Research On Digital Video Format Conversion Algorithms

With the rapid development of digital technology, many video formats coexist in current video fields and the variety of video formats leads a new subject that it is necessary to convert video signals among various formats. After the simulations, comparisons and analysis, the author establishes the algorithm for de-interlacing and frame rate up conversion. Our research can be summarized as follows:Firstly, we have reviewed the existing de-interlacing algorithms and specifically analyzed their performances in details; hence found out that every algorithm has its own preference. Therefore we tried to combine all their capabilities using motion adaptive filter to gain relatively better performance in overall aspects.Secondly, we have reviewed the newly developed and dominating algorithm-3 dimension search block matching motion estimation. We point out its priority in vector insistence and easier implementation, which is taken as the basic algorithm in our following work. Considering widely existing global motion, the 3-D algorithm is not suitable, so we improve this algorithm using global motion parameter which has effect on the precision of motion vector in optimal matching criteria so as to reflect the global motion in better way.Thirdly, we have analyzed the deficiency of field duplicate and simple average motion compensation algorithm with conversion and pointed out that inaccurate procession to picture detail could results in artifacts, and then we developed a new algorithm based on the motion compensation interpolation.