caozhong
Results
5
comments of
caozhong
他就是个简单的交叉熵吧，一共能检测四个角度（文字方向检测 0、90、180、270度检测（支持dnn/tensorflow）），简单分类就能做到
in this code: proc = cv2.resize(img,(.....),......),the parameter "img" is empty,you should check the code about load image
> @caozhongbjtu Hi, it shouldn't be that bad. > you can try to play with text_low, text_threshold, link_threshold to get best result depend on size of your input image. >...
> 我有，留个邮箱吧。 您好，这是我的邮箱[email protected]