遗传神经网络在红外目标识别中的应用

遗传算法和神经网络都是解决非线性不确定问题的有力工具,在图像处理方面各有优缺点,BP神经网络的局限是收敛速度慢且易限入局部最小而遗传算法的优势则是具有较强的全局搜索能力,却易陷入过早收敛.针对遗传算法和神经网络的优缺点,取长补短,介绍了利用遗传算法对BP神经网络进行权值优化的红外小目标图像处理识别技术.计算机仿真实验结果表明了该方法的收敛速度明显优越于其他传统方法,具有较强的实用性和鲁棒性.

Application of Genetic Neural Network in Infrared Target Discrimination

Genetic algorithm and neural network are all powerful tools in solving nonlinear and uncertain problems, which there are advantages and disadvantages in image processing individually. The limitation of BP neural network is the slow rapidity of convergence and it is trapped in local minimum easily. The advantage of Genetic Algorithm is the powerful ability in overall searching but it is converged in advance easily. It is put forward in this paper that infrared small target image processing and discriminating technology based on genetic algorithm does right-value optimization to BP neural network according to the advantages and disadvantages in genetic algorithm and neural network. Computer simulations are given to demonstrate that the convergence rate of the model is faster than other conventional methods and possess stronger practicability and robustness.