基于数据挖掘的激光雷达图像识别技术  

Laser Radar Image Recognition Technology Based on Data Mining

作　　者：杨帆[1] 周华[1] 王嫣[1] 

机构地区：[1]中州大学,郑州450044

出　　处：《激光杂志》2016年第7期39-42,共4页Laser Journal

基　　金：河南省科技发展计划项目(152400410314)

摘　　要：图像识别是激光雷达技术应用的基础,为了提高激光雷达图像的识别准确性,提出一种数据挖掘的激光雷达图像识别方法。首先采用小波变换对激光雷达图像进行去噪处理,并提取特征向量,然后采用数据挖掘技术-深度层次网络根据特征建立激光雷达图像识别模型,最后进行激光雷达图像识别仿真实验。结果表明,本文方法可以提取激光雷达图像有效特征,可以得了较好的激光雷达图像识别结果。Image recognition is the basis of the application for laser radar technology, in order to improve the recognition accuracy of the laser radar image, a new method of image recognition based on data mining is proposed in this paper. Firstly, wavelet transform is used to deal with the laser radar image denoising, and feature vectors are extracted, secondly, data mining technology--depth layer network is used to establish the model of the laser radar image recognition according to features. Lastly, the simulation experiment of laser radar image recognition is carried out. The results show that the proposed method can extract the effective features of the laser radar image, which can get good results of the laser radar image recognition.

关 键 词：激光技术 雷达图像 识别模型 数据挖掘 深度层次网络 

分 类 号：TN91[电子电信—通信与信息系统]