一种遥感卫星影像色差仿真模型构建和色差校正方法
By constructing a color difference simulation model and color difference correction method for remote sensing satellite images, and utilizing the UNet architecture's convolutional neural network and post-processing algorithms, the problem of color difference in optical remote sensing satellite images was solved, achieving efficient image correction and quality improvement.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- CHANGGUANG SATELLITE TECH CO LTD
- Filing Date
- 2025-08-28
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies cannot completely remove chromatic aberration in optical remote sensing satellite images in practical applications, affecting image quality and radiometric consistency.
A color difference simulation model for remote sensing satellite imagery is constructed. A convolutional neural network based on the UNet architecture is used for color difference correction. Two color difference simulation functions are designed by combining laboratory calibration data and on-orbit imaging characteristics. Color difference patterns are generated, and pre-training and fine-tuning datasets are constructed through block operations. Post-processing algorithms are used to correct boundary effects and achieve color difference correction.
It effectively removes chromatic aberration in remote sensing images, improves the radiometric quality and visual effects of the images, and enhances the application value of optical remote sensing satellite images.
Smart Images
Figure CN121095124B_ABST