Skip to content

GitLab

Commit 73f85035 authored by myhloli's avatar myhloli
Browse files

refactor: optimize OCR batch processing and enhance image cropping logic

parent 101b12a1
Show whitespace changes
Inline Side-by-side
Showing with 136 additions and 136 deletions
mineru/backend/pipeline/batch_analyze.py
View file @ 73f85035
...@@ -127,7 +127,7 @@ class BatchAnalyze: ...@@ -127,7 +127,7 @@ class BatchAnalyze:
if not lang_crop_list: if not lang_crop_list:
continue continue
# logger.info(f"Processing OCR detection for language {lang} with {len(lang_crop_list)} images") logger.info(f"Processing OCR detection for language {lang} with {len(lang_crop_list)} images")
# 获取OCR模型 # 获取OCR模型
ocr_model = atom_model_manager.get_atom_model( ocr_model = atom_model_manager.get_atom_model(
...@@ -171,7 +171,7 @@ class BatchAnalyze: ...@@ -171,7 +171,7 @@ class BatchAnalyze:
# 批处理检测 # 批处理检测
batch_size = min(len(batch_images), self.batch_ratio * 16) # 增加批处理大小 batch_size = min(len(batch_images), self.batch_ratio * 16) # 增加批处理大小
# logger.debug(f"OCR-det batch: {batch_size} images, target size: {target_h}x{target_w}") logger.debug(f"OCR-det batch: {batch_size} images, target size: {target_h}x{target_w}")
batch_results = ocr_model.text_detector.batch_predict(batch_images, batch_size) batch_results = ocr_model.text_detector.batch_predict(batch_images, batch_size)
# 处理批处理结果 # 处理批处理结果
......
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