论文部分内容阅读
介绍一种新的具有低复杂度的可用于实时传输的图像编码器,它使用了一种基于局域方差分析的边缘保护方法来提高被压缩图像的可视性和可识别性.这样的分析和压缩方法是通过将图像划分为多个块处理来实现的,所开发的能使边界效应最小化的提升小波滤波器组,通过提升系数的量化实现定点运算,并用位移和相加来取代乘法运算以实现运算复杂度的最小化.同时提出一种改进的快速SPIHT算法,通过使用更多的位来对小波系数进行编码和节省传输分类过程所需的位数以提高压缩性能,这种算法在可变的位码率中可减少系数的相关性.实验结果显示,这种编码对于观察到的压缩图像和进行量化的性能测量,均取得良好效果,同时还能为无线传输提供具有信道误差的弹性功能,最后还使用了带有随机误差的仿真传输信道来对这种功能进行评估和比较.
This paper introduces a new low complexity image encoder that can be used for real-time transmission, which uses a method of edge protection based on local variance analysis to improve the visibility and identifiability of compressed images. And the compression method is achieved by dividing the image into a plurality of block processes. The developed lifting wavelet filter set that minimizes the boundary effect is achieved by increasing the quantization of fixed coefficients, and replacing the multiplication by the displacement and the addition Computing to achieve the minimum computational complexity.An improved SPI SPI algorithm is also proposed to improve the compression performance by using more bits to encode the wavelet coefficients and save the number of bits needed to transmit the classification process The correlation between the coefficients can be reduced in a variable bit rate.The experimental results show that this coding has good effect on the observed compressed images and the performance measurement of quantization, at the same time, it can also provide a channel error for wireless transmission Finally, we also use a simulation transmission channel with random errors to evaluate and compare this function.