BP神经网络在图像信息提取中的应用研究

The Applied Research About BP Neural Network of Image Information Extraction

本文阐述BP神经网络在图像信息提取中的运用,包括:遥感技术、图像识别原理和模式识别常用的神经网络模型.针对目前在目标识别中应用最多的前馈神经网络模型研究其采用的BP算法,讨论了由于标准的BP算法存在训练时间长、收敛速度慢和易陷入局部极小值等问题,提出了BP神经网络分类器的设计方法、样本的选择及组织的改进方法.

This paper first introduces BP neural network image information extraction include： remote sensing tech- nology, image recognition principle and several pattern recognition used neural network model. Then BP algorithm for target recognition application feedforward neural network model which uses standard BP algorithm training time long, slow convergence and easy to fall into local minimum value. Several BP algorithm in BP neural network classi- fier design method, the improved method of sample selection and organization were put forward.