文献类型：期刊文献

中文题名：遥感影像融合下自然资源地类特征提取仿真

英文题名：Simulation of Natural Resource Land Feature Extraction Based on Remote Sensing Image Fusion

作者：蒙友波 廖艳梅 覃锋 王晓红

第一作者：蒙友波

机构：[1]贵州省自然资源勘测规划研究院,贵州贵阳550000;[2]贵州理工学院矿业工程学院,贵州贵阳550003;[3]贵州大学林学院,贵州贵阳550025

第一机构：贵州省自然资源勘测规划研究院,贵州贵阳550000

年份：2023

卷号：40

期号：9

起止页码：162-166

中文期刊名：计算机仿真

外文期刊名：Computer Simulation

收录：CSTPCD;;北大核心:【北大核心2020】；

语种：中文

中文关键词：遥感影像;影像融合;自然资源;地类特征;聚类算法

外文关键词：Remote sensing image;Image fusion;Natural resource;Land type characteristic;Clustering algorithm

摘要：目前方法对自然资源地类图像进行特征提取时,由于未融合处理遥感图像,导致存在平均正确识别率低、平均查全率低以及平均误报率高的问题。提出遥感影像融合下自然资源地类特征提取方法。方法首先对遥感图像进行分解处理,使用拉普拉斯法对分解图像进行融合;再通过小波变换法恢复图像像素分量,建立遥感图像的目标函数并确定正确的滤波器参数;最后通过聚类算法对图像进行计算,从而实现自然资源地类图像的特征提取。实验结果表明,运用上述方法对自然资源地类图像进行特征提取时,平均正确识别率高、平均查全率高以及平均误报率低。
At present,when extracting features from natural resource land class images,there are problems such as low average correct recognition rate,low average recall rate,and high average false alarm rate due to the lack of fusion processing of remote sensing images.Based on remote sensing image fusion,this paper puts forward a method to extract natural resource land type features.Firstly,the remote sensing image was decomposed,and then fused by the Laplace method.Secondly,image pixel components were restored by wavelet transform.Thirdly,objective functions of remote sensing images were constructed.And correct filter parameters were determined.Finally,the image was computed by clustering algorithm.Thus the feature extraction for natural resource land classification images was achieved.Experimental results show that the proposed method has a higher average correct recognition rate and average recall rate as well as a lower average false alarm rate.