OCR PSEnet定位

github源码：

git clone https://github.com/WenmuZhou/PSENet.pytorch

git clone https://gitee.com/zhou_rx/PSENet.pytorch.git

pip install Polygon3，pip install colorlog

数据格式：

训练：

python train.py

问题：

PSENet.pytorch/pse$ make
......
g++: error: unrecognized command line option ‘-fno-plt’
Makefile:10: recipe for target 'pse.so' failed
make: *** [pse.so] Error 1

解决方法：

https://blog.csdn.net/forest_world/article/details/105883339

测试（需要继续训练和调参）：

def draw_bbox(img_path, result, color=(255, 0, 0),thickness=2):
    if isinstance(img_path, str):
        img_path = cv2.imread(img_path)
        # img_path = cv2.cvtColor(img_path, cv2.COLOR_BGR2RGB)
    img_path = img_path.copy()
    for point in result:
        point = point.astype(int)
        cv2.line(img_path, tuple(point[0]), tuple(point[1]), (0, 0, 255), thickness)
        cv2.line(img_path, tuple(point[1]), tuple(point[2]), (0, 255, 0), thickness)
        cv2.line(img_path, tuple(point[2]), tuple(point[3]), (255, 0, 0), thickness)
        cv2.line(img_path, tuple(point[3]), tuple(point[0]), (0, 0, 0), thickness)
    return img_path

                                                                                            测试图片来源百度