| 摘要 |
<p>Disclosed is a video encoding device for maintaining high image quality and increasing the compression rate of a video signal. In order to minimize the number of bits necessary for residual blocks, transform coefficients quantized according to the size of a transform encoding block are adaptively divided into a plurality of subsets, and quantization coefficients which are not 0 are scanned and encoded in each subset unit. Further, after valid reference pixels of a current block, which are adjacent to the current block, and generated reference pixels have been filtered, an intra prediction mode of the current block is determined by using the filtered reference pixels. Thus, an amount of data required for residual blocks of the current block is minimized.</p> |
