论文部分内容阅读
首先分析了传统的非均匀性校正方法的缺点,指出自适应校正红外焦平面器件非均匀性的必要性。根据焦平面器件的非均匀性噪声特性和算法研究的需要,介绍了非均匀性失真图像的产生方法。在上述工作的基础上,研究了基于神经网络的自适应非均匀性校正算法,探讨了最近4邻域像素平均、最近4邻域像素灰度加权和8邻域像素灰度加权等三种情况。实验结果表明,8邻域灰度加权算法校正效果较好
Firstly, the shortcoming of the traditional non-uniformity correction method is analyzed, and the necessity of adaptively correcting the non-uniformity of the infrared focal plane device is pointed out. According to the non-uniform noise characteristics of focal plane devices and the need of algorithm research, a method of generating non-uniform distorted images is introduced. Based on the above work, the adaptive nonuniformity correction algorithm based on neural network is studied, and three kinds of cases such as pixel average of the 4 neighborhoods, gray scale of the nearest 4 neighborhood pixels and gray scale of 8 neighborhood pixels are discussed . The experimental results show that the 8-neighborhood gray-scale weighting algorithm has a good correction effect