应用OB-HMAD算法与光谱特性的遥感图像动态变化检测

刘琤

刘琤.应用OB-HMAD算法与光谱特性的遥感图像动态变化检测[J].贵州地质,2023,40(2):173-176, 172

应用OB-HMAD算法与光谱特性的遥感图像动态变化检测

Dynamic Change Detection of Remote Sensing Images by Using OB-HMAD Algorithm and Spectral Characteristics

投稿时间：2023-01-04

DOI：

中文关键词: OB-HMAD算法光谱特性自然资源遥感图像动态变化光谱异质性形状异质性

英文关键词:OB-HMAD algorithm Spectral characteristic Remote sensing images of natural resources Dynamic change Spectral heterogeneity Shape heterogeneity

基金项目:

作者	单位
刘琤	武夷山市自然资源局，福建南平354300

摘要点击次数: 205

全文下载次数: 374

中文摘要:

在对图像动态变化情况进行检测的过程中，对原始图像的分割精度较低，导致检测结果存在较大误差。为此，应用OB-HMAD算法与光谱特性，进行自然资源遥感图像动态变化检测。在计算了影像各波段的光谱异质性和形状异质性特征值参量后，结合各波段对应的权重系数，考虑自然资源遥感图像的光谱异质性和形状紧凑度、形状平滑度，利用OB-HMAD算法对自然资源遥感图像进行多尺度分割处理。在图像动态变化检测阶段，根据自然资源遥感图像变化与光谱特征参数之间的关系，计算了各分割图像的曲率参数，通过计算曲率参数之差完成对目标图像参数动态变化的计算。实验结果表明，该方法应用下，图像NDVI、RVI和SAVI变化量的检测结果与实际值的误差分别仅为0000 9，0005 7和0004 8。以上数据证明该方法完成了遥感图像动态变化检测，且检测效果较佳。

英文摘要:

In the process of detecting dynamic changes in images，the low segmentation accuracy of the original image，there are significant errors in the detection results. Therefore，the OB-HMAD algorithm and spectral characteristics are applied to detect dynamic changes in natural resource remote sensing images. After calculating the spectral heterogeneity and shape heterogeneity feature parameters of each band of the image，combined with the corresponding weight coefficients of each band，considering the spectral heterogeneity，shape compactness，and shape smoothness of natural resource remote sensing images，the OB-HMAD algorithm is used for multi-scale segmentation of natural resource remote sensing images. In the stage of image dynamic change detection，the curvature parameters of each segmented image are calculated based on the relationship between the changes in natural resource remote sensing images and spectral feature parameters. The calculation of the dynamic changes in the target image parameters is completed by calculating the difference between the curvature parameters. The experimental results show that under the application of this method，the errors between the detection results of NDVI，RVI，and SAVI changes in images and the actual values are only 0.0009，00057 and 00048 respectively. The above data proved that it completed the detection of dynamic changes in remote sensing images and achieved good detection results.

查看全文查看/发表评论下载PDF阅读器