Research On Key Frames Extraction In Videos

The Key frames extracted from videos information need to be able to represent thesemantic information of the video. Key frame extraction is the basis of the followingimage processing and the critical step in image processing analysis from video to theimages, so it is significant to image processing. The extraction for Uyghur text frame isthe basis of location and recognition, the content is as follows:(1)According to key frame extraction is based on video shot detection, the articleproposed several video shot detection and key frame extraction algorithms.(2)Understanding the structural characteristics of Uyghur text, analyzing theircharacteristics of the extracted data.(3)The paper proposed a new text extraction method based on the baseline, themethod can quickly and efficiently extract the frames which contain the semanticcontent in the video, verified to be higher precision and accuracy.(4)Because of the redundancy of extracted text key frames which we have detectedand for the convenience of text location later, we reduce redundancy by taking theHarris corner feature of the text area matching algorithm.(5)To analyze the main features of the video information further, and test the effectof the text key frames extraction, we created a video Uyghur subtitles keyframeextraction system.In this paper,we extracted video text keyframes from video based on characteristicsof Uyghur. By analysing the basis features of Chinese, English, and Uyghur, weproposed a new frames extracted algorithm based on Uyghur Special features, andexperimental results show the algorithm have a good extraction rate.This researchwould play an important role in subsequent text positioning and identification ofUyghur,as the same time, it would have an extremely important significance in thedevelopment and dissemination of the minority languages western in China.