CAJ | 학술논문

地面景物边缘轮廓是其红外图像和可见光图像的共有特征,边缘信息量和边缘连续性是评价图像边缘特性的重要指标,也是预测图像匹配性能的依据.提出了一种评价红外场景边缘连续性的改进计算方法,利用该方法和POE指标,对机载试验采集的不同类型场景、不同时段、不同气象条件下红外图像与可见光图像边缘特性及其变化规律进行了统计研究.试验表明:地面景物红外图像和可见光图像的边缘信息量和边缘连续性,与图像匹配概率和定位精度存在正比关系.提出的景物边缘信息量和边缘连续性指标的评价方法有效,能够作为预测场景可匹配性、场景成像稳定性的根据.
지면경물변연륜곽시기홍외도상화가견광도상적공유특정,변연신식량화변연련속성시평개도상변연특성적중요지표,야시예측도상필배성능적의거.제출료일충평개홍외장경변연련속성적개진계산방법,이용해방법화POE지표,대궤재시험채집적불동류형장경、불동시단、불동기상조건하홍외도상여가견광도상변연특성급기변화규률진행료통계연구.시험표명:지면경물홍외도상화가견광도상적변연신식량화변연련속성,여도상필배개솔화정위정도존재정비관계.제출적경물변연신식량화변연련속성지표적평개방법유효,능구작위예측장경가필배성、장경성상은정성적근거.
As common characterizations, the edge information content of infrared image and edge continuity of ground scene are important features to analyze the object' edge image, and to estimate the matching performance of MWIR and visible images. In this paper, a novel method is presented, combined with POE metrics, which is designed to quantitatively analyze the edge continuity of the global scene. In order to investigate the relationship between edge characterization and matching performance, a great deal of MWIR and visible aerial image sequences are selected from experiments, which shot rural and factory district scenes at different time of a day under different meteorologic conditions. As shown in experimental results, the edge information content and edge continuity is proportional to the correlation peak value, and inversely proportional to the registration error for the same scene. It's feasible that the presented metrics of edge information content and its continuity are applied to predict the matching performance and imaging stability.