Adaptive correction method for intersection error of surveying and mapping satellite

Abstract

The invention relates to an adaptive correction method for an intersection error of a surveying and mapping satellite. The method comprises: with random left and right visual stereo images of a same region, calculation of an affine transformation basis matrix F is put forward by using RPC parameters and SRTM data based on same-name point matching; on the basis of an SIFT algorithm, key point extraction, matching and optimization are carried out on the left and right visual stereo images, and calculation is performed to obtain a translation correction matrix T of an intersection error; after translation correction, a residual error reaches a sub-pixel level, so that the elevation precision is guaranteed effectively. According to the invention, intersection error correction is performed adaptively without a ground control point under the condition of a plurality of stereo images with different base-height ratios, so that the elevation accuracy of satellite surveying and mapping is improved substantially.

Description

technical field [0001] The invention relates to an adaptive correction method for the intersection error of surveying and mapping satellites, belongs to the technical field of optical remote sensing satellites, and is particularly suitable for improving and guaranteeing the elevation accuracy of surveying and mapping satellites. Background technique [0002] The basic principles of satellite stereoscopic mapping are as follows: figure 1 As shown, the two cameras are at the satellite height H, and the baseline length is B (the center C of the two images 1 and C 2 The distance between points), B / H is called the base height ratio. The same target with a height of h is imaged multiple times, and the height information h' is reconstructed by correlation processing using the parallax between the left and right stereo images. The height accuracy Δh mainly depends on the intersection error dx divided by the base height ratio B / H. Especially when the base height is relatively sma...

AI Technical Summary

Problems solved by technology

Even for the high-resolution WorldView-3 satellite currently in orbit, the plane positioning accuracy without control points is about 3.5m, and the intersection error of the two stereo imaging will be close to or even exceed 3.5m. If the base height ratio is 0.6, the elevation The accuracy will reach about 6m, which obviously cannot meet the needs of high-precision surveying and mapping without control points

Images