Merge remote-tracking branch 'origin/dygraph' into dy1

deploy/cpp_infer/src/postprocess_op.cpp

@@ -13,6 +13,7 @@
 // limitations under the License.
 
 #include <include/postprocess_op.h>
+#include <include/clipper.cpp>
 
 namespace PaddleOCR {
 
@@ -159,6 +160,53 @@ std::vector<std::vector<float>> PostProcessor::GetMiniBoxes(cv::RotatedRect box,
   return array;
 }
 
+float PostProcessor::PolygonScoreAcc(std::vector<cv::Point> contour,
+                                     cv::Mat pred) {
+  int width = pred.cols;
+  int height = pred.rows;
+  std::vector<float> box_x;
+  std::vector<float> box_y;
+  for (int i = 0; i < contour.size(); ++i) {
+    box_x.push_back(contour[i].x);
+    box_y.push_back(contour[i].y);
+  }
+
+  int xmin =
+      clamp(int(std::floor(*(std::min_element(box_x.begin(), box_x.end())))), 0,
+            width - 1);
+  int xmax =
+      clamp(int(std::ceil(*(std::max_element(box_x.begin(), box_x.end())))), 0,
+            width - 1);
+  int ymin =
+      clamp(int(std::floor(*(std::min_element(box_y.begin(), box_y.end())))), 0,
+            height - 1);
+  int ymax =
+      clamp(int(std::ceil(*(std::max_element(box_y.begin(), box_y.end())))), 0,
+            height - 1);
+
+  cv::Mat mask;
+  mask = cv::Mat::zeros(ymax - ymin + 1, xmax - xmin + 1, CV_8UC1);
+
+
+  cv::Point* rook_point = new cv::Point[contour.size()];
+   
+  for (int i = 0; i < contour.size(); ++i) {
+    rook_point[i] = cv::Point(int(box_x[i]) - xmin, int(box_y[i]) - ymin);
+  }
+  const cv::Point *ppt[1] = {rook_point};
+  int npt[] = {int(contour.size())};
+
+
+  cv::fillPoly(mask, ppt, npt, 1, cv::Scalar(1));
+
+  cv::Mat croppedImg;
+  pred(cv::Rect(xmin, ymin, xmax - xmin + 1, ymax - ymin + 1)).copyTo(croppedImg);
+  float score = cv::mean(croppedImg, mask)[0];
+
+  delete []rook_point;
+  return score;
+}
+
 float PostProcessor::BoxScoreFast(std::vector<std::vector<float>> box_array,
                                   cv::Mat pred) {
   auto array = box_array;
@@ -197,10 +245,9 @@ float PostProcessor::BoxScoreFast(std::vector<std::vector<float>> box_array,
   return score;
 }
 
-std::vector<std::vector<std::vector<int>>>
-PostProcessor::BoxesFromBitmap(const cv::Mat pred, const cv::Mat bitmap,
-                               const float &box_thresh,
-                               const float &det_db_unclip_ratio) {
+std::vector<std::vector<std::vector<int>>> PostProcessor::BoxesFromBitmap(
+    const cv::Mat pred, const cv::Mat bitmap, const float &box_thresh,
+    const float &det_db_unclip_ratio, const bool &use_polygon_score) {
   const int min_size = 3;
   const int max_candidates = 1000;
 
@@ -234,7 +281,12 @@ PostProcessor::BoxesFromBitmap(const cv::Mat pred, const cv::Mat bitmap,
     }
 
     float score;
-    score = BoxScoreFast(array, pred);
+    if (use_polygon_score)
+      /* compute using polygon*/
+      score = PolygonScoreAcc(contours[_i], pred);
+    else
+      score = BoxScoreFast(array, pred);
+
     if (score < box_thresh)
       continue;
 

deploy/cpp_infer/src/preprocess_op.cpp

@@ -47,16 +47,13 @@ void Normalize::Run(cv::Mat *im, const std::vector<float> &mean,
     e /= 255.0;
   }
   (*im).convertTo(*im, CV_32FC3, e);
-  for (int h = 0; h < im->rows; h++) {
-    for (int w = 0; w < im->cols; w++) {
-      im->at<cv::Vec3f>(h, w)[0] =
-          (im->at<cv::Vec3f>(h, w)[0] - mean[0]) * scale[0];
-      im->at<cv::Vec3f>(h, w)[1] =
-          (im->at<cv::Vec3f>(h, w)[1] - mean[1]) * scale[1];
-      im->at<cv::Vec3f>(h, w)[2] =
-          (im->at<cv::Vec3f>(h, w)[2] - mean[2]) * scale[2];
-    }
+  std::vector<cv::Mat> bgr_channels(3);
+  cv::split(*im, bgr_channels);
+  for (auto i = 0; i < bgr_channels.size(); i++) {
+    bgr_channels[i].convertTo(bgr_channels[i], CV_32FC1, 1.0 * scale[i],
+                              (0.0 - mean[i]) * scale[i]);
   }
+  cv::merge(bgr_channels, *im);
 }
 
 void ResizeImgType0::Run(const cv::Mat &img, cv::Mat &resize_img,
@@ -77,28 +74,13 @@ void ResizeImgType0::Run(const cv::Mat &img, cv::Mat &resize_img,
 
   int resize_h = int(float(h) * ratio);
   int resize_w = int(float(w) * ratio);
-  if (resize_h % 32 == 0)
-    resize_h = resize_h;
-  else if (resize_h / 32 < 1 + 1e-5)
-    resize_h = 32;
-  else
-    resize_h = (resize_h / 32) * 32;
 
-  if (resize_w % 32 == 0)
-    resize_w = resize_w;
-  else if (resize_w / 32 < 1 + 1e-5)
-    resize_w = 32;
-  else
-    resize_w = (resize_w / 32) * 32;
-  if (!use_tensorrt) {
-    cv::resize(img, resize_img, cv::Size(resize_w, resize_h));
-    ratio_h = float(resize_h) / float(h);
-    ratio_w = float(resize_w) / float(w);
-  } else {
-    cv::resize(img, resize_img, cv::Size(640, 640));
-    ratio_h = float(640) / float(h);
-    ratio_w = float(640) / float(w);
-  }
+  resize_h = max(int(round(float(resize_h) / 32) * 32), 32);
+  resize_w = max(int(round(float(resize_w) / 32) * 32), 32);
+
+  cv::resize(img, resize_img, cv::Size(resize_w, resize_h));
+  ratio_h = float(resize_h) / float(h);
+  ratio_w = float(resize_w) / float(w);
 }
 
 void CrnnResizeImg::Run(const cv::Mat &img, cv::Mat &resize_img, float wh_ratio,
@@ -117,23 +99,12 @@ void CrnnResizeImg::Run(const cv::Mat &img, cv::Mat &resize_img, float wh_ratio,
     resize_w = imgW;
   else
     resize_w = int(ceilf(imgH * ratio));
-  if (!use_tensorrt) {
-    cv::resize(img, resize_img, cv::Size(resize_w, imgH), 0.f, 0.f,
-               cv::INTER_LINEAR);
-    cv::copyMakeBorder(resize_img, resize_img, 0, 0, 0,
-                       int(imgW - resize_img.cols), cv::BORDER_CONSTANT,
-                       {127, 127, 127});
-  } else {
-    int k = int(img.cols * 32 / img.rows);
-    if (k >= 100) {
-      cv::resize(img, resize_img, cv::Size(100, 32), 0.f, 0.f,
-                 cv::INTER_LINEAR);
-    } else {
-      cv::resize(img, resize_img, cv::Size(k, 32), 0.f, 0.f, cv::INTER_LINEAR);
-      cv::copyMakeBorder(resize_img, resize_img, 0, 0, 0, int(100 - k),
-                         cv::BORDER_CONSTANT, {127, 127, 127});
-    }
-  }
+
+  cv::resize(img, resize_img, cv::Size(resize_w, imgH), 0.f, 0.f,
+             cv::INTER_LINEAR);
+  cv::copyMakeBorder(resize_img, resize_img, 0, 0, 0,
+                     int(imgW - resize_img.cols), cv::BORDER_CONSTANT,
+                     {127, 127, 127});
 }
 
 void ClsResizeImg::Run(const cv::Mat &img, cv::Mat &resize_img,
@@ -151,15 +122,11 @@ void ClsResizeImg::Run(const cv::Mat &img, cv::Mat &resize_img,
   else
     resize_w = int(ceilf(imgH * ratio));
 
-  if (!use_tensorrt) {
-    cv::resize(img, resize_img, cv::Size(resize_w, imgH), 0.f, 0.f,
-               cv::INTER_LINEAR);
-    if (resize_w < imgW) {
-      cv::copyMakeBorder(resize_img, resize_img, 0, 0, 0, imgW - resize_w,
-                         cv::BORDER_CONSTANT, cv::Scalar(0, 0, 0));
-    }
-  } else {
-    cv::resize(img, resize_img, cv::Size(100, 32), 0.f, 0.f, cv::INTER_LINEAR);
+  cv::resize(img, resize_img, cv::Size(resize_w, imgH), 0.f, 0.f,
+             cv::INTER_LINEAR);
+  if (resize_w < imgW) {
+    cv::copyMakeBorder(resize_img, resize_img, 0, 0, 0, imgW - resize_w,
+                       cv::BORDER_CONSTANT, cv::Scalar(0, 0, 0));
   }
 }
 

deploy/cpp_infer/src/utility.cpp

@@ -12,12 +12,14 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
+#include <dirent.h>
+#include <include/utility.h>
 #include <iostream>
 #include <ostream>
+#include <sys/stat.h>
+#include <sys/types.h>
 #include <vector>
 
-#include <include/utility.h>
-
 namespace PaddleOCR {
 
 std::vector<std::string> Utility::ReadDict(const std::string &path) {
@@ -57,4 +59,92 @@ void Utility::VisualizeBboxes(
             << std::endl;
 }
 
+// list all files under a directory
+void Utility::GetAllFiles(const char *dir_name,
+                          std::vector<std::string> &all_inputs) {
+  if (NULL == dir_name) {
+    std::cout << " dir_name is null ! " << std::endl;
+    return;
+  }
+  struct stat s;
+  lstat(dir_name, &s);
+  if (!S_ISDIR(s.st_mode)) {
+    std::cout << "dir_name is not a valid directory !" << std::endl;
+    all_inputs.push_back(dir_name);
+    return;
+  } else {
+    struct dirent *filename; // return value for readdir()
+    DIR *dir;                // return value for opendir()
+    dir = opendir(dir_name);
+    if (NULL == dir) {
+      std::cout << "Can not open dir " << dir_name << std::endl;
+      return;
+    }
+    std::cout << "Successfully opened the dir !" << std::endl;
+    while ((filename = readdir(dir)) != NULL) {
+      if (strcmp(filename->d_name, ".") == 0 ||
+          strcmp(filename->d_name, "..") == 0)
+        continue;
+      // img_dir + std::string("/") + all_inputs[0];
+      all_inputs.push_back(dir_name + std::string("/") +
+                           std::string(filename->d_name));
+    }
+  }
+}
+
+cv::Mat Utility::GetRotateCropImage(const cv::Mat &srcimage,
+                            std::vector<std::vector<int>> box) {
+  cv::Mat image;
+  srcimage.copyTo(image);
+  std::vector<std::vector<int>> points = box;
+
+  int x_collect[4] = {box[0][0], box[1][0], box[2][0], box[3][0]};
+  int y_collect[4] = {box[0][1], box[1][1], box[2][1], box[3][1]};
+  int left = int(*std::min_element(x_collect, x_collect + 4));
+  int right = int(*std::max_element(x_collect, x_collect + 4));
+  int top = int(*std::min_element(y_collect, y_collect + 4));
+  int bottom = int(*std::max_element(y_collect, y_collect + 4));
+
+  cv::Mat img_crop;
+  image(cv::Rect(left, top, right - left, bottom - top)).copyTo(img_crop);
+
+  for (int i = 0; i < points.size(); i++) {
+    points[i][0] -= left;
+    points[i][1] -= top;
+  }
+
+  int img_crop_width = int(sqrt(pow(points[0][0] - points[1][0], 2) +
+                                pow(points[0][1] - points[1][1], 2)));
+  int img_crop_height = int(sqrt(pow(points[0][0] - points[3][0], 2) +
+                                 pow(points[0][1] - points[3][1], 2)));
+
+  cv::Point2f pts_std[4];
+  pts_std[0] = cv::Point2f(0., 0.);
+  pts_std[1] = cv::Point2f(img_crop_width, 0.);
+  pts_std[2] = cv::Point2f(img_crop_width, img_crop_height);
+  pts_std[3] = cv::Point2f(0.f, img_crop_height);
+
+  cv::Point2f pointsf[4];
+  pointsf[0] = cv::Point2f(points[0][0], points[0][1]);
+  pointsf[1] = cv::Point2f(points[1][0], points[1][1]);
+  pointsf[2] = cv::Point2f(points[2][0], points[2][1]);
+  pointsf[3] = cv::Point2f(points[3][0], points[3][1]);
+
+  cv::Mat M = cv::getPerspectiveTransform(pointsf, pts_std);
+
+  cv::Mat dst_img;
+  cv::warpPerspective(img_crop, dst_img, M,
+                      cv::Size(img_crop_width, img_crop_height),
+                      cv::BORDER_REPLICATE);
+
+  if (float(dst_img.rows) >= float(dst_img.cols) * 1.5) {
+    cv::Mat srcCopy = cv::Mat(dst_img.rows, dst_img.cols, dst_img.depth());
+    cv::transpose(dst_img, srcCopy);
+    cv::flip(srcCopy, srcCopy, 0);
+    return srcCopy;
+  } else {
+    return dst_img;
+  }
+}
+
 } // namespace PaddleOCR
\ No newline at end of file

deploy/cpp_infer/tools/build.sh

@@ -12,9 +12,10 @@ cmake .. \
     -DWITH_MKL=ON \
     -DWITH_GPU=OFF \
     -DWITH_STATIC_LIB=OFF \
-    -DUSE_TENSORRT=OFF \
+    -DWITH_TENSORRT=OFF \
     -DOPENCV_DIR=${OPENCV_DIR} \
     -DCUDNN_LIB=${CUDNN_LIB_DIR} \
     -DCUDA_LIB=${CUDA_LIB_DIR} \
+    -DTENSORRT_DIR=${TENSORRT_DIR} \
 
 make -j

deploy/cpp_infer/tools/config.txt

-# model load config
-use_gpu 0
-gpu_id  0
-gpu_mem  4000
-cpu_math_library_num_threads  10
-use_mkldnn 0
-
-# det config
-max_side_len  960
-det_db_thresh  0.3
-det_db_box_thresh  0.5
-det_db_unclip_ratio  1.6
-det_model_dir  ./inference/ch_ppocr_mobile_v2.0_det_infer/
-
-# cls config
-use_angle_cls 0
-cls_model_dir  ./inference/ch_ppocr_mobile_v2.0_cls_infer/
-cls_thresh  0.9
-
-# rec config
-rec_model_dir  ./inference/ch_ppocr_mobile_v2.0_rec_infer/
-char_list_file ../../ppocr/utils/ppocr_keys_v1.txt
-
-# show the detection results
-visualize 1
-
-# use_tensorrt
-use_tensorrt 0
-use_fp16   0
-

deploy/cpp_infer/tools/run.sh

-
-./build/ocr_system ./tools/config.txt ../../doc/imgs/12.jpg

deploy/hubserving/ocr_cls/module.py

@@ -6,6 +6,7 @@ from __future__ import print_function
 import os
 import sys
 sys.path.insert(0, ".")
+import copy
 
 from paddlehub.common.logger import logger
 from paddlehub.module.module import moduleinfo, runnable, serving
@@ -14,6 +15,8 @@ import paddlehub as hub
 
 from tools.infer.utility import base64_to_cv2
 from tools.infer.predict_cls import TextClassifier
+from tools.infer.utility import parse_args
+from deploy.hubserving.ocr_cls.params import read_params
 
 
 @moduleinfo(
@@ -28,8 +31,7 @@ class OCRCls(hub.Module):
         """
         initialize with the necessary elements
         """
-        from ocr_cls.params import read_params
-        cfg = read_params()
+        cfg = self.merge_configs()
 
         cfg.use_gpu = use_gpu
         if use_gpu:
@@ -48,6 +50,20 @@ class OCRCls(hub.Module):
 
         self.text_classifier = TextClassifier(cfg)
 
+    def merge_configs(self, ):
+        # deafult cfg
+        backup_argv = copy.deepcopy(sys.argv)
+        sys.argv = sys.argv[:1]
+        cfg = parse_args()
+
+        update_cfg_map = vars(read_params())
+
+        for key in update_cfg_map:
+            cfg.__setattr__(key, update_cfg_map[key])
+
+        sys.argv = copy.deepcopy(backup_argv)
+        return cfg
+
     def read_images(self, paths=[]):
         images = []
         for img_path in paths:

deploy/hubserving/ocr_det/module.py

@@ -7,6 +7,8 @@ import os
 import sys
 sys.path.insert(0, ".")
 
+import copy
+
 from paddlehub.common.logger import logger
 from paddlehub.module.module import moduleinfo, runnable, serving
 import cv2
@@ -15,6 +17,8 @@ import paddlehub as hub
 
 from tools.infer.utility import base64_to_cv2
 from tools.infer.predict_det import TextDetector
+from tools.infer.utility import parse_args
+from deploy.hubserving.ocr_system.params import read_params
 
 
 @moduleinfo(
@@ -29,8 +33,7 @@ class OCRDet(hub.Module):
         """
         initialize with the necessary elements
         """
-        from ocr_det.params import read_params
-        cfg = read_params()
+        cfg = self.merge_configs()
 
         cfg.use_gpu = use_gpu
         if use_gpu:
@@ -49,6 +52,20 @@ class OCRDet(hub.Module):
 
         self.text_detector = TextDetector(cfg)
 
+    def merge_configs(self, ):
+        # deafult cfg
+        backup_argv = copy.deepcopy(sys.argv)
+        sys.argv = sys.argv[:1]
+        cfg = parse_args()
+
+        update_cfg_map = vars(read_params())
+
+        for key in update_cfg_map:
+            cfg.__setattr__(key, update_cfg_map[key])
+
+        sys.argv = copy.deepcopy(backup_argv)
+        return cfg
+
     def read_images(self, paths=[]):
         images = []
         for img_path in paths:

deploy/hubserving/ocr_det/params.py

@@ -22,6 +22,7 @@ def read_params():
     cfg.det_db_box_thresh = 0.5
     cfg.det_db_unclip_ratio = 1.6
     cfg.use_dilation = False
+    cfg.det_db_score_mode = "fast"
 
     # #EAST parmas
     # cfg.det_east_score_thresh = 0.8

deploy/hubserving/ocr_rec/module.py

@@ -6,6 +6,7 @@ from __future__ import print_function
 import os
 import sys
 sys.path.insert(0, ".")
+import copy
 
 from paddlehub.common.logger import logger
 from paddlehub.module.module import moduleinfo, runnable, serving
@@ -14,6 +15,8 @@ import paddlehub as hub
 
 from tools.infer.utility import base64_to_cv2
 from tools.infer.predict_rec import TextRecognizer
+from tools.infer.utility import parse_args
+from deploy.hubserving.ocr_rec.params import read_params
 
 
 @moduleinfo(
@@ -28,8 +31,7 @@ class OCRRec(hub.Module):
         """
         initialize with the necessary elements
         """
-        from ocr_rec.params import read_params
-        cfg = read_params()
+        cfg = self.merge_configs()
 
         cfg.use_gpu = use_gpu
         if use_gpu:
@@ -48,6 +50,20 @@ class OCRRec(hub.Module):
 
         self.text_recognizer = TextRecognizer(cfg)
 
+    def merge_configs(self, ):
+        # deafult cfg
+        backup_argv = copy.deepcopy(sys.argv)
+        sys.argv = sys.argv[:1]
+        cfg = parse_args()
+
+        update_cfg_map = vars(read_params())
+
+        for key in update_cfg_map:
+            cfg.__setattr__(key, update_cfg_map[key])
+
+        sys.argv = copy.deepcopy(backup_argv)
+        return cfg
+
     def read_images(self, paths=[]):
         images = []
         for img_path in paths:

deploy/hubserving/ocr_system/module.py

@@ -6,6 +6,7 @@ from __future__ import print_function
 import os
 import sys
 sys.path.insert(0, ".")
+import copy
 
 import time
 
@@ -17,6 +18,8 @@ import paddlehub as hub
 
 from tools.infer.utility import base64_to_cv2
 from tools.infer.predict_system import TextSystem
+from tools.infer.utility import parse_args
+from deploy.hubserving.ocr_system.params import read_params
 
 
 @moduleinfo(
@@ -31,8 +34,7 @@ class OCRSystem(hub.Module):
         """
         initialize with the necessary elements
         """
-        from ocr_system.params import read_params
-        cfg = read_params()
+        cfg = self.merge_configs()
 
         cfg.use_gpu = use_gpu
         if use_gpu:
@@ -51,6 +53,20 @@ class OCRSystem(hub.Module):
 
         self.text_sys = TextSystem(cfg)
 
+    def merge_configs(self, ):
+        # deafult cfg
+        backup_argv = copy.deepcopy(sys.argv)
+        sys.argv = sys.argv[:1]
+        cfg = parse_args()
+
+        update_cfg_map = vars(read_params())
+
+        for key in update_cfg_map:
+            cfg.__setattr__(key, update_cfg_map[key])
+
+        sys.argv = copy.deepcopy(backup_argv)
+        return cfg
+
     def read_images(self, paths=[]):
         images = []
         for img_path in paths:

deploy/hubserving/ocr_system/params.py

@@ -22,6 +22,7 @@ def read_params():
     cfg.det_db_box_thresh = 0.5
     cfg.det_db_unclip_ratio = 1.6
     cfg.use_dilation = False
+    cfg.det_db_score_mode = "fast"
 
     #EAST parmas
     cfg.det_east_score_thresh = 0.8

deploy/hubserving/readme.md

@@ -29,7 +29,8 @@ deploy/hubserving/ocr_system/
 ### 1. 准备环境
 ```shell
 # 安装paddlehub  
-pip3 install paddlehub --upgrade -i https://pypi.tuna.tsinghua.edu.cn/simple
+# paddlehub 需要 python>3.6.2
+pip3 install paddlehub==2.1.0 --upgrade -i https://pypi.tuna.tsinghua.edu.cn/simple
 ```
 
 ### 2. 下载推理模型

deploy/hubserving/readme_en.md

@@ -30,7 +30,8 @@ The following steps take the 2-stage series service as an example. If only the d
 ### 1. Prepare the environment
 ```shell
 # Install paddlehub  
-pip3 install paddlehub --upgrade -i https://pypi.tuna.tsinghua.edu.cn/simple
+# python>3.6.2 is required bt paddlehub
+pip3 install paddlehub==2.1.0 --upgrade -i https://pypi.tuna.tsinghua.edu.cn/simple
 ```
 
 ### 2. Download inference model

deploy/lite/Makefile

+ARM_ABI = arm8
+export ARM_ABI
+
+include ../Makefile.def
+
+LITE_ROOT=../../../
+
+THIRD_PARTY_DIR=${LITE_ROOT}/third_party
+
+OPENCV_VERSION=opencv4.1.0
+
+OPENCV_LIBS = ../../../third_party/${OPENCV_VERSION}/arm64-v8a/libs/libopencv_imgcodecs.a \
+              ../../../third_party/${OPENCV_VERSION}/arm64-v8a/libs/libopencv_imgproc.a \
+              ../../../third_party/${OPENCV_VERSION}/arm64-v8a/libs/libopencv_core.a \
+              ../../../third_party/${OPENCV_VERSION}/arm64-v8a/3rdparty/libs/libtegra_hal.a \
+              ../../../third_party/${OPENCV_VERSION}/arm64-v8a/3rdparty/libs/liblibjpeg-turbo.a \
+              ../../../third_party/${OPENCV_VERSION}/arm64-v8a/3rdparty/libs/liblibwebp.a \
+              ../../../third_party/${OPENCV_VERSION}/arm64-v8a/3rdparty/libs/liblibpng.a \
+              ../../../third_party/${OPENCV_VERSION}/arm64-v8a/3rdparty/libs/liblibjasper.a \
+              ../../../third_party/${OPENCV_VERSION}/arm64-v8a/3rdparty/libs/liblibtiff.a \
+              ../../../third_party/${OPENCV_VERSION}/arm64-v8a/3rdparty/libs/libIlmImf.a \
+              ../../../third_party/${OPENCV_VERSION}/arm64-v8a/3rdparty/libs/libtbb.a \
+              ../../../third_party/${OPENCV_VERSION}/arm64-v8a/3rdparty/libs/libcpufeatures.a
+
+OPENCV_INCLUDE = -I../../../third_party/${OPENCV_VERSION}/arm64-v8a/include
+
+CXX_INCLUDES = $(INCLUDES) ${OPENCV_INCLUDE} -I$(LITE_ROOT)/cxx/include
+
+CXX_LIBS = ${OPENCV_LIBS} -L$(LITE_ROOT)/cxx/lib/ -lpaddle_light_api_shared $(SYSTEM_LIBS)
+
+###############################################################
+# How to use one of static libaray:                           #
+#  `libpaddle_api_full_bundled.a`                             #
+#  `libpaddle_api_light_bundled.a`                            #
+###############################################################
+# Note: default use lite's shared library.                    #
+###############################################################
+# 1. Comment above line using `libpaddle_light_api_shared.so`
+# 2. Undo comment below line using `libpaddle_api_light_bundled.a`
+
+#CXX_LIBS = $(LITE_ROOT)/cxx/lib/libpaddle_api_light_bundled.a $(SYSTEM_LIBS)
+
+ocr_db_crnn: fetch_clipper fetch_opencv ocr_db_crnn.o crnn_process.o db_post_process.o clipper.o cls_process.o
+	$(CC) $(SYSROOT_LINK) $(CXXFLAGS_LINK) ocr_db_crnn.o crnn_process.o db_post_process.o clipper.o cls_process.o -o ocr_db_crnn  $(CXX_LIBS) $(LDFLAGS)
+
+ocr_db_crnn.o: ocr_db_crnn.cc
+	$(CC) $(SYSROOT_COMPLILE) $(CXX_DEFINES) $(CXX_INCLUDES) $(CXX_FLAGS) -o ocr_db_crnn.o -c ocr_db_crnn.cc
+
+crnn_process.o: fetch_opencv crnn_process.cc
+	$(CC) $(SYSROOT_COMPLILE) $(CXX_DEFINES) $(CXX_INCLUDES) $(CXX_FLAGS) -o crnn_process.o -c crnn_process.cc
+
+cls_process.o: fetch_opencv cls_process.cc
+	$(CC) $(SYSROOT_COMPLILE) $(CXX_DEFINES) $(CXX_INCLUDES) $(CXX_FLAGS) -o cls_process.o -c cls_process.cc
+
+db_post_process.o: fetch_clipper fetch_opencv db_post_process.cc
+	$(CC) $(SYSROOT_COMPLILE) $(CXX_DEFINES) $(CXX_INCLUDES) $(CXX_FLAGS) -o db_post_process.o -c db_post_process.cc
+
+clipper.o: fetch_clipper
+	$(CC) $(SYSROOT_COMPLILE) $(CXX_DEFINES) $(CXX_INCLUDES) $(CXX_FLAGS) -o clipper.o -c clipper.cpp
+
+fetch_clipper:
+	@test -e clipper.hpp || \
+        ( echo "Fetch clipper " && \
+        wget -c https://paddle-inference-dist.cdn.bcebos.com/PaddleLite/Clipper/clipper.hpp)
+	@ test -e clipper.cpp || \
+        wget -c https://paddle-inference-dist.cdn.bcebos.com/PaddleLite/Clipper/clipper.cpp
+
+fetch_opencv:
+	@ test -d ${THIRD_PARTY_DIR} ||  mkdir ${THIRD_PARTY_DIR}
+	@ test -e ${THIRD_PARTY_DIR}/${OPENCV_VERSION}.tar.gz || \
+      (echo "fetch opencv libs" && \
+      wget -P ${THIRD_PARTY_DIR} https://paddle-inference-dist.bj.bcebos.com/${OPENCV_VERSION}.tar.gz)
+	@ test -d ${THIRD_PARTY_DIR}/${OPENCV_VERSION} || \
+      tar -zxvf ${THIRD_PARTY_DIR}/${OPENCV_VERSION}.tar.gz -C ${THIRD_PARTY_DIR}
+
+
+.PHONY: clean
+clean:
+	rm -f ocr_db_crnn.o clipper.o db_post_process.o crnn_process.o cls_process.o
+	rm -f ocr_db_crnn

deploy/lite/cls_process.cc

+// Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+#include "cls_process.h" //NOLINT
+#include <algorithm>
+#include <memory>
+#include <string>
+
+const std::vector<int> rec_image_shape{3, 48, 192};
+
+cv::Mat ClsResizeImg(cv::Mat img) {
+  int imgC, imgH, imgW;
+  imgC = rec_image_shape[0];
+  imgH = rec_image_shape[1];
+  imgW = rec_image_shape[2];
+
+  float ratio = static_cast<float>(img.cols) / static_cast<float>(img.rows);
+
+  int resize_w, resize_h;
+  if (ceilf(imgH * ratio) > imgW)
+    resize_w = imgW;
+  else
+    resize_w = int(ceilf(imgH * ratio));
+  cv::Mat resize_img;
+  cv::resize(img, resize_img, cv::Size(resize_w, imgH), 0.f, 0.f,
+             cv::INTER_LINEAR);
+  if (resize_w < imgW) {
+    cv::copyMakeBorder(resize_img, resize_img, 0, 0, 0, imgW - resize_w,
+                       cv::BORDER_CONSTANT, cv::Scalar(0, 0, 0));
+  }
+  return resize_img;
+}
\ No newline at end of file

deploy/lite/cls_process.h

+// Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+#pragma once
+
+#include <cstring>
+#include <fstream>
+#include <iostream>
+#include <memory>
+#include <string>
+#include <vector>
+
+#include "math.h" //NOLINT
+#include "opencv2/core.hpp"
+#include "opencv2/imgcodecs.hpp"
+#include "opencv2/imgproc.hpp"
+
+cv::Mat ClsResizeImg(cv::Mat img);
\ No newline at end of file

deploy/lite/config.txt

+max_side_len  960
+det_db_thresh  0.3
+det_db_box_thresh  0.5
+det_db_unclip_ratio  1.6
+det_db_use_dilate 0
+det_use_polygon_score 1
+use_direction_classify  1
\ No newline at end of file

deploy/lite/crnn_process.cc

+// Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+#include "crnn_process.h" //NOLINT
+#include <algorithm>
+#include <memory>
+#include <string>
+
+const std::vector<int> rec_image_shape{3, 32, 320};
+
+cv::Mat CrnnResizeImg(cv::Mat img, float wh_ratio) {
+  int imgC, imgH, imgW;
+  imgC = rec_image_shape[0];
+  imgW = rec_image_shape[2];
+  imgH = rec_image_shape[1];
+
+  imgW = int(32 * wh_ratio);
+
+  float ratio = static_cast<float>(img.cols) / static_cast<float>(img.rows);
+  int resize_w, resize_h;
+  if (ceilf(imgH * ratio) > imgW)
+    resize_w = imgW;
+  else
+    resize_w = static_cast<int>(ceilf(imgH * ratio));
+  cv::Mat resize_img;
+  cv::resize(img, resize_img, cv::Size(resize_w, imgH), 0.f, 0.f,
+             cv::INTER_LINEAR);
+
+  return resize_img;
+}
+
+std::vector<std::string> ReadDict(std::string path) {
+  std::ifstream in(path);
+  std::string filename;
+  std::string line;
+  std::vector<std::string> m_vec;
+  if (in) {
+    while (getline(in, line)) {
+      m_vec.push_back(line);
+    }
+  } else {
+    std::cout << "no such file" << std::endl;
+  }
+  return m_vec;
+}
+
+cv::Mat GetRotateCropImage(cv::Mat srcimage,
+                           std::vector<std::vector<int>> box) {
+  cv::Mat image;
+  srcimage.copyTo(image);
+  std::vector<std::vector<int>> points = box;
+
+  int x_collect[4] = {box[0][0], box[1][0], box[2][0], box[3][0]};
+  int y_collect[4] = {box[0][1], box[1][1], box[2][1], box[3][1]};
+  int left = int(*std::min_element(x_collect, x_collect + 4));
+  int right = int(*std::max_element(x_collect, x_collect + 4));
+  int top = int(*std::min_element(y_collect, y_collect + 4));
+  int bottom = int(*std::max_element(y_collect, y_collect + 4));
+
+  cv::Mat img_crop;
+  image(cv::Rect(left, top, right - left, bottom - top)).copyTo(img_crop);
+
+  for (int i = 0; i < points.size(); i++) {
+    points[i][0] -= left;
+    points[i][1] -= top;
+  }
+
+  int img_crop_width =
+      static_cast<int>(sqrt(pow(points[0][0] - points[1][0], 2) +
+                            pow(points[0][1] - points[1][1], 2)));
+  int img_crop_height =
+      static_cast<int>(sqrt(pow(points[0][0] - points[3][0], 2) +
+                            pow(points[0][1] - points[3][1], 2)));
+
+  cv::Point2f pts_std[4];
+  pts_std[0] = cv::Point2f(0., 0.);
+  pts_std[1] = cv::Point2f(img_crop_width, 0.);
+  pts_std[2] = cv::Point2f(img_crop_width, img_crop_height);
+  pts_std[3] = cv::Point2f(0.f, img_crop_height);
+
+  cv::Point2f pointsf[4];
+  pointsf[0] = cv::Point2f(points[0][0], points[0][1]);
+  pointsf[1] = cv::Point2f(points[1][0], points[1][1]);
+  pointsf[2] = cv::Point2f(points[2][0], points[2][1]);
+  pointsf[3] = cv::Point2f(points[3][0], points[3][1]);
+
+  cv::Mat M = cv::getPerspectiveTransform(pointsf, pts_std);
+
+  cv::Mat dst_img;
+  cv::warpPerspective(img_crop, dst_img, M,
+                      cv::Size(img_crop_width, img_crop_height),
+                      cv::BORDER_REPLICATE);
+
+  const float ratio = 1.5;
+  if (static_cast<float>(dst_img.rows) >=
+      static_cast<float>(dst_img.cols) * ratio) {
+    cv::Mat srcCopy = cv::Mat(dst_img.rows, dst_img.cols, dst_img.depth());
+    cv::transpose(dst_img, srcCopy);
+    cv::flip(srcCopy, srcCopy, 0);
+    return srcCopy;
+  } else {
+    return dst_img;
+  }
+}

deploy/lite/crnn_process.h

+// Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+#pragma once
+
+#include <cstring>
+#include <fstream>
+#include <iostream>
+#include <memory>
+#include <string>
+#include <vector>
+
+#include "math.h" //NOLINT
+#include "opencv2/core.hpp"
+#include "opencv2/imgcodecs.hpp"
+#include "opencv2/imgproc.hpp"
+
+cv::Mat CrnnResizeImg(cv::Mat img, float wh_ratio);
+
+std::vector<std::string> ReadDict(std::string path);
+
+cv::Mat GetRotateCropImage(cv::Mat srcimage, std::vector<std::vector<int>> box);
+
+template <class ForwardIterator>
+inline size_t Argmax(ForwardIterator first, ForwardIterator last) {
+  return std::distance(first, std::max_element(first, last));
+}