一种改进的SPIHT图像编码算法

在研究图像小波分解系数特点和SPIHT算法的基础上，提出了一种改进的SPIHT图像编码压缩算法.针对SPIHT算法没有利用同一子带相邻重要系数的相似性及在阈值为4、2、1这三层的大编码量，提出了在这三层编码用邻域内重要系数的绝对值均值代替各重要系数的绝对值来进行编解码；并且加深对大系数的重视程度，在前几个阈值时，对首次发现的重要系数进行三位二进制位的精细化.实验表明，在中、高压缩比情况下，本算法的重构图像峰值信噪比（PSNR）要优于原算法，并且视觉质量更好.

An improved image coding algorithm based on SPIHT

After the studying of image wavelet decomposing coefficients and SPTHT,an improved image coding algorithm is introduced based on SPTHT.It aims at the great coding amount in the scanning process of threshold 4,2 and 1,and proposes that,in those processes,average absolute value of adjacent domain's important coefficients is coded instead of each important coefficient of that adjacent domain,utilizing the similarity of adjacent important coefficients in the same sub-band.The improved algorithm also attaches much value to bigger coefficients.In the scanning process of bigger thresholds,it employs three bits for the refinement of the important coefficients,which are detected to be important for the first time.The experimental results show that the Peak Signal-to-Noise Ratio(PSNR) of the reconstructed image of the improved algorithm is better than that of the original algorithm,also with a better visual quality.