High-resolution aerial image partition method based on LEGION and under assisting of LiDAR

Abstract

The invention relates to a high-resolution aerial image partition method based on LEGION and under assisting of a LiDAR. The high-resolution aerial image partition method comprises the following steps of utilizing LiDAR laser point cloud data to generate a laser echo intensity image, conducting registration of the laser echo intensity image and a remote sensing image panchromatic wave band, using a principal component analysis (PCA) to conduct fusing of the laser echo intensity image and the remote sensing image panchromatic wave band after registration of the laser echo intensity image and the remote sensing image panchromatic wave band, obtaining a fusing image, and conducting LEGION image partition for the fusing image. Compared with the prior art, the PCA is utilized to fuse the laser echo intensity image and the high-resolution remote sensing image panchromatic wave band, two different data sources are comprehensively utilized to be used as LEGION partition method input data, LiDAR data and characteristics of a panchromatic wave band image are well reserved, and partition precision of a high-resolution remote sensing image is effectively improved. The LiDAR data and information of the remote sensing image are comprehensively utilized to conduct LEGION partition, and practice shows that the technology can effectively conduct remote sensing image partition and obtain satisfying partition effect.

Description

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine

Login to View More

AI Technical Summary

Benefits of technology

The present invention combines laser echo intensity images and high-resolution remote sensing images using principal component analysis (PCA) to improve the accuracy of image segmentation. By fusing these two data sources, the invention retains the spectral characteristics of LiDAR data and the structural features of high-resolution remote sensing images. This fusion provides a better input for the segmentation method used in the invention. The segmentation results are similar to how human eyes recognize ground objects, especially those with consistent texture characteristics. The method is also more resistant to noise and provides a stronger and more consistent segmentation of ground objects compared to other methods. Overall, the invention effectively performs remote sensing image segmentation and achieves satisfactory results.

Problems solved by technology

Technology Problem addressed in this patents relates to enabling efficient and effective analysis of remotely sensing videos captured at varying levels of detail due to factors like environmental conditions and lightning strikes. Current techniques involve extracting representative regions containing all possible background scenes without considering these issues, leading to incomplete understanding of the video content.

Images