The Image Compression Method Based on Adaptive Segment and Adaptive Quantified

An image compression method based on adaptive segment and adaptive quantified is presented in this paper. The image is divided into smooth and detail areas using the multi-threshold-segmentation method based on the potential function. The quantization table varied with the different image. The 2-D DCT is decomposed into 1-D fast DCT based on base-2 FFT. The Compression-to-PSNR Rate can evaluate the image compression result more effectively than that of the compression rate or PSNR respectively. The experiment shows that the algorithm gives the superior performance compared with traditional JPEG.

本文提出了一种基于自适应分割和自适应量化的图像压缩方法。利用基于势函数的多阈值分割方法将图像划分为平滑区域和细节区域。量化表随图像不同而变化。基于基 - 2快速傅里叶变换将二维离散余弦变换分解为一维快速离散余弦变换。压缩比 - 峰值信噪比（Compression - to - PSNR Rate）比压缩率或峰值信噪比分别更能有效地评估图像压缩结果。实验表明，与传统的JPEG相比，该算法具有更优的性能。