论文部分内容阅读
对图像压缩编码算法进行了改进。首先,将小波分解后的3个高频系数进行预处理:将高频部分进行球坐标变换,降低了同一尺度内系数的相关性;基于小波域和球坐标域的两个前提,定义了多尺度模积的概念,用来控制收缩函数对小波高频部分进行收缩处理。这样,可以去除那些不影响视觉效果的小波系数以及噪声信息,达到较高的压缩比。然后,对小波变换的低频部分进行单独编码(DPCM),对球坐标下的高频部分采用改进的多级树集合分裂(SPIHT)编码。针对SPIHT编码中重复扫描的问题,引入了最大值矩阵MMP(matrix of maximum pixel),这种策略能够有效降低比较次数。仿真实验表明,本文提出的算法具有较好的编码效率。
The image compression coding algorithm has been improved. Firstly, the three high-frequency coefficients after wavelet decomposition are preprocessed: the high-frequency part is transformed by spherical coordinates, which reduces the correlation of the coefficients in the same scale. Based on the two preconditions of wavelet domain and sphere coordinate domain, The concept of scale modular product is used to control the contraction function to shrink the high frequency part of the wavelet. In this way, you can remove those does not affect the visual effects of wavelet coefficients and noise information, to achieve a higher compression ratio. Then, the low frequency part of the wavelet transform is encoded separately (DPCM), and the high frequency part of the spherical coordinate is encoded by improved multi-level tree splitting (SPIHT). In order to solve the problem of repetitive scanning in SPIHT coding, the matrix of maximum pixel (MMP) is introduced. This strategy can effectively reduce the number of comparisons. Simulation results show that the proposed algorithm has better coding efficiency.