控制图象灰度失真的高保真压缩算法

High Fidelity Compression Algorithm Based on Limiting Image Grey Error

为实现遥感图象的高保真压缩 ,在借鉴 JPEG- L S近无损压缩思想的基础上 ,提出了 3项改进措施 ,设计与实现了比 JPEG- L S压缩倍数高、图象恢复质量更好的视觉无失真压缩算法——“控制图象灰度失真的高保真压缩算法 (L IGE)”.实验结果表明 ,该算法既可限制图象最大灰度误差 ,又能控制恢复图象的峰值信噪比 ,从而有效地控制图象失真度 ,压缩倍数为 4时 ,数据处理速度与图象恢复质量两方面 ,均优于基于小波变换和嵌入式零树编码的 SPIHT算法 .该研究成果将对发展我国未来的高分辨率卫星、小卫星通信系统、星 -天 -地信息网提供有力的技术支撑 .

In order to resolve the contradiction between the need of high image quality and low data rates in data transmission and storage in the fields of satellite remote sensing, a new idea of using adaptive block coding technique and multi mode adaptive quantization technique to improve the JPEG LS is proposed. As a result, a new visually loss less coding algorithm-LIGE( Limiting Image Grey Error) is presented. The performance of this algorithm is much better than JPEG LS. Contrast to the DWT based SPIHT, LIGE has the following outstanding characters:Given a threshold Q, the image distortion and the PSNR of the reconstructed image can be predicted in advance, which can guarantee the required reconstructed image quality and avoid excessive loss of original image information. There is no floating point calculation and transform in the algorithm, so the compression speed of the coder is 3 times faster than that of SPIHT. At compression ratio 4:1, the PSNR of the reconstructed images of LIGE is higher than that of SPIHT, especially for the remote sensing image, the quality of the reconstructed image is even better.The result of this paper will be benefitial to the development of the communication system of Chinese satellites.