出版社：The Institute of Image Information and Television Engineers
摘要：This paper proposes a rate reduction method for high frame-rate picture coding based on repeatedly displaying sharp and blurred pictures. Our contribution is a picture coding method utilizing a spatiotemporal frequency model derived from human perceptual limitation of high frame-rate pictures. This method removes the high frequency parts of residual coefficients based on the spatiotemporal properties derived from local image data. Experimental results show that the proposed method achieves an average of 15% bit savings over the conventional MPEG-4 AVC/H.264 coding method without subjective degradation.
