Dynamic parameter control method for process data compression ratio
Jun Fang
Abstract
Jun Fang
Abstract
For lossy compression algorithm, such as SDT, box car, backward slope, it is difficult to find a suitable compression deviation △E . So compression deviation seriously limits the compression algorithm’s practicability and convenience. In view of this, it puts forward a dynamic parameter control method based on compressed ratio CR(CR=1- number of data compression/ number of data before compression). It assumes that the compression ratio CR and compression deviation have linear relation. Given the compression ratio CR dest , it can dynamically adjust △E so as to CR CR dest . The method makes compression ratio under control, and reduces the actual operation difficulty for related compression algorithm.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
For lossy compression algorithm, such as SDT, box car, backward slope, it is difficult to find a suitable compression deviation △E . So compression deviation seriously limits the compression algorithm’s practicability and convenience. In view of this, it puts forward a dynamic parameter control method based on compressed ratio CR(CR=1- number of data compression/ number of data before compression). It assumes that the compression ratio CR and compression deviation have linear relation. Given the compression ratio CR dest , it can dynamically adjust △E so as to CR CR dest . The method makes compression ratio under control, and reduces the actual operation difficulty for related compression algorithm.
Key concepts: Lossy compression, Compression ratio, Compression (physics), Data compression, Data compression ratio, Lossless compression, Algorithm, Computer science