基于大数据分析的红外图像非均匀性自动化校对  

Automated Verification Method for Infrared Image Non-uniformity Based on Big Data Analysis

作　　者：陆玉[1] LU Yu(Fuyang Institute of Technology,Fuyang Anhui 236000,China)

机构地区：[1]阜阳职业技术学院,安徽阜阳236000

出　　处：《佳木斯大学学报(自然科学版)》2025年第1期39-42,共4页Journal of Jiamusi University:Natural Science Edition

基　　金：阜阳职业技术学院2022年校级科研项目服务地方专项项目(2022KYXM11)。

摘　　要：为提升红外图像质量,降低非均匀性噪声对于图像的干扰,提出基于大数据分析的红外图像非均匀性自动化校对方法。该方法采用K-means聚类算法对红外图像进行亮度均衡化处理后,再通过生成对抗网络模型进行图像的非均匀性自动化校对,增强红外图像质量。测试结果显示:该方法可有效完成红外图像亮度均衡化处理,处理后红外图像的色差结果均在0.11以下,校对后红外图像质量较好,粗糙度的结果均在-0.72以下,自然图像质量均在40以下。In order to improve the quality of infrared image and reduce the interference of non-uniformity noise,an automatic proofreading method for infrared image non-uniformity based on big data analysis is proposed.After using the K-means clustering algorithm to perform brightness equalization on infrared images,an adversarial network model is generated for automatic correction of image non-uniformity,enhancing the quality of infrared images.The test results show that this method can effectively complete the brightness equalization processing of infrared images.After processing,the color difference results of infrared images are below 0.11,the quality of infrared images is better after proofreading,the roughness results are below-0.72,and the quality of natural images is below 40.

关 键 词：大数据分析 红外图像 非均匀性 自动化校对 亮度均衡化 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]