.462V38.588z" />
<linearGradient
id="path166_2_"
gradientUnits="userSpaceOnUse"
x1="-149.0464"
y1="251.1436"
x2="-149.0464"
y2="436.303"
gradientTransform="matrix(0.1875 0 0 -0.1875 51.5 83.75)">
<stop
offset="0"
style="stop-color:#B4E2FF"
id="stop598" />
<stop
offset="1"
style="stop-color:#006DFF"
id="stop599" />
<a:midPointStop
offset="0"
style="stop-color:#B4E2FF"
id="midPointStop600" />
<a:midPointStop
offset="0.5"
style="stop-color:#B4E2FF"
id="midPointStop601" />
<a:midPointStop
offset="1"
style="stop-color:#006DFF"
id="midPointStop602" />
</linearGradient>
<path
id="path166"
style="fill:url(#path166_2_);"
d="M40.654,38.588c0,1.36-1.074,2.463-2.399,2.463H8.854c-1.326,0-2.4-1.103-2.4-2.463V8.42 c0-1.36,1.074-2.462,2.4-2.462h29.401c1.325,0,2.399,1.103,2.399,2.462V38.588z" />
<path
style="fill:#FFFFFF;"
d="M8.854,6.521c-1.013,0-1.837,0.852-1.837,1.9v30.167c0,1.048,0.824,1.9,1.837,1.9h29.401 c1.013,0,1.837-0.853,1.837-1.9V8.42c0-1.048-0.824-1.9-1.837-1.9H8.854z"
id="path604" />
<linearGradient
id="XMLID_1_"
gradientUnits="userSpaceOnUse"
x1="7.3057"
y1="7.2559"
x2="50.7728"
y2="50.7231">
<stop
offset="0"
style="stop-color:#94CAFF"
id="stop606" />
<stop
offset="1"
style="stop-color:#006DFF"
id="stop607" />
<a:midPointStop
offset="0"
style="stop-color:#94CAFF"
id="midPointStop608" />
<a:midPointStop
offset="0.5"
style="stop-color:#94CAFF"
id="midPointStop609" />
<a:midPointStop
offset="1"
style="stop-color:#006DFF"
id="midPointStop610" />
</linearGradient>
<path
style="fill:url(#XMLID_1_);"
d="M8.854,6.521c-1.013,0-1.837,0.852-1.837,1.9v30.167c0,1.048,0.824,1.9,1.837,1.9h29.401 c1.013,0,1.837-0.853,1.837-1.9V8.42c0-1.048-0.824-1.9-1.837-1.9H8.854z"
id="path611" />
<linearGradient
id="XMLID_2_"
gradientUnits="userSpaceOnUse"
x1="23.5039"
y1="2.187"
x2="23.5039"
y2="34.4368">
<stop
offset="0"
style="stop-color:#428AFF"
id="stop613" />
<stop
offset="1"
style="stop-color:#C9E6FF"
id="stop614" />
<a:midPointStop
offset="0"
style="stop-color:#428AFF"
id="midPointStop615" />
<a:midPointStop
offset="0.5"
style="stop-color:#428AFF"
id="midPointStop616" />
<a:midPointStop
offset="1"
style="stop-color:#C9E6FF"
id="midPointStop617" />
</linearGradient>
<path
style="fill:url(#XMLID_2_);"
d="M36.626,6.861c0,0-26.184,0-26.914,0c0,0.704,0,16.59,0,17.294c0.721,0,26.864,0,27.583,0 c0-0.704,0-16.59,0-17.294C36.988,6.861,36.626,6.861,36.626,6.861z"
id="path618" />
<polygon
id="path186_1_"
style="fill:#FFFFFF;"
points="35.809,6.486 10.221,6.486 10.221,23.405 36.788,23.405 36.788,6.486 " />
<linearGradient
id="path186_2_"
gradientUnits="userSpaceOnUse"
x1="-104.5933"
y1="411.6699"
x2="-206.815"
y2="309.4482"
gradientTransform="matrix(0.1875 0 0 -0.1875 51.5 83.75)">
<stop
offset="0"
style="stop-color:#CCCCCC"
id="stop621" />
<stop
offset="1"
style="stop-color:#F0F0F0"
id="stop622" />
<a:midPointStop
offset="0"
style="stop-color:#CCCCCC"
id="midPointStop623" />
<a:midPointStop
offset="0.5"
style="stop-color:#CCCCCC"
id="midPointStop624" />
<a:midPointStop
offset="1"
style="stop-color:#F0F0F0"
id="midPointStop625" />
</linearGradient>
<polygon
id="path186"
style="fill:url(#path186_2_);"
points="35.809,6.486 10.221,6.486 10.221,23.405 36.788,23.405 36.788,6.486 " />
<path
style="fill:#FFFFFF;stroke:#FFFFFF;stroke-width:0.1875;"
d="M11.488,7.019c0,0.698,0,14.542,0,15.239c0.716,0,23.417,0,24.133,0c0-0.698,0-14.541,0-15.239 C34.904,7.019,12.204,7.019,11.488,7.019z"
id="path627" />
<linearGradient
id="XMLID_3_"
gradientUnits="userSpaceOnUse"
x1="34.5967"
y1="3.5967"
x2="18.4087"
y2="19.7847">
<stop
offset="0"
style="stop-color:#FFFFFF"
id="stop629" />
<stop
offset="0.5506"
style="stop-color:#E6EDFF"
id="stop630" />
<stop
offset="1"
style="stop-color:#FFFFFF"
id="stop631" />
<a:midPointStop
offset="0"
style="stop-color:#FFFFFF"
id="midPointStop632" />
<a:midPointStop
offset="0.5"
style="stop-color:#FFFFFF"
id="midPointStop633" />
<a:midPointStop
offset="0.5506"
style="stop-color:#E6EDFF"
id="midPointStop634" />
<a:midPointStop
offset="0.5"
style="stop-color:#E6EDFF"
id="midPointStop635" />
<a:midPointStop
offset="1"
style="stop-color:#FFFFFF"
id="midPointStop636" />
</linearGradient>
<path
style="fill:url(#XMLID_3_);stroke:#FFFFFF;stroke-width:0.1875;"
d="M11.488,7.019c0,0.698,0,14.542,0,15.239c0.716,0,23.417,0,24.133,0c0-0.698,0-14.541,0-15.239 C34.904,7.019,12.204,7.019,11.488,7.019z"
id="path637" />
<linearGradient
id="path205_1_"
gradientUnits="userSpaceOnUse"
x1="-174.4409"
y1="300.0908"
x2="-108.8787"
y2="210.2074"
gradientTransform="matrix(0.1875 0 0 -0.1875 51.5 83.75)">
<stop
offset="0"
style="stop-color:#003399"
id="stop639" />
<stop
offset="0.2697"
style="stop-color:#0035ED"
id="stop640" />
<stop
offset="1"
style="stop-color:#57ADFF"
id="stop641" />
<a:midPointStop
offset="0"
style="stop-color:#003399"
id="midPointStop642" />
<a:midPointStop
offset="0.5"
style="stop-color:#003399"
id="midPointStop643" />
<a:midPointStop
offset="0.2697"
style="stop-color:#0035ED"
id="midPointStop644" />
<a:midPointStop
offset="0.5"
style="stop-color:#0035ED"
id="midPointStop645" />
<a:midPointStop
offset="1"
style="stop-color:#57ADFF"
id="midPointStop646" />
</linearGradient>
<rect
id="path205"
x="12.154"
y="26.479"
style="fill:url(#path205_1_);"
width="22.007"
height="13.978" />
<linearGradient
id="XMLID_4_"
gradientUnits="userSpaceOnUse"
x1="21.8687"
y1="25.1875"
x2="21.8687"
y2="44.6251">
<stop
offset="0"
style="stop-color:#DFDFDF"
id="stop649" />
<stop
offset="1"
style="stop-color:#7D7D99"
id="stop650" />
<a:midPointStop
offset="0"
style="stop-color:#DFDFDF"
id="midPointStop651" />
<a:midPointStop
offset="0.5"
style="stop-color:#DFDFDF"
id="midPointStop652" />
<a:midPointStop
offset="1"
style="stop-color:#7D7D99"
id="midPointStop653" />
</linearGradient>
<path
style="fill:url(#XMLID_4_);"
d="M13.244,27.021c-0.311,0-0.563,0.252-0.563,0.563v13.104c0,0.312,0.252,0.563,0.563,0.563h17.249 c0.311,0,0.563-0.251,0.563-0.563V27.583c0-0.311-0.252-0.563-0.563-0.563H13.244z M18.85,30.697c0,0.871,0,5.078,0,5.949 c-0.683,0-2.075,0-2.759,0c0-0.871,0-5.078,0-5.949C16.775,30.697,18.167,30.697,18.85,30.697z"
id="path654" />
<linearGradient
id="XMLID_5_"
gradientUnits="userSpaceOnUse"
x1="-158.0337"
y1="288.0684"
x2="-158.0337"
y2="231.3219"
gradientTransform="matrix(0.1875 0 0 -0.1875 51.5 83.75)">
<stop
offset="0"
style="stop-color:#F0F0F0"
id="stop656" />
<stop
offset="0.6348"
style="stop-color:#CECEDB"
id="stop657" />
<stop
offset="0.8595"
style="stop-color:#B1B1C5"
id="stop658" />
<stop
offset="1"
style="stop-color:#FFFFFF"
id="stop659" />
<a:midPointStop
offset="0"
style="stop-color:#F0F0F0"
id="midPointStop660" />
<a:midPointStop
offset="0.5"
style="stop-color:#F0F0F0"
id="midPointStop661" />
<a:midPointStop
offset="0.6348"
style="stop-color:#CECEDB"
id="midPointStop662" />
<a:midPointStop
offset="0.5"
style="stop-color:#CECEDB"
id="midPointStop663" />
<a:midPointStop
offset="0.8595"
style="stop-color:#B1B1C5"
id="midPointStop664" />
<a:midPointStop
offset="0.5"
style="stop-color:#B1B1C5"
id="midPointStop665" />
<a:midPointStop
offset="1"
style="stop-color:#FFFFFF"
id="midPointStop666" />
</linearGradient>
<path
style="fill:url(#XMLID_5_);"
d="M13.244,27.583v13.104h17.249V27.583H13.244z M19.413,37.209h-3.884v-7.074h3.884V37.209z"
id="path667" />
<linearGradient
id="path228_1_"
gradientUnits="userSpaceOnUse"
x1="-68.1494"
y1="388.4561"
x2="-68.1494"
y2="404.6693"
gradientTransform="matrix(0.1875 0 0 -0.1875 51.5 83.75)">
<stop
offset="0"
style="stop-color:#3399FF"
id="stop669" />
<stop
offset="1"
style="stop-color:#000000"
id="stop670" />
<a:midPointStop
offset="0"
style="stop-color:#3399FF"
id="midPointStop671" />
<a:midPointStop
offset="0.5"
style="stop-color:#3399FF"
id="midPointStop672" />
<a:midPointStop
offset="1"
style="stop-color:#000000"
id="midPointStop673" />
</linearGradient>
<rect
id="path228"
x="37.83"
y="9.031"
style="fill:url(#path228_1_);"
width="1.784"
height="1.785" />
<polyline
id="_x3C_Slice_x3E_"
style="fill:none;"
points="0,48 0,0 48,0 48,48 " />
</g>
</svg>
saveAsDetail=將标签保存到其他文件
changeSaveDir=改变存放目录
openFile=打开文件
shapeLineColorDetail=更改线条颜色
resetAll=重置界面与保存地址
crtBox=矩形标注
crtBoxDetail=创建一个新的区块
dupBoxDetail=复制区块
verifyImg=验证图像
zoominDetail=放大
verifyImgDetail=验证图像
saveDetail=保存标签文件
openFileDetail=打开图像文件
fitWidthDetail=调整宽度适应到窗口宽度
tutorial=PaddleOCR地址
editLabel=编辑标签
openAnnotationDetail=打开标签文件
quit=退出
shapeFillColorDetail=更改填充颜色
closeCurDetail=关闭当前文件
closeCur=关闭文件
deleteImg=删除图像
deleteImgDetail=删除当前图像
fitWin=调整到窗口大小
delBox=删除选择的区块
boxLineColorDetail=选择线框颜色
originalsize=原始大小
resetAllDetail=重置所有设定
zoomoutDetail=放大画面
save=确认
saveAs=另存为
fitWinDetail=缩放到当前窗口大小
openDir=打开目录
copyPrevBounding=复制当前图像中的上一个边界框
showHide=显示/隐藏标签
changeSaveFormat=更改存储格式
shapeFillColor=填充颜色
quitApp=退出程序
dupBox=复制区块
delBoxDetail=删除区块
zoomin=放大画面
info=信息
openAnnotation=开启标签
prevImgDetail=上一个图像
fitWidth=缩放到跟当前画面一样宽
zoomout=缩小画面
changeSavedAnnotationDir=更改保存标签文件的预设目录
nextImgDetail=下一个图像
originalsizeDetail=放大到原始大小
prevImg=上一张
tutorialDetail=显示示范内容
shapeLineColor=形状线条颜色
boxLineColor=区块线条颜色
editLabelDetail=修改当前所选的区块颜色
nextImg=下一张
useDefaultLabel=使用预设标签
useDifficult=有难度的
boxLabelText=区块的标签
labels=标签
autoSaveMode=自动保存模式
singleClsMode=单一类别模式
displayLabel=显示类别
fileList=文件列表
files=文件
advancedMode=专家模式
advancedModeDetail=切换到专家模式
showAllBoxDetail=显示所有区块
hideAllBoxDetail=隐藏所有区块
annoPanel=标注面板
anno=标注
addNewBbox=新框
reLabel=重标注
choosemodel=选择模型
tipchoosemodel=选择OCR模型
ImageResize=图片缩放
IR=图片缩放
autoRecognition=自动标注
reRecognition=重新识别
mfile=文件
medit=编辑
mview=视图
mhelp=帮助
iconList=缩略图
detectionBoxposition=检测框位置
recognitionResult=识别结果
creatPolygon=四点标注
drawSquares=正方形标注
saveRec=保存识别结果
tempLabel=待识别
steps=操作步骤
choseModelLg=选择模型语言
cancel=取消
ok=确认
autolabeling=自动标注中
hideBox=隐藏所有标注
showBox=显示所有标注
saveLabel=保存标记结果
\ No newline at end of file
saveAsDetail=將標籤保存到其他文件
changeSaveDir=改變存放目錄
openFile=開啟檔案
shapeLineColorDetail=更改線條顏色
resetAll=重置
crtBox=創建區塊
crtBoxDetail=畫一個區塊
dupBoxDetail=複製區塊
verifyImg=驗證圖像
zoominDetail=放大
verifyImgDetail=驗證圖像
saveDetail=將標籤存到
openFileDetail=打開圖像
fitWidthDetail=調整到窗口寬度
tutorial=YouTube教學
editLabel=編輯標籤
openAnnotationDetail=打開標籤文件
quit=結束
shapeFillColorDetail=更改填充顏色
closeCurDetail=關閉目前檔案
closeCur=關閉
deleteImg=刪除圖像
deleteImgDetail=刪除目前圖像
fitWin=調整到跟窗口一樣大小
delBox=刪除選取區塊
boxLineColorDetail=選擇框線顏色
originalsize=原始大小
resetAllDetail=重設所有設定
zoomoutDetail=畫面放大
save=儲存
saveAs=另存為
fitWinDetail=縮放到窗口一樣
openDir=開啟目錄
copyPrevBounding=複製當前圖像中的上一個邊界框
showHide=顯示/隱藏標籤
changeSaveFormat=更改儲存格式
shapeFillColor=填充顏色
quitApp=離開本程式
dupBox=複製區塊
delBoxDetail=刪除區塊
zoomin=放大畫面
info=資訊
openAnnotation=開啟標籤
prevImgDetail=上一個圖像
fitWidth=縮放到跟畫面一樣寬
zoomout=縮小畫面
changeSavedAnnotationDir=更改預設標籤存的目錄
nextImgDetail=下一個圖像
originalsizeDetail=放大到原始大小
prevImg=上一個圖像
tutorialDetail=顯示示範內容
shapeLineColor=形狀線條顏色
boxLineColor=日期分隔線顏色
editLabelDetail=修改所選區塊的標籤
nextImg=下一張圖片
useDefaultLabel=使用預設標籤
useDifficult=有難度的
boxLabelText=區塊的標籤
labels=標籤
autoSaveMode=自動儲存模式
singleClsMode=單一類別模式
displayLabel=顯示類別
fileList=檔案清單
files=檔案
iconList=XX
icon=XX
advancedMode=進階模式
advancedModeDetail=切到進階模式
showAllBoxDetail=顯示所有區塊
hideAllBoxDetail=隱藏所有區塊
openFile=Open
openFileDetail=Open image or label file
quit=Quit
quitApp=Quit application
openDir=Open Dir
copyPrevBounding=Copy previous Bounding Boxes in the current image
changeSavedAnnotationDir=Change default saved Annotation dir
openAnnotation=Open Annotation
openAnnotationDetail=Open an annotation file
changeSaveDir=Change Save Dir
nextImg=Next Image
nextImgDetail=Open the next Image
prevImg=Prev Image
prevImgDetail=Open the previous Image
verifyImg=Verify Image
verifyImgDetail=Verify Image
save=Check
saveDetail=Save the labels to a file
changeSaveFormat=Change save format
saveAs=Save As
saveAsDetail=Save the labels to a different file
closeCur=Close
closeCurDetail=Close the current file
deleteImg=Delete current image
deleteImgDetail=Delete the current image
resetAll=Reset Interface and Save Dir
resetAllDetail=Reset All
boxLineColor=Box Line Color
boxLineColorDetail=Choose Box line color
crtBox=Create RectBox
crtBoxDetail=Draw a new box
delBox=Delete RectBox
delBoxDetail=Remove the box
dupBox=Duplicate RectBox
dupBoxDetail=Create a duplicate of the selected box
tutorial=PaddleOCR url
tutorialDetail=Show demo
info=Information
zoomin=Zoom In
zoominDetail=Increase zoom level
zoomout=Zoom Out
zoomoutDetail=Decrease zoom level
originalsize=Original size
originalsizeDetail=Zoom to original size
fitWin=Fit Window
fitWinDetail=Zoom follows window size
fitWidth=Fit Width
fitWidthDetail=Zoom follows window width
editLabel=Edit Label
editLabelDetail=Modify the label of the selected Box
shapeLineColor=Shape Line Color
shapeLineColorDetail=Change the line color for this specific shape
shapeFillColor=Shape Fill Color
shapeFillColorDetail=Change the fill color for this specific shape
showHide=Show/Hide Label Panel
useDefaultLabel=Use default label
useDifficult=Difficult
boxLabelText=Box Labels
labels=Labels
autoSaveMode=Auto Save mode
singleClsMode=Single Class Mode
displayLabel=Display Labels
fileList=File List
files=Files
advancedMode=Advanced Mode
advancedModeDetail=Swtich to advanced mode
showAllBoxDetail=Show all bounding boxes
hideAllBoxDetail=Hide all bounding boxes
annoPanel=anno Panel
anno=anno
addNewBbox=new bbox
reLabel=reLabel
choosemodel=Choose OCR model
tipchoosemodel=Choose OCR model from dir
ImageResize=Image Resize
IR=Image Resize
autoRecognition=Auto Recognition
reRecognition=Re-recognition
mfile=File
medit=Eidt
mview=View
mhelp=Help
iconList=Icon List
detectionBoxposition=Detection box position
recognitionResult=Recognition result
creatPolygon=Create Quadrilateral
drawSquares=Draw Squares
saveRec=Save Recognition Result
tempLabel=TEMPORARY
steps=Steps
choseModelLg=Choose Model Language
cancel=Cancel
ok=OK
autolabeling=Automatic Labeling
hideBox=Hide All Box
showBox=Show All Box
saveLabel=Save Label
\ No newline at end of file
[bumpversion]
commit = True
tag = True
[bumpversion:file:setup.py]
[bdist_wheel]
universal = 1
# Copyright (c) <2015-Present> Tzutalin
# Copyright (C) 2013 MIT, Computer Science and Artificial Intelligence Laboratory. Bryan Russell, Antonio Torralba,
# William T. Freeman. Permission is hereby granted, free of charge, to any person obtaining a copy of this software and
# associated documentation files (the "Software"), to deal in the Software without restriction, including without
# limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the
# Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
# The above copyright notice and this permission notice shall be included in all copies or substantial portions of
# the Software. THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT
# NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT
# SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF
# CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
# THE SOFTWARE.
#!/usr/bin/env python
# -*- coding: utf-8 -*-
from setuptools import setup, find_packages, Command
from sys import platform as _platform
from shutil import rmtree
import sys
import os
here = os.path.abspath(os.path.dirname(__file__))
NAME = 'labelImg'
REQUIRES_PYTHON = '>=3.0.0'
REQUIRED_DEP = ['pyqt5', 'lxml']
about = {}
with open(os.path.join(here, 'libs', '__init__.py')) as f:
exec(f.read(), about)
with open('README.rst') as readme_file:
readme = readme_file.read()
with open('HISTORY.rst') as history_file:
history = history_file.read()
# OS specific settings
SET_REQUIRES = []
if _platform == "linux" or _platform == "linux2":
# linux
print('linux')
elif _platform == "darwin":
# MAC OS X
SET_REQUIRES.append('py2app')
required_packages = find_packages()
required_packages.append('labelImg')
APP = [NAME + '.py']
OPTIONS = {
'argv_emulation': True,
'iconfile': 'resources/icons/app.icns'
}
class UploadCommand(Command):
"""Support setup.py upload."""
description=readme + '\n\n' + history,
user_options = []
@staticmethod
def status(s):
"""Prints things in bold."""
print('\033[1m{0}\033[0m'.format(s))
def initialize_options(self):
pass
def finalize_options(self):
pass
def run(self):
try:
self.status('Removing previous builds…')
rmtree(os.path.join(here, 'dist'))
except OSError:
self.status('Fail to remove previous builds..')
pass
self.status('Building Source and Wheel (universal) distribution…')
os.system(
'{0} setup.py sdist bdist_wheel --universal'.format(sys.executable))
self.status('Uploading the package to PyPI via Twine…')
os.system('twine upload dist/*')
self.status('Pushing git tags…')
os.system('git tag -d v{0}'.format(about['__version__']))
os.system('git tag v{0}'.format(about['__version__']))
# os.system('git push --tags')
sys.exit()
setup(
app=APP,
name=NAME,
version=about['__version__'],
description="LabelImg is a graphical image annotation tool and label object bounding boxes in images",
long_description=readme + '\n\n' + history,
author="TzuTa Lin",
author_email='tzu.ta.lin@gmail.com',
url='https://github.com/tzutalin/labelImg',
python_requires=REQUIRES_PYTHON,
package_dir={'labelImg': '.'},
packages=required_packages,
entry_points={
'console_scripts': [
'labelImg=labelImg.labelImg:main'
]
},
include_package_data=True,
install_requires=REQUIRED_DEP,
license="MIT license",
zip_safe=False,
keywords='labelImg labelTool development annotation deeplearning',
classifiers=[
'Development Status :: 5 - Production/Stable',
'Intended Audience :: Developers',
'License :: OSI Approved :: MIT License',
'Natural Language :: English',
'Programming Language :: Python :: 3',
'Programming Language :: Python :: 3.3',
'Programming Language :: Python :: 3.4',
'Programming Language :: Python :: 3.5',
'Programming Language :: Python :: 3.6',
'Programming Language :: Python :: 3.7',
],
package_data={'data/predefined_classes.txt': ['data/predefined_classes.txt']},
options={'py2app': OPTIONS},
setup_requires=SET_REQUIRES,
# $ setup.py publish support.
cmdclass={
'upload': UploadCommand,
}
)
......@@ -9,19 +9,44 @@
## PaddleOCR常见问题汇总(持续更新)
* [近期更新(2020.12.07)](#近期更新)
* [【精选】OCR精选10个问题](#OCR精选10个问题)
* [【理论篇】OCR通用21个问题](#OCR通用问题)
* [基础知识3](#基础知识)
* [数据集4](#数据集)
* [模型训练调优6](#模型训练调优)
* [预测部署8](#预测部署)
* [【实战篇】PaddleOCR实战53个问题](#PaddleOCR实战问题)
* [使用咨询17](#使用咨询)
* [数据集9](#数据集)
* [模型训练调优13](#模型训练调优)
* [预测部署14](#预测部署)
* [【理论篇】OCR通用30个问题](#OCR通用问题)
* [基础知识7](#基础知识)
* [数据集7](#数据集2)
* [模型训练调优7](#模型训练调优2)
* [预测部署9](#预测部署2)
* [【实战篇】PaddleOCR实战84个问题](#PaddleOCR实战问题)
* [使用咨询20](#使用咨询)
* [数据集17](#数据集3)
* [模型训练调优24](#模型训练调优3)
* [预测部署23](#预测部署3)
<a name="近期更新"></a>
## 近期更新(2020.12.07)
#### Q2.4.9:弯曲文本有试过opencv的TPS进行弯曲校正吗?
**A**:opencv的tps需要标出上下边界对应的点,这些点很难通过传统方法或者深度学习方法获取。PaddleOCR里StarNet网络中的tps模块实现了自动学点,自动校正,可以直接尝试这个。
#### Q3.3.20: 文字检测时怎么模糊的数据增强?
**A**: 模糊的数据增强需要修改代码进行添加,以DB为例,参考[Normalize](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/ppocr/data/imaug/operators.py#L60) ,添加模糊的增强就行
#### Q3.3.21: 文字检测时怎么更改图片旋转的角度,实现360度任意旋转?
**A**: 将[这里](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/ppocr/data/imaug/iaa_augment.py#L64) 的(-10,10) 改为(-180,180)即可
#### Q3.3.22: 训练数据的长宽比过大怎么修改shape
**A**: 识别修改[这里](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/configs/rec/ch_ppocr_v2.0/rec_chinese_lite_train_v2.0.yaml#L75) ,
检测修改[这里](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml#L85)
#### Q3.4.23:安装paddleocr后,提示没有paddle
**A**:这是因为paddlepaddle gpu版本和cpu版本的名称不一致,现在已经在[whl的文档](./whl.md)里做了安装说明。
<a name="OCR精选10个问题"></a>
## 【精选】OCR精选10个问题
......@@ -106,12 +131,14 @@
#### Q1.1.10:PaddleOCR中,对于模型预测加速,CPU加速的途径有哪些?基于TenorRT加速GPU对输入有什么要求?
**A**:(1)CPU可以使用mkldnn进行加速;对于python inference的话,可以把enable_mkldnn改为true,[参考代码](https://github.com/PaddlePaddle/PaddleOCR/blob/549108fe0aa0d87c0a3b2d471f1c653e89daab80/tools/infer/utility.py#L73),对于cpp inference的话,在配置文件里面配置use_mkldnn 1即可,[参考代码](https://github.com/PaddlePaddle/PaddleOCR/blob/549108fe0aa0d87c0a3b2d471f1c653e89daab80/deploy/cpp_infer/tools/config.txt#L6)
**A**:(1)CPU可以使用mkldnn进行加速;对于python inference的话,可以把enable_mkldnn改为true,[参考代码](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/tools/infer/utility.py#L84),对于cpp inference的话,在配置文件里面配置use_mkldnn 1即可,[参考代码](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/deploy/cpp_infer/tools/config.txt#L6)
(2)GPU需要注意变长输入问题等,TRT6 之后才支持变长输入
<a name="OCR通用问题"></a>
## 【理论篇】OCR通用问题
<a name="基础知识"></a>
### 基础知识
#### Q2.1.1:CRNN能否识别两行的文字?还是说必须一行?
......@@ -127,7 +154,20 @@
**A**:端到端在文字分布密集的业务场景,效率会比较有保证,精度的话看自己业务数据积累情况,如果行级别的识别数据积累比较多的话two-stage会比较好。百度的落地场景,比如工业仪表识别、车牌识别都用到端到端解决方案。
#### Q2.1.4 印章如何识别
**A**: 1. 使用带tps的识别网络或abcnet,2.使用极坐标变换将图片拉平之后使用crnn
#### Q2.1.5 多语言的字典里是混合了不同的语种,这个是有什么讲究吗?统一到一个字典里会对精度造成多大的损失?
**A**:统一到一个字典里,会造成最后一层FC过大,增加模型大小。如果有特殊需求的话,可以把需要的几种语言合并字典训练模型,合并字典之后如果引入过多的形近字,可能会造成精度损失,字符平衡的问题可能也需要考虑一下。在PaddleOCR里暂时将语言字典分开。
#### Q2.1.6 预处理部分,图片的长和宽为什么要处理成32的倍数?
**A**:以检测中的resnet骨干网络为例,图像输入网络之后,需要经过5次2倍降采样,共32倍,因此建议输入的图像尺寸为32的倍数。
#### Q2.1.7:类似泰语这样的小语种,部分字会占用两个字符甚至三个字符,请问如何制作字典。
**A**:处理字符的时候,把多字符的当作一个字就行,字典中每行是一个字。
<a name="数据集2"></a>
### 数据集
#### Q2.2.1:支持空格的模型,标注数据的时候是不是要标注空格?中间几个空格都要标注出来么?
......@@ -146,6 +186,19 @@
**A**:可以根据实际场景做不同的尝试,共享一个类别是可以收敛,效果也还不错。但是如果分开训练,同类样本之间一致性更好,更容易收敛,识别效果会更优。
#### Q2.2.5: 文本行较紧密的情况下如何准确检测?
**A**:使用基于分割的方法,如DB,检测密集文本行时,最好收集一批数据进行训练,并且在训练时,并将生成二值图像的shrink_ratio参数调小一些。
#### Q2.2.6: 当训练数据量少时,如何获取更多的数据?
**A**: 当训练数据量少时,可以尝试以下三种方式获取更多的数据:(1)人工采集更多的训练数据,最直接也是最有效的方式。(2)基于PIL和opencv基本图像处理或者变换。例如PIL中ImageFont, Image, ImageDraw三个模块将文字写到背景中,opencv的旋转仿射变换,高斯滤波等。(3)利用数据生成算法合成数据,例如pix2pix等算法。
#### Q2.2.7: 论文《Editing Text in the Wild》中文本合成方法SRNet有什么特点?
**A**: SRNet是借鉴GAN中图像到图像转换、风格迁移的想法合成文本数据。不同于通用GAN的方法只选择一个分支,SRNet将文本合成任务分解为三个简单的子模块,提升合成数据的效果。这三个子模块为不带背景的文本风格迁移模块、背景抽取模块和融合模块。PaddleOCR计划将在2020年12月中旬开源基于SRNet的实用模型。
<a name="模型训练调优2"></a>
### 模型训练调优
#### Q2.3.1:如何更换文本检测/识别的backbone?
......@@ -179,6 +232,15 @@
**A**:在中文识别模型训练时,并不是采用直接将训练样本缩放到[3,32,320]进行训练,而是先等比例缩放图像,保证图像高度为32,宽度不足320的部分补0,宽高比大于10的样本直接丢弃。预测时,如果是单张图像预测,则按上述操作直接对图像缩放,不做宽度320的限制。如果是多张图预测,则采用batch方式预测,每个batch的宽度动态变换,采用这个batch中最长宽度。
#### Q2.3.7:识别训练时,训练集精度已经到达90了,但验证集精度一直在70,涨不上去怎么办?
**A**:训练集精度90,测试集70多的话,应该是过拟合了,有两个可尝试的方法:
(1)加入更多的增广方式或者调大增广prob的[概率](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/ppocr/data/imaug/rec_img_aug.py#L341),默认为0.4。
(2)调大系统的[l2 dcay值](https://github.com/PaddlePaddle/PaddleOCR/blob/a501603d54ff5513fc4fc760319472e59da25424/configs/rec/ch_ppocr_v1.1/rec_chinese_lite_train_v1.1.yml#L47)
<a name="预测部署2"></a>
### 预测部署
#### Q2.4.1:请问对于图片中的密集文字,有什么好的处理办法吗?
......@@ -221,10 +283,16 @@
**A**:表格目前学术界比较成熟的解决方案不多 ,可以尝试下分割的论文方案。
#### Q2.4.9:弯曲文本有试过opencv的TPS进行弯曲校正吗?
**A**:opencv的tps需要标出上下边界对应的点,这个点很难通过传统方法或者深度学习方法获取。PaddleOCR里StarNet网络中的tps模块实现了自动学点,自动校正,可以直接尝试这个。
<a name="PaddleOCR实战问题"></a>
## 【实战篇】PaddleOCR实战问题
<a name="使用咨询"></a>
### 使用咨询
#### Q3.1.1:OSError: [WinError 126] 找不到指定的模块。mac pro python 3.4 shapely import 问题
......@@ -261,7 +329,7 @@
#### Q3.1.9:模型的解码部分有后处理?
**A**:有的检测的后处理在ppocr/postprocess路径下,识别的后处理均在ppocr/utils/character.py文件内
**A**:有的检测的后处理在ppocr/postprocess路径下
#### Q3.1.10:PaddleOCR中文模型是否支持数字识别?
......@@ -269,15 +337,15 @@
#### Q3.1.11:PaddleOCR如何做到横排和竖排同时支持的?
**A**:合成了一批竖排文字,逆时针旋转90度后加入训练集与横排一起训练。预测时根据图片长比判断是否为竖排,若为竖排则将crop出的文本逆时针旋转90度后送入识别网络。
**A**:合成了一批竖排文字,逆时针旋转90度后加入训练集与横排一起训练。预测时根据图片长比判断是否为竖排,若为竖排则将crop出的文本逆时针旋转90度后送入识别网络。
#### Q3.1.12:如何获取检测文本框的坐标?
**A**:文本检测的结果有box和文本信息, 具体 [参考代码](https://github.com/PaddlePaddle/PaddleOCR/blob/9d33e36df550762b204d5fbfd7977a25e31b2c44/tools/infer/predict_system.py#L13)
**A**:文本检测的结果有box和文本信息, 具体 [参考代码](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/tools/infer/predict_system.py)
#### Q3.1.13:识别模型框出来的位置太紧凑,会丢失边缘的文字信息,导致识别错误
**A**: 可以在命令中加入 --det_db_unclip_ratio ,参数[定义位置](https://github.com/PaddlePaddle/PaddleOCR/blob/develop/tools/infer/utility.py#L49),这个参数是检测后处理时控制文本框大小的,默认2.0,可以尝试改成2.5或者更大,反之,如果觉得文本框不够紧凑,也可以把该参数调小。
**A**: 可以在命令中加入 --det_db_unclip_ratio ,参数[定义位置](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/tools/infer/utility.py#L48),这个参数是检测后处理时控制文本框大小的,默认1.6,可以尝试改成2.5或者更大,反之,如果觉得文本框不够紧凑,也可以把该参数调小。
#### Q3.1.14:英文手写体识别有计划提供的预训练模型吗?
......@@ -305,7 +373,24 @@
|8.6M超轻量中文OCR模型|MobileNetV3+MobileNetV3|det_mv3_db.yml|rec_chinese_lite_train.yml|
|通用中文OCR模型|Resnet50_vd+Resnet34_vd|det_r50_vd_db.yml|rec_chinese_common_train.yml|
#### !!Q3.1.18:如何加入自己的检测算法?
**A**:1. 在ppocr/modeling对应目录下分别选择backbone,head。如果没有可用的可以新建文件并添加
2. 在ppocr/data下选择对应的数据处理处理方式,如果没有可用的可以新建文件并添加
3. 在ppocr/losses下新建文件并编写loss
4. 在ppocr/postprocess下新建文件并编写后处理算法
5. 将上面四个步骤里新添加的类或函数参照yml文件写到配置中
#### !!Q3.1.19:训练的时候报错`reader raised an exception`,但是具体不知道是啥问题?
**A**:这个一般是因为标注文件格式有问题或者是标注文件中的图片路径有问题导致的,在[tools/train.py](../../tools/train.py)文件中有一个`test_reader`的函数,基于这个去检查一下数据的格式以及标注,确认没问题之后再进行模型训练。
#### Q3.1.20:PaddleOCR与百度的其他OCR产品有什么区别?
**A**:PaddleOCR主要聚焦通用ocr,如果有垂类需求,您可以用PaddleOCR+垂类数据自己训练;
如果缺少带标注的数据,或者不想投入研发成本,建议直接调用开放的API,开放的API覆盖了目前比较常见的一些垂类。
<a name="数据集3"></a>
### 数据集
#### Q3.2.1:如何制作PaddleOCR支持的数据格式
......@@ -359,6 +444,45 @@
**A**:可以主要参考可视化效果,通用模型更倾向于检测一整行文字,轻量级可能会有一行文字被分成两段检测的情况,不是数量越多,效果就越好。
#### Q3.2.10:crnn+ctc模型训练所用的垂直文本(旋转至水平方向)是如何生成的?
**A**:方法与合成水平方向文字一致,只是将字体替换成了垂直字体。
#### Q3.2.11:有哪些标注工具可以标注OCR数据集?
**A**:您可以参考:https://github.com/PaddlePaddle/PaddleOCR/blob/develop/doc/doc_en/data_annotation_en.md。
我们计划推出高效标注OCR数据的标注工具,请您持续关注PaddleOCR的近期更新。
#### Q3.2.12:一些特殊场景的数据识别效果差,但是数据量很少,不够用来finetune怎么办?
**A**:您可以合成一些接近使用场景的数据用于训练。
我们计划推出基于特定场景的文本数据合成工具,请您持续关注PaddleOCR的近期更新。
#### Q3.2.13:特殊字符(例如一些标点符号)识别效果不好怎么办?
**A**:首先请您确认要识别的特殊字符是否在字典中。
如果字符在已经字典中但效果依然不好,可能是由于识别数据较少导致的,您可以增加相应数据finetune模型。
#### Q3.2.14:PaddleOCR可以识别灰度图吗?
**A**:PaddleOCR的模型均为三通道输入。如果您想使用灰度图作为输入,建议直接用3通道的模式读入灰度图,
或者将单通道图像转换为三通道图像再识别。例如,opencv的cvtColor函数就可以将灰度图转换为RGB三通道模式。
#### Q3.2.15: 文本标注工具PPOCRLabel有什么特色?
**A**: PPOCRLabel是一个半自动文本标注工具,它使用基于PPOCR的中英文OCR模型,预先预测文本检测和识别结果,然后用户对上述结果进行校验和修正就行,大大提高用户的标注效率。同时导出的标注结果直接适配PPOCR训练所需要的数据格式,
#### Q3.2.16: 文本标注工具PPOCRLabel,可以更换模型吗?
**A**: PPOCRLabel中OCR部署方式采用的基于pip安装whl包快速推理,可以参考相关文档更换模型路径,进行特定任务的标注适配。基于pip安装whl包快速推理的文档如下,https://github.com/PaddlePaddle/PaddleOCR/blob/develop/doc/doc_ch/whl.md。
#### Q3.2.17: 文本标注工具PPOCRLabel支持的运行环境有哪些?
**A**: PPOCRLabel可运行于Linux、Windows、MacOS等多种系统。操作步骤可以参考文档,https://github.com/PaddlePaddle/PaddleOCR/blob/develop/PPOCRLabel/README.md
<a name="模型训练调优3"></a>
### 模型训练调优
#### Q3.3.1:文本长度超过25,应该怎么处理?
......@@ -368,12 +492,12 @@
#### Q3.3.2:配置文件里面检测的阈值设置么?
**A**:有的,检测相关的参数主要有以下几个:
``max_side_len:预测时图像resize的长边尺寸
thresh: 用于二值化输出图的阈值
box_thresh:用于过滤文本框的阈值,低于此阈值的文本框不要
unclip_ratio: 文本框扩张的系数,关系到文本框的大小``
``det_limit_side_len:预测时图像resize的长边尺寸
det_db_thresh: 用于二值化输出图的阈值
det_db_box_thresh:用于过滤文本框的阈值,低于此阈值的文本框不要
det_db_unclip_ratio: 文本框扩张的系数,关系到文本框的大小``
这些参数的默认值见[代码](https://github.com/PaddlePaddle/PaddleOCR/blob/develop/tools/infer/utility.py#L40),可以通过从命令行传递参数进行修改。
这些参数的默认值见[代码](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/tools/infer/utility.py#L42),可以通过从命令行传递参数进行修改。
#### Q3.3.3:我想请教一下,你们在训练识别时候,lsvt里的非矩形框文字,你们是怎么做处理的呢。忽略掉还是去最小旋转框?
......@@ -383,50 +507,103 @@ unclip_ratio: 文本框扩张的系数,关系到文本框的大小``
**A**:可以通过下面的脚本终止所有包含train.py字段的进程,
```
```shell
ps -axu | grep train.py | awk '{print $2}' | xargs kill -9
```
#### Q3.3.5:读数据进程数设置4~8时训练一会进程接连defunct后gpu利用率一直为0卡死
**A**:修改多进程的队列数后解决, 将[代码段]( https://github.com/PaddlePaddle/PaddleOCR/blob/549108fe0aa0d87c0a3b2d471f1c653e89daab80/ppocr/data/reader_main.py#L75 ) 修改为:
```
return paddle.reader.multiprocess_reader(readers, False, queue_size=320)
```
#### Q3.3.6:可不可以将pretrain_weights设置为空呢?想从零开始训练一个model
#### Q3.3.5:可不可以将pretrain_weights设置为空呢?想从零开始训练一个model
**A**:这个是可以的,在训练通用识别模型的时候,pretrain_weights就设置为空,但是这样可能需要更长的迭代轮数才能达到相同的精度。
#### Q3.3.7:PaddleOCR默认不是200个step保存一次模型吗?为啥文件夹下面都没有生成
#### Q3.3.6:PaddleOCR默认不是200个step保存一次模型吗?为啥文件夹下面都没有生成
**A**:因为默认保存的起始点不是0,而是4000,将eval_batch_step [4000, 5000]改为[0, 2000] 就是从第0次迭代开始,每2000迭代保存一次模型
#### Q3.3.8:如何进行模型微调?
#### Q3.3.7:如何进行模型微调?
**A**:注意配置好合适的数据集,对齐数据格式,然后在finetune训练时,可以加载我们提供的预训练模型,设置配置文件中Global.pretrain_weights 参数为要加载的预训练模型路径。
#### Q3.3.9:文本检测换成自己的数据没法训练,有一些”###”是什么意思?
#### Q3.3.8:文本检测换成自己的数据没法训练,有一些”###”是什么意思?
**A**:数据格式有问题,”###” 表示要被忽略的文本区域,所以你的数据都被跳过了,可以换成其他任意字符或者就写个空的。
#### Q3.3.10:copy_from_cpu这个地方,这块input不变(t_data的size不变)连续调用两次copy_from_cpu()时,这里面的gpu_place会重新malloc GPU内存吗?还是只有当ele_size变化时才会重新在GPU上malloc呢?
#### Q3.3.9:copy_from_cpu这个地方,这块input不变(t_data的size不变)连续调用两次copy_from_cpu()时,这里面的gpu_place会重新malloc GPU内存吗?还是只有当ele_size变化时才会重新在GPU上malloc呢?
**A**:小于等于的时候都不会重新分配,只有大于的时候才会重新分配
#### Q3.3.11:自己训练出来的未inference转换的模型 可以当作预训练模型吗?
#### Q3.3.10:自己训练出来的未inference转换的模型 可以当作预训练模型吗?
**A**:可以的,但是如果训练数据量少的话,可能会过拟合到少量数据上,泛化性能不佳。
#### Q3.3.12:使用带TPS的识别模型预测报错
#### Q3.3.11:使用带TPS的识别模型预测报错
**A**:TPS模块暂时不支持导出,后续更新。
#### Q3.3.12:如何更换文本检测/识别的backbone?报错信息:``Input(X) dims[3] and Input(Grid) dims[2] should be equal, but received X dimension[3](320) != Grid dimension[2](100) ``
**A**:直接更换配置文件里的Backbone.name即可,格式为:网络文件路径,网络Class名词。如果所需的backbone在PaddleOCR里没有提供,可以参照PaddleClas里面的网络结构,进行修改尝试。具体修改原则可以参考OCR通用问题中 "如何更换文本检测/识别的backbone" 的回答。
#### Q3.3.13: 训练中使用的字典需要与加载的预训练模型使用的字典一样吗?
**A**:分情况,1. 不改变识别字符,训练的字典与你使用该模型进行预测的字典需要保持一致的。
2. 改变识别的字符,这种情况可以不一样,最后一层会重新训练
#### Q3.3.14: 如何对检测模型finetune,比如冻结前面的层或某些层使用小的学习率学习?
**A**
**A**:如果是冻结某些层,可以将变量的stop_gradient属性设置为True,这样计算这个变量之前的所有参数都不会更新了,参考:https://www.paddlepaddle.org.cn/documentation/docs/zh/develop/faq/train_cn.html#id4
如果对某些层使用更小的学习率学习,静态图里还不是很方便,一个方法是在参数初始化的时候,给权重的属性设置固定的学习率,参考:https://www.paddlepaddle.org.cn/documentation/docs/zh/develop/api/paddle/fluid/param_attr/ParamAttr_cn.html#paramattr
实际上我们实验发现,直接加载模型去fine-tune,不设置某些层不同学习率,效果也都不错
#### Q3.3.15: 使用通用中文模型作为预训练模型,更改了字典文件,出现ctc_fc_b not used的错误
**A**:修改了字典之后,识别模型的最后一层FC纬度发生了改变,没有办法加载参数。这里是一个警告,可以忽略,正常训练即可。
#### Q3.3.16: cpp_infer 在Windows下使用vs2015编译不通过
**A**:1. windows上建议使用VS2019工具编译,具体编译细节参考[链接](https://github.com/PaddlePaddle/PaddleOCR/blob/develop/deploy/cpp_infer/docs/windows_vs2019_build.md)
**A**:2. 在release模式下而不是debug模式下编译,参考[issue](https://github.com/PaddlePaddle/PaddleOCR/issues/1023)
#### Q3.3.17: No module named 'tools.infer'
**A**:1. 确保在PaddleOCR/目录下执行的指令,执行'export PYTHONPATH=.'
**A**:2. 拉取github上最新代码,这个问题在10月底已修复。
#### Q3.3.18: 训练模型和测试模型的检测结果差距较大
**A**:1. 检查两个模型使用的后处理参数是否是一样的,训练的后处理参数在配置文件中的PostProcess部分,测试模型的后处理参数在tools/infer/utility.py中,最新代码中两个后处理参数已保持一致。
#### Q3.3.19: 使用合成数据精调小模型后,效果可以,但是还没开源的小infer模型效果好,这是为什么呢?
**A**
**A**:直接更换配置文件里的Backbone.function即可,格式为:网络文件路径,网络Class名词。如果所需的backbone在PaddleOCR里没有提供,可以参照PaddleClas里面的网络结构,进行修改尝试。具体修改原则可以参考OCR通用问题中 "如何更换文本检测/识别的backbone" 的回答。
(1)要保证使用的配置文件和pretrain weights是对应的;
#### Q3.3.13:如何更换文本检测/识别的backbone?报错信息:``Input(X) dims[3] and Input(Grid) dims[2] should be equal, but received X dimension[3](320) != Grid dimension[2](100) ``
(2)在微调时,一般都需要真实数据,如果使用合成数据,效果反而可能会有下降,PaddleOCR中放出的识别inference模型也是基于预训练模型在真实数据上微调得到的,效果提升比较明显;
**A**:TPS模块暂时无法支持变长的输入,请设置 ``--rec_image_shape='3,32,100' --rec_char_type='en' 固定输入shape``
(3)在训练的时候,文本长度超过25的训练图像都会被丢弃,因此需要看下真正参与训练的图像有多少,太少的话也容易过拟合。
#### Q3.3.20: 文字检测时怎么模糊的数据增强?
**A**: 模糊的数据增强需要修改代码进行添加,以DB为例,参考[Normalize](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/ppocr/data/imaug/operators.py#L60) ,添加模糊的增强就行
#### Q3.3.21: 文字检测时怎么更改图片旋转的角度,实现360度任意旋转?
**A**: 将[这里](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/ppocr/data/imaug/iaa_augment.py#L64) 的(-10,10) 改为(-180,180)即可
#### Q3.3.22: 训练数据的长宽比过大怎么修改shape
**A**: 识别修改[这里](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/configs/rec/ch_ppocr_v2.0/rec_chinese_lite_train_v2.0.yaml#L75) ,
检测修改[这里](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml#L85)
<a name="预测部署3"></a>
### 预测部署
......@@ -481,15 +658,56 @@ return paddle.reader.multiprocess_reader(readers, False, queue_size=320)
#### Q3.4.11:libopenblas.so找不到是什么意思?
**A**:目前包括mkl和openblas两种版本的预测库,推荐使用mkl的预测库,如果下载的预测库是mkl的,编译的时候也需要勾选`with_mkl`选项
,以Linux下编译为例,需要在设置这里为ON,`-DWITH_MKL=ON`[参考链接](https://github.com/PaddlePaddle/PaddleOCR/blob/8a78af26df0dd8f15b734cc8db13e25d2a3656a2/deploy/cpp_infer/tools/build.sh#L12)。此外,使用预测库时,推荐在Linux或者Windows上进行开发,不推荐在MacOS上开发。
,以Linux下编译为例,需要在设置这里为ON,`-DWITH_MKL=ON`[参考链接](https://github.com/PaddlePaddle/PaddleOCR/blob/569deedc41c2fa5e126a4d14b6c0c46a6bca43b8/deploy/cpp_infer/tools/build.sh#L12) 。此外,使用预测库时,推荐在Linux或者Windows上进行开发,不推荐在MacOS上开发。
#### Q3.4.12:使用自定义字典训练,inference时如何修改
**A**:使用了自定义字典的话,用inference预测时,需要通过 --rec_char_dict_path 修改字典路径。详细操作可参考[文档](https://github.com/PaddlePaddle/PaddleOCR/blob/develop/doc/doc_ch/inference.md#%E8%87%AA%E5%AE%9A%E4%B9%89%E6%96%87%E6%9C%AC%E8%AF%86%E5%88%AB%E5%AD%97%E5%85%B8%E7%9A%84%E6%8E%A8%E7%90%86)
**A**:使用了自定义字典的话,用inference预测时,需要通过 --rec_char_dict_path 修改字典路径。详细操作可参考[文档](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/doc/doc_ch/inference.md#4-%E8%87%AA%E5%AE%9A%E4%B9%89%E6%96%87%E6%9C%AC%E8%AF%86%E5%88%AB%E5%AD%97%E5%85%B8%E7%9A%84%E6%8E%A8%E7%90%86)
#### Q3.4.13:能否返回单字字符的位置?
**A**:训练的时候标注是整个文本行的标注,所以预测的也是文本行位置,如果要获取单字符位置信息,可以根据预测的文本,计算字符数量,再去根据整个文本行的位置信息,估计文本块中每个字符的位置。
#### Q3.4.14:PaddleOCR模型部署方式有哪几种?
**A**:目前有Inference部署,serving部署和手机端Paddle Lite部署,可根据不同场景做灵活的选择:Inference部署适用于本地离线部署,serving部署适用于云端部署,Paddle Lite部署适用于手机端集成。
#### Q3.4.15: hubserving、pdserving这两种部署方式区别是什么?
**A**:hubserving原本是paddlehub的配套服务部署工具,可以很方便的将paddlehub内置的模型部署为服务,paddleocr使用了这个功能,并将模型路径等参数暴露出来方便用户自定义修改。paddle serving是面向所有paddle模型的部署工具,文档中可以看到我们提供了快速版和标准版,其中快速版和hubserving的本质是一样的,而标准版基于rpc,更稳定,更适合分布式部署。
#### Q3.4.16: hub serving部署服务时如何多gpu同时利用起来,export CUDA_VISIBLE_DEVICES=0,1 方式吗?
**A**:hubserving的部署方式目前暂不支持多卡预测,除非手动启动多个serving,不同端口对应不同卡。或者可以使用paddleserving进行部署,部署工具已经发布:https://github.com/PaddlePaddle/PaddleOCR/tree/develop/deploy/pdserving ,在启动服务时--gpu_id 0,1 这样就可以
#### Q3.4.17: 预测内存泄漏问题
**A**:1. 使用hubserving出现内存泄漏,该问题为已知问题,预计在paddle2.0正式版中解决。相关讨论见[issue](https://github.com/PaddlePaddle/PaddleHub/issues/682)
**A**:2. C++ 预测出现内存泄漏,该问题已经在paddle2.0rc版本中解决,建议安装paddle2.0rc版本,并更新PaddleOCR代码到最新。
#### Q3.4.18:对于一些尺寸较大的文档类图片,在检测时会有较多的漏检,怎么避免这种漏检的问题呢?
**A**:PaddleOCR中在图像最长边大于960时,将图像等比例缩放为长边960的图像再进行预测,对于这种图像,可以通过修改det_limit_side_len,增大检测的最长边:[tools/infer/utility.py#L42](../../tools/infer/utility.py#L42)
#### Q3.4.19:在使用训练好的识别模型进行预测的时候,发现有很多重复的字,这个怎么解决呢?
**A**:可以看下训练的尺度和预测的尺度是否相同,如果训练的尺度为`[3, 32, 320]`,预测的尺度为`[3, 64, 640]`,则会有比较多的重复识别现象。
#### Q3.4.20:文档场景中,使用DB模型会出现整行漏检的情况应该怎么解决?
**A**:可以在预测时调小 det_db_box_thresh 阈值,默认为0.5, 可调小至0.3观察效果。
#### Q3.4.21:自己训练的det模型,在同一张图片上,inference模型与eval模型结果差别很大,为什么?
**A**:这是由于图片预处理不同造成的。如果训练的det模型图片输入并不是默认的shape[600, 600],eval的程序中图片预处理方式与train时一致
(由xxx_reader.yml中的test_image_shape参数决定缩放大小,但predict_eval.py中的图片预处理方式由程序里的preprocess_params决定,
最好不要传入max_side_len,而是传入和训练时一样大小的test_image_shape。
#### Q3.4.22:训练ccpd车牌数据集,训练集准确率高,测试均是错误的,这是什么原因?
**A**:这是因为训练时将shape修改为[3, 70, 220], 预测时对图片resize,会把高度压缩至32,影响测试结果。注释掉[resize代码](https://github.com/PaddlePaddle/PaddleOCR/blob/569deedc41c2fa5e126a4d14b6c0c46a6bca43b8/tools/infer/predict_rec.py#L56-L57) 可以解决问题。
#### Q3.4.23:安装paddleocr后,提示没有paddle
**A**:这是因为paddlepaddle gpu版本和cpu版本的名称不一致,现在已经在[whl的文档](./whl.md)里做了安装说明。
\ No newline at end of file
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment