Always used tesseract to recognize Chinese Text, its recognition accuracy is very clear when the image is very clear, but as long as there are some noise, the accuracy rate will be very bad. Do not know ocropuse there is no ready-made example of Chinese recognition, I would like to test in a bad environment compared to the effect of tesseract, Thanks! :)