论文部分内容阅读
将粗糙集理论和神经网络结合,提出了一种基于粗神经网络的信息融合算法,用于图像融合的研究,该方法不仅可以接受定量的输入,而且可以接受定性的输入,即输入是一个范围,或者在观测时间内输入是变化的。由于粗神经网络的误差传递函数不可微,不能用BP算法训练粗神经网络,所以采用遗传算法训练粗神经网络。仿真实验结果表明,基于粗神经网络的信息融合方法可以有效地融合多幅带有噪声的图像,并取得了良好的融合效果。
Combining rough set theory and neural network, an information fusion algorithm based on rough neural network is proposed for image fusion research. This method can not only accept quantitative input but also qualitative input, that is, the input is a range , Or the input changes during the observation time. Because of the error transfer function of rough neural network is not differentiable, can not use BP algorithm to train rough neural network, so using GA to train rough neural network. The simulation results show that the method based on rough neural network can fuse many images with noises effectively and achieve a good fusion effect.