An intelligent color uniformization method based on online remote sensing image service

By using an intelligent color balancing method based on online remote sensing image services, the problems of difficulty in selecting reference images and low processing efficiency are solved, achieving efficient and reliable color balancing of large-scale remote sensing images, which is applicable to various remote sensing data processing scenarios.

CN121544472BActive Publication Date: 2026-06-19ZHONGKE XINGTU DIGITAL EARTH HEFEI CO LTD
View PDF 3 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
ZHONGKE XINGTU DIGITAL EARTH HEFEI CO LTD
Filing Date
2025-08-25
Publication Date
2026-06-19

AI Technical Summary

Technical Problem

Existing remote sensing image color balancing methods suffer from difficulties in selecting reference images, low processing efficiency, and a lack of solutions that can be directly applied to engineering practice, resulting in poor color balancing consistency and low processing efficiency in large-area remote sensing images.

Method used

Based on online remote sensing image services, the location and quantity of reference image tile data are calculated hierarchically to generate a cross-scale coefficient of variation distribution map. Multi-scale fusion is performed by combining mean, variance and coefficient of variation. Wallis filter is used for image color balancing and quality assessment is conducted.

Benefits of technology

It achieves automated and intelligent color balancing of large-scale remote sensing images, improving processing efficiency and result consistency. It is suitable for remote sensing data processing needs in different scenarios and meets the requirements of real-time processing and batch automation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121544472B_ABST
    Figure CN121544472B_ABST
Patent Text Reader

Abstract

This invention discloses an intelligent color balancing method based on online remote sensing image services. The method includes: determining the position and number of reference image tiles at different levels according to the resolution and latitude / longitude coordinates of the image to be sized; calculating the mean and variance of tiles at each level to generate a cross-scale coefficient of variation distribution map; fusing the reference images at multiple scales based on the above statistical characteristics; balancing the image to be sized using the fusion result as a benchmark; outputting the result and evaluating its quality. This invention utilizes image analysis and machine learning techniques to achieve automated multi-scale color balancing, quickly locating the color balancing benchmark, outputting a visual quality inspection report, supporting GPU acceleration and embedded deployment, and meeting the needs of scenarios such as real-time satellite processing and batch color balancing at ground stations, thereby improving the automation and reliability of remote sensing data color balancing.
Need to check novelty before this filing date? Find Prior Art