基于小波图像编码算法的改进及实现

Implementation of Improved Embedded Zerotree Wavelet Image Coding Algorithm

针对现有SPIHT算法存在的不足,提出了一种新的改进编码方案。该算法通过引入图像平滑预处理和边缘检测,很好地提高了图像视觉效果和重构图像的PSNR值,并且进一步改进了编码传输策略,对最低频子带单独进行编码,对高频子带采用无链表SPIHT算法。实验结果表明:该算法有效地克服了SPIHT算法存在的不足,其编码速度和图像恢复质量,均优于SPIHT算法(特别是对纹理丰富的图像)。

In view of the insufficiency of existing SPIHT algorithm, this paper proposes a new improved coding scheme. By introducing image smoothing preprocessing and edge detection, the algorithm enhances the image vision effect and PSNR value of reconstruction image, and further improves coding transmission strategy. The lowest frequency sub-band coding is done alone, the high-frequency sub-band uses Listless SPIHT coding algorithm. The experimental results show that this algorithm can effectively overcome the insufficiency of SPIHT algorithm and its coding speed and image quality are better than that of SPIHT algorithm （especially to image with rich texture rich）.