Abstract: One aspect of the present invention discloses an image compression method based on local dynamic quantization. The method includes the following steps: (1) obtaining an image in a raw format, and parsing information about the obtained image; (2) performing image block partitioning on raw data collected from the image; (3) obtaining statistics of pixel values of the image blocks; (4) calculating a quantization step; (5) calculating an offset of each pixel; (6) performing quantization by: performing a division operation on the relative amplitude of each pixel of the current image block and the quantization step; and establishing a new gray-scale value; (7) reconstructing a block-based data structure; and (8) outputting a file. By the method, information of the raw data can be maintained to the greatest extent, and a data storage volume also can be reduced to meet the requirements on data recovery and reediting.

Abstract: One aspect of the present invention discloses an image compression method based on local dynamic quantization. The method includes the following steps: (1) obtaining an image in a raw format, and parsing information about the obtained image; (2) performing image block partitioning on raw data collected from the image; (3) obtaining statistics of pixel values of the image blocks; (4) calculating a quantization step; (5) calculating an offset of each pixel; (6) performing quantization by: performing a division operation on the relative amplitude of each pixel of the current image block and the quantization step; and establishing a new gray-scale value; (7) reconstructing a block-based data structure; and (8) outputting a file. By means of the method, information of the raw data can be maintained to the greatest extent, and a data storage volume also can be reduced to meet the requirements on data recovery and reediting.