Quick splicing method for remote sensing image of unmanned aerial vehicle

Abstract

A quick splicing method for a remote sensing image of an unmanned aerial vehicle comprises the steps of (1), performing coarse extraction of characteristic points based on a SURF algorithm; (2), eliminating mis-extracted characteristic points by means of an RANSAC algorithm; (3), performing accurate matching of the characteristic points of an image by means of a least square matching algorithm; (4), performing an image fusion process; and (5), splicing the image, thereby obtaining the remote sensing image of the unmanned aerial vehicle. According to the quick splicing method, the characteristic points of the remote sensing image of the unmanned aerial vehicle is coarsely extracted by means of the SURF algorithm firstly, and then the mis-extracted characteristic points are eliminated by means of a RANSAC algorithm. Finally accurate matching of the characteristic points and real-time splicing of the image are realized by means of the least square image matching algorithm. The least square image matching algorithm can realize a good splicing effect and high splicing efficiency, and furthermore can satisfy a requirement for real-time splicing. The longitudinal overlap degree of the to-be-spliced unmanned aerial vehicle remote sensing image is larger than 60%, and the side overlap degree is larger than 30%. The quick splicing method has advantages of low calculation amount, high splicing effect, high splicing efficiency, high matching degree and effective realization of real-time image splicing.

Description

technical field [0001] The invention belongs to the technical field of remote sensing of unmanned aerial vehicles, and in particular relates to a method for quickly splicing remote sensing images of unmanned aerial vehicles. Background technique [0002] Real-time splicing of UAV remote sensing images can reflect the status of the survey area in real time and lay the foundation for subsequent image processing. In the stitching method of UAV remote sensing images based on the SURF algorithm, a large number of incorrect feature points are extracted during the stitching process, resulting in a large amount of calculation, low matching efficiency, and cumbersome procedures, which cannot meet the requirements of real-time stitching, and the distance threshold matching method is adopted. The matching accuracy is low and the splicing effect is poor. Contents of the invention [0003] In order to solve the deficiencies in the prior art, the present invention provides a method for...

AI Technical Summary

Problems solved by technology

In the stitching method of UAV remote sensing images based on the SURF algorithm, a large number of incorrect feature points are extracted during the stitching process, resulting in a large amount of calculation, low matching efficiency, and cumbersome procedures, which cannot meet the requirements of real-time stitching, and the distance threshold matching method is used. Low matching accuracy and poor splicing effect

Images