摘要
从某种意义来说,图像退化的过程可以等价为传输函数和噪声对原始图像矩阵进行线性变换,因而此像恢复过程可以等价于在了解传输函数的基础上,尝试由最小二乘给出解答。当传输函数可分离时,该问题可进一步转化为Toeplitz矩阵的求逆。通过仿真实验,验证了该方法的有效性并给出了算法的数值稳定性分析。
Image degradation process is equivalent to linear transformation of original image matrix processed by transfer function and noise,while image restoration process is equivalent to trying to get the original image using the least square method.When transfer function is separable,the problem is then transformed into finding the reverse matrix of Toeplitz matrix.Simulation results verify the validity of the method and analyze its numerical stability.
出处
《计算机与数字工程》
2009年第10期120-123,共4页
Computer & Digital Engineering
基金
国家"八六三"项目"基于多智能体的传感器网络协同目标跟踪技术研究"(编号:2007ADA299)资助
