Research On Shape Error Concealment Techniques Of Video Object Based On MPEG-4

With the Internet and wireless network’s increasingly popular, video applicationsbecome more general, users pay more attention to the quality of video transmission. Asvideo coding is very sensitive to channel disturbances due to the use of compressionalgorithm, therefore, error concealment techniques are becoming increasingly importantfor getting high quality video image. The newest video coding standard MPEG-4 whichis based on video object has been widely used recently. Various methods have beenproposed for error concealment of this video coding standard. The error concealmenttechnique at the decoder plays an important role on reducing the impact on video bytransmission errors and improving the quality of reconstructed video pictures. Therefore,the research direction of this paperis the error concealment technique based on theMPEG-4 video coding standard.Firstly, the paper analyzes the importance and significance of the research on errorconcealment technology for video communication, and it summarizes thenewcharacteristics, error control and error concealment techniques of MPEG-4 videocoding standard, and follows with a summary of its performance advantages.Secondly, after a comprehensive study, the paper makes a general analysis of theerror concealment techniques, and divides the techniques based on MPEG-4 videocompression coding standard into three aspects: error concealment techniquesat theencoder, error concealment techniques at the decoder and error concealment techniquesusing the two aspects information, and then the paper focuses on the error concealmenttechniques at the decoder, and makes a general analysis of its all types. At the same time,the paper makes a detailed analysis on the shape error concealment which are lessresearched, respectively from two aspects of time and space domains, and simulates thealgorithms by experiments.Then, the paper does a research on the video object shape information error hidingtechniques, analyzes the existing algorithms in the spatial domains. On this basis, thepaper improves the traditional algorithms. As bezier curves can not be modified partly, the paper uses the spline curves instead of Bezier curves to curvefitting, and the fittingcurve is more close to the lost contour. And the recovered images through the simulationresults show that the improved method can obtain higher quality .Finally, By comparing the advantages and disadvantages of time domain andairspace shape error concealment algorithms, the paper combines existing algorithm toachieve adaptative options, depending on the type of the damaged areas. And it verifiesthe feasibility of the algorithm through experimental simulation.