Text Detecting And Removal In Video Image

With the advance of today’s digital process, multimedia and communication industrieshave been developed rapidly in the daily life. Digital video and digital image have graduallybecome the mainstream media of information exchange, a variety of digital video and imageexist everywhere, corresponding to the increasing amount of data. In order to meet thedifferent and growing needs, a variety of technologies of digital video images processing hasbeen made substantial progress.The current major source of digital video include digital movies, variety shows, medicalimages and video surveillance for other purposes. Techniques of these multimedia dataprocessing are being extensively studied, which have great commercial potential. In recentyears, in the field of image analysis and processing, text detection and image restorationhave been paid more attention. The study of this paper mainly focuses on the detecting andinpainting of the text in the video. My main work contains:(1) Firstly, this paper describes text detection algorithms which are commonly used,then compare and analyze the advantages and disadvantages of each method, on the basis ofthe problems faced by this paper; this paper propose a method combined with the informat-ion between frames in the text area detecting. Improve the accuracy of text detecting with thecontinuity of the video image sequences.(2) In the research on the image inpainting, this paper to introduce the classicCriminisi’s repair algorithm based on textures synthesis from sample. And then analyze thestructure of this algorithm and the results. On this basis, two key steps of the algorithm havebeen improved in this paper; by comparing the experiments of this improved algorithmprove the advantage in maintaining the structural and viewing.(3) In the research of video image sequence inpainting, this paper describes severalmethods which have been commonly used in video restoration, analyze and discuss thescope of application of various methods and the advantages and disadvantages. On this basis,the paper proposed a method for video inpainting based on the information between frames,focusing on continuous motion and consistent content in video sequences. Finally, differenttypes of video sequences have been chosen to verify the effectiveness of the algorithm; withthe comparative experiments between general methods and this improved algorithm, thismethod can guarantee a good result of the structure and textures in the image that has beenrepaired, and maintain the continuity of image sequence; make a smooth and natural view.