An Automatic Orthorectification Method Based on Remote Sensing Image of Environment-1 Satellite

Abstract

An automatic orthorectification method based on the remote sensing image of the HJ-1 satellite, the steps are as follows: (1) Eliminate the projection deformation caused by the ground fluctuation in the reference image according to the principle of geometric projection transformation, and generate a simulated projection image; (2) Use Harris The corner extraction algorithm searches for possible feature points in the reference image and the original image; (3) uses the correlation coefficient method to find possible points with the same name on the basis of step (2); (4) uses a polynomial model based on least squares Solve the relationship coefficient between the reference image and the original image; (5) use the relationship coefficient between the reference image and the original image to correct the original image and eliminate the distortion caused by terrain fluctuations again according to the principle of geometric projection transformation, and finally complete the orthorectification. The method of the invention has the characteristics of high degree of automation, no need for manual selection of control points, high correction accuracy and high production efficiency.

Description

technical field [0001] The invention relates to an automatic orthorectification method based on the remote sensing image of the Environment-1 satellite, which belongs to the technical field of remote-sensing image processing, and is suitable for the precise positioning of the remote-sensing image of the Environment-1 satellite, the extraction of image information and the preprocessing work of the production of special products, etc. . Background technique [0002] During the imaging process of satellite imagery, it is affected by many factors such as projection, camera axis tilt, earth curvature, and terrain fluctuations, resulting in different degrees of geometric deformation and distortion of each image point in the image. The orthorectification image can eliminate the image distortion caused by various factors in the imaging process, and restore a remote sensing image with accurate geographical location. Therefore, orthorectification of remote sensing images has become a...

AI Technical Summary

Problems solved by technology

However, stereo images are not easy to obtain for remote sensing images, the cost is high, and a certain number of control points need to be manually selected.

For the few existing remote sensing image automatic correction methods that use image matching and general empirical models, they only use corner point matching and quadratic polynomial models to apply them simply, and a fixed threshold needs to be set in each step. flexible, poor applicability

[0004] The shortcomings of the above methods are mainly reflected in: (1) manual selection of control points is required, and the process of selecting control points is extremely cumbersome, which greatly reduces the efficiency of image orthorectification; (2) due to the high dependence on manual work, the image correction The accuracy is uncontrollable and greatly affected by manual work; (3) stereo images are difficult to obtain for remote sensing images, and the cost is high; (4) the universality of thresholds in existing automatic correction methods is not high, and the calculation efficiency and calculation The precision of the result cannot be improved

Images