Abstract:

Existing compression schemes usually scan an image by a fixed way. However, each image has its own characteristics, which makes it difficult to obtain a good coding performance with a fixed scanning manner. To solve this problem, an adaptive scanning-based compression method is proposed. Firstly, the wavelet transform is performed. Then the scanning order among the subbands is determined by the content of the image. Secondly, for the scanning of a subband, in order to preserving the texture information as much as possible, the scanning method is designed with the considering of the characteristic of the subband. Finally, the binary tree code is exploited to encode the one-dimensional coefficient sequence. Experiments show that, under the same conditions, the proposed compression method is superior to JPEG2000 and the state of the art compression methods based on binary tree coding adaptively (BTCA).