Large-range multi-view inclined image connection point quickand accurate matching method

Abstract

The invention discloses a large-range multi-view inclined image connection point quick and accurate matching method, which comprises the following steps of calculating a downward-view camera placementangle error by utilizing a downward-viewimage in a small-range local area, and correcting an external azimuth angle element of thedownward-view image, performingdownward-view image connection point matching by fusing and utilizing strategies such as pyramid image matching, block feature point extraction and relaxation iteration method matching, and obtaining accuratedownward-view image exterior orientation elements and connection point object space coordinates through block adjustment, on the basis, fully using the spatial geometric constraint relation between the downward-view image and theoblique-view image, and achieving turning point matching from a downward-view image connection point to the oblique-view image, and refining the coordinates of the image points of the connection points by using multi-view object space matching to realize rapid and accurate matching of the connection points of the inclined images. The geometric constraint between the downward-view image and the oblique-view image is fully utilized, only the feature points are extracted from the downward-view image for matching, and the matching speed and quality of the connection points of the large-range oblique image are effectively guaranteed.

Description

technical field [0001] The invention belongs to the field of surveying and mapping science and technology, and relates to a fast and accurate matching method for connection points of large-scale multi-view oblique images. This method is suitable for fast automatic aerial triangulation of large-scale multi-view oblique images, which is conducive to improving the matching speed, accuracy and overlap of connection points in automatic aerial triangulation of multi-view oblique images, and provides more information for the calculation of block adjustment of oblique images. Excellent image point observation data. Background technique [0002] Multi-view aerial photography has changed the operation mode of traditional aerial photography, which only shoots vertically. It can obtain multi-view images from downward and different oblique directions at the same time, and obtain richer multi-view image information. It has been used in large-scale mapping, cadastral surveying, Widely use...

AI Technical Summary

Problems solved by technology

[0010] The purpose of the present invention is to provide a method for fast and accurate matching of connection points of large-scale multi-view oblique images, which mainly solves the problem of low matching efficiency of connection points, low matching quality and utilization of multi-view image information in automatic aerial triangulation of multi-view oblique images. Insufficient, empty three layers and other problems

Images