An Improvement Method of Subjective Picture Quality within Concerned Regions for H.264 Video Coding

Video compression is very crucial method which is used to reduce the quantity of data used to represent digital video pictures. However, most video compression is loss when the video pictures data is quantized. So a quantizer is key factors for adjusting the bitrate and controlling the picture quality. Importantly, if the concerned region of video sequence takes on high level picture quality, the subjective picture quality could be improved. First, on the basis of the subjectively concerned region which can be determined using motion vectors within a video sequence, a classification method is suggested. Furthermore, we propose a method that allocation a relatively small quantization step-size for the concerned region compared with the other regions within a video sequence and thus improves picture quality. The measurement of the objective PSNR and the subjective image quality of result presented that subjective picture quality can be improved by proposed method.