探究空间探测目标影像清晰度自动评价方法  

Exploring the Automatic Evaluation Method of Image Clarity of Space Exploration Target

作　　者：王德营 梁志锋 王渊 郭伟 Wang Deying;Lliang Zhifeng;Wang Yuan;Guo Wei(Geovis Technology Co.,Ltd.,Hefei Anhui 230011)

机构地区：[1]中科星图测控技术(合肥)有限公司,安徽合肥230011

出　　处：《现代工业经济和信息化》2023年第5期210-212,共3页Modern Industrial Economy and Informationization

摘　　要：空间探测目标影像目前仍多采用人工目视的主观清晰度评定方法进行数据筛选分类,但随着空间探测遥感影像数据规模的不断增长,亟需寻找一种可替代人工目视的清晰度自动评价方法。以典型的不同清晰度的空目影像来对8种通用影像清晰度客观评价方法进行筛选。结果表明,针对空目影像,灰度方差及灰度方差乘积两种评价方法的评价结果符合性较好,适用于该应用场景下的清晰度自动评价。At present,the subjective sharpness evaluation method of artificial vision is still used for data screening and classification of space exploration target images.However,with the continuous increase of the scale of space exploration remote sensing image data,it is urgent to find an alternative automatic sharpness evaluation method.Eight general objective evaluation methods for image clarity were screened by typical aerial images of different resolutions.The results show that for the aerial image,the evaluation results of the two evaluation methods of gray variance and gray variance product are in good agreement,and they are suitable for the automatic evaluation of sharpness in this application scenario.

关 键 词：空目影像 清晰度 梯度函数 灰度方差 熵函数 

分 类 号：V443.5[航空宇航科学与技术—飞行器设计]