论文部分内容阅读
高空间分辨率遥感影像中地物目标内部光谱信息复杂性的增强,使得传统基于光谱特征值的数据处理方法效果不再显著,影像分割为解决这一问题提供了一种思路,成为当前高空间分辨率遥感影像处理的研究焦点。时刻独立脉冲耦合神经网络具有状态相近、空间相邻神经元相互耦合同步脉冲激发和区域之间神经元脉冲激发时刻独立两大特点,已被应用于非遥感影像分割中,并取得较好效果。本文结合高空间分辨率遥感影像特点,通过对网络参数进行实验和分析,提出一个基于时刻独立脉冲耦合神经网络的高空间分辨率遥感影像分割方法,并利用空间分辨率0.3m的航空影像进行了数据试验,将分割结果进行讨论并与现有时刻独立脉冲耦合神经网络方法和ISODATA方法分割结果进行对比分析。结果表明:时刻独立脉冲耦合神经网络在高空间分辨率遥感影像分割处理中具有很好的应用前景。
The complexity of the spectral information inside the object in high spatial resolution remote sensing image is enhanced, which makes the effect of traditional data processing method based on spectral eigenvalue no longer significant. Image segmentation provides an idea to solve this problem and becomes the current high space Research focus of resolution remote sensing image processing. Time-independent impulsive coupled neural networks have two characteristics, that is, the state is similar, the spatially adjacent neurons are synchronized with each other and the neuron pulse excitation time is independent, which has been applied to non-remote sensing image segmentation and achieved good results. In this paper, based on the characteristics of high spatial resolution remote sensing images, this paper presents a method of high spatial resolution remote sensing image segmentation based on time independent pulse coupled neural network by experiment and analysis of network parameters. Using aerial images with spatial resolution of 0.3m, Data experiments were carried out to discuss the segmentation results and to compare with the existing independent pulse coupled neural network method and the ISODATA method. The results show that time independent pulse coupled neural network has a good application prospect in high spatial resolution remote sensing image segmentation.