Adaptation of frames for GOP using NSEW affine translation for video compression

Video compression standards follow Group of Pictures (GOP) which varies for every conventional standards. The frame identification plays a very important role in GOP. Instead of following the precoded GOP, based on some matching criteria if frames are chosen adaptively further processing of GOP could be effectively done. If GOP is dependent on frame identification the quality of the reconstructed video after compression can be improved subjectively. It is observed that B-frames takes more computation time since it depends on the previous and next frame to be processed. So, a NSEW affine translation technique is proposed to convert B-frames to either P-frames or I-frames depending on the similarity between adjacent frames. The result shows that the performance objective evaluation coincides with the subjective evaluation.