土地环境高分遥感图像智能识别方法研究  

Research on Intelligent Recognition Technology of Land Environment High Resolution Remote Sensing Image

作　　者：王伟[1] WANG Wei(Department of Information Engineering,Henan Industry and Trade Vocational College,Zhengzhou Henan 451191,China)

机构地区：[1]河南工业贸易职业学院信息工程学院,河南郑州451191

出　　处：《信息与电脑》2021年第10期166-168,共3页Information & Computer

基　　金：河南省科技攻关项目“面向黄河流域生态保护的遥感图像智能处理系统”(项目编号:212102310085);河南省高等学校重点科研项目“土地环境高分遥感图像智能识别技术研究”(项目编号:21B520004)。

摘　　要：为提高土地环境的识别准确性,提出并设计了一种土地环境高分遥感图像智能识别方法.利用传感器输出图像的灰度值,获得地面目标大气真实反射率.根据遥感图像提供的校正参数计算出大气的振幅和亮度,通过寻找最优的分割尺度,完成对目标对象的多尺度分割.根据对多尺度分割影像分类并融合处理,实现土地环境高分遥感图像智能识别。实验结果表明,所研究的识别技术较传统技术识别准确性高,并且花费的时间也较少.In order to improve the accuracy of land environment recognition, an intelligent recognition method of high resolution remote sensing image of land environment is proposed and designed. Using the gray value of the sensor output image, the real atmospheric reflectance of the ground target is obtained. According to the correction parameters provided by the remote sensing image, the amplitude and brightness of the atmosphere are calculated, and the multi-scale segmentation of the target object is completed by finding the optimal segmentation scale. According to the classification and fusion of multi-scale segmentation image, the intelligent recognition of land environment high resolution remote sensing image is realized. The experimental results show that the recognition technology is more accurate than the traditional technology, and it takes less time.

关 键 词：土地环境 高分遥感图像 智能识别 

分 类 号：P407[天文地球—大气科学及气象学]