An Object-Oriented Orthophoto Quality Inspection Method Based on Gray Correlation

Abstract

The invention provides an object-oriented orthophoto quality inspection method based on gray-scale correlation, including technical links such as multi-spectral conversion to panchromatic, 16-bit to 8-bit conversion, superpixel segmentation, and gray-scale correlation calculation. In the method, the effective area of ​​the image to be processed is firstly calculated, and the processing tasks (ie, edge joining and positioning tasks) are determined according to the overlap of the effective area. Block. Then, 16-bit to 8-bit processing is performed on the block data, and superpixel segmentation is performed. Finally, bordering and positioning offset calculations are performed on the superpixel image blocks. The technical scheme of the invention truly realizes the automatic, rapid and accurate quality inspection of the orthophoto image of the optical remote sensing satellite.

Description

technical field [0001] The invention belongs to the field of surveying and mapping science and technology, and relates to an object-oriented orthophoto quality inspection technology based on gray scale correlation. Background technique [0002] Represented by Ziyuan No. 3, Gaofen No. 1 and Gaofen No. 2, the design indicators of optical remote sensing satellites have gradually reached the international advanced level, the earth observation system is becoming more and more perfect, the data volume of satellite images is increasing rapidly, and the degree of marketization is increasing year by year improve. However, not all remote sensing image products can meet the requirements of image positioning processing. How to quickly, accurately and automatically evaluate the quality of each image product is particularly important. [0003] After the optical remote sensing satellite image is geometrically corrected, most of the areas have better border and positioning accuracy, but it...

AI Technical Summary

Problems solved by technology

[0003] After the optical remote sensing satellite image is geometrically corrected, most of the areas have better border and positioning accuracy, but it cannot be guaranteed that all images and the processing results of each area of ​​the image can meet the production requirements. For example, for the 2-meter data of Gaofen-2, the general production The edge joint and positioning accuracy are required to be within two pixels. Due to the limitation of geometric processing technology and image quality, it is often impossible to meet the production requirements in some areas. Therefore, it is necessary to perform quality inspection on each orthophoto and each area on the image. , in order to extract the borders on the image or the areas with large positioning errors

Traditional methods often use manual methods to compare orthophotos with high-precision base maps and inspect them area by area, which requires a lot of manpower and material resources, and is inefficient

Images