论文部分内容阅读
在红外搜索与跟踪系统中,背景杂波抑制效果将直接影响到低信噪比条件下点状运动目标的检测及跟踪性能。利用RBF神经网络的非线性映射能力和遗传算法的全局搜索机制,本文研究了一种利用遗传算法(GA)优化RBF神经网络的背景杂波抑制技术。杂波抑制后,残留噪声的高斯性和独立性通过Kendall秩相关法和计算Friedman统计量的方法进行了验证,背景杂波抑制效果与BP神经网络和常用的Uniform加权函数进行了比较,结果表明本文研究方法可行有效。
In the infrared search and tracking system, the background clutter suppression effect will directly affect the detection and tracking performance of the point-like moving target under low signal-to-noise ratio conditions. Based on the non-linear mapping ability of RBF neural network and the global search mechanism of genetic algorithm, this paper studies a background clutter suppression technology based on genetic algorithm (GA) to optimize RBF neural network. After the clutter suppression, the Gaussianity and independence of the residual noise are verified by the method of Kendall rank correlation and Friedman statistics calculation. The background clutter suppression effect is compared with the BP neural network and the commonly used Uniform weighting function. The results show that This research method is feasible and effective.