Web page text recognition and extraction based on OCR

At present, various translation software and web-page built-in translation functions have greatly facilitated people’s lives. However, the text in webpage images cannot be directly translated because there are no relevant characters in the web-page source code. To make it easier for users to know the content of webpages when browsing non-native webpages, and to extract characters from webpage images more conveniently, this study intends to design a plug-in/program that uses image recognition technology in computer vision and calls the mainstream OCR interface on the Internet. When users select a specified area in the webpage with the mouse, the program will extract and translate the characters in the selected area, bringing users a better web browsing experience.