Method and device for detecting and extracting subtitle

The invention discloses a method for detecting and extracting a subtitle. The method comprises the steps of: detecting whether a subtitle exists in a frame of image; positioning a position of the subtitle and generating a frame including the subtitle; enhancing characters in a subtitle region; performing self-adaptive segmentation on the characters in the video subtitle region to form single characters; and subjecting the single characters obtained through segmentation to ORC identification so as to obtain text information related to a current video frame. The invention further discloses a device for detecting and extracting the subtitle. With the adoption of the method and the device provided by the invention, massive video programs can be classified and managed automatically, the manual storage/retrieval mode of the traditional video programs is changed, the efficiency is improved, and the accuracy rate of the extracted subtitle information is improved. 本发明公开了种字幕检测及提取的方法，包括：检测某帧图像是否有字幕存在；定位字幕的位置，并产生个包括字幕的框；