人工智能技术的大视场光电测量系统畸变校正  

Distortion correction of wild field photoelectric measurement system based on artificial intelligence technology

作　　者：王月新[1] 田竹梅[1] 任国凤[1] 邵贵成[1] WANG Yuexin;TIAN Zhumei;REN Guofeng;SHAO Guicheng(Xinzhou Normal University,Xinzhou Shanxi 034000,China)

机构地区：[1]忻州师范学院,山西忻州034000

出　　处：《激光杂志》2020年第8期59-62,共4页Laser Journal

基　　金：忻州师范学院科研项目(No.2018KY15);山西省高等学校科技创新项目(No.STIP2019L0830);山西省教学改革创新项目(No.J2018162)。

摘　　要：针对传统大视场光电测量系统畸变校正存在的错误大、精度低等不足,提出了基于人工智能技术的大视场光电测量系统。首先获取理想和畸变节点坐标间的关系,建立畸变模型、对畸变数据筛选;然后根据理想半径与畸变半径之间的关系,确定图像畸变范围,对畸变数据的二次鉴别,并利用二次曲面拟合法和图像边缘部位投影校正畸变点的对应位置,最后实验测试结果:与传统方法相比,所提方法的筛选、判别能力更强,降低了畸变校正误差。For the shortcomings of traditional wild field photoelectric measurement system such as large error and low accuracy,a wild field photoelectric measurement system based on artificial intelligence technology is proposed.Firstly,the relationship between the ideal and the distortion node coordinates is obtained,the distortion model is established,and the distortion data is screened.Then,according to the relationship between the ideal radius and the distortion radius,the image distortion range is determined.After the distortion data is identified twice,the corresponding positions of the distortion points are corrected by the quadric surface fitting method and the projection of image edge.Finally,the experimental test results are as follows:compared with the traditional method,the screening and discrimination abilities of this method proposed are stronger,and the error of distortion correction is reduced.

关 键 词：人工智能技术 大视场光电测量系统 畸变 校正方法 

分 类 号：TN29[电子电信—物理电子学]