TABLE OF CONTENTS As with image coding, the objectives of video coding are bitrate reduction in storage and transmission by the use of both statistical and subjective redundancies, and to take advantage of probabilistic information using entropy coding techniques to reduce the symbol encoding length losslessly. There exists a practical tradeoff between coding gain (compression ratio) and video quality, as well as between the compression ratio and implementation complexity.
High video compression is achieved by degrading the video quality, as normally measured by the frame-by-frame (or temporal average of) peak signal-to-noise ratio (PSNR), Eq. (4.10), which gives the objective quality of the decoded image frames in relation to the original image frames prior to encoding. It should be noted that the degree of video quality degradation (due to both objective degradation and visible artifacts) depends on the complexity of the video scene as much as on the sophistication of the compression technique, i.e., for simple textures in image frames and low video activity a good video reconstruction with no visible artifacts may be achieved even with simple video compression techniques. In general, the compression techniques used in most video coding algorithms can be summarized as in Figure 5.3 [15].
Most simple techniques are used mainly for very-low-complexity operations or for lossless coding purposes. For example, the truncation technique can be adopted to reduce the number of bits used in representing...
