摘要
提出的针对外场红外测量图像的目标提取方法很好的解决了外场红外测量图像包含很多"奇异点"的干扰问题;该方法对强红外辐射的目标和弱的目标同样具有很好的提取效果,具有较强的实用性;同时利用外场红外测量图像的连续性,较好地解决了目标在图像中的定位问题,大大减少了目标提取的计算速度;算法利用了用户输入初始目标矩形范围和初始阈值,根据输入的矩形范围,确认背景区域,并根据背景区域计算背景灰度均值,利用该均匀与输入的阈值,初步确认目标像素点,再对初步确认的目标点进行连通性检验,取连通点数最多的区域为目标区域。
The proposed method of target extraction for external infrared image measurement is a good solution to the problem that the infrared measurement image contains many ＂singular points＂. This method has a good extraction effect on the target of strong infrared radiation,and the target with weak background control. At the same time,the continuity of the image is measured by the infrared field,and the positioning problem of the target in the image is solved. And the calculation speed of the target extraction can be greatly reduced. The algorithm uses the user to input the initial target rectangle range and the initial threshold. The background area is confirmed according to the input rectangular range,and the background gray value is calculated according to the background area. The target pixel is confirmed by the uniform and input threshold. The initial confirmation of the target point for connectivity testing was processed,and the area with the largest number of connected points is the target area.
出处
《兵器装备工程学报》
CAS
2017年第7期160-162,共3页
Journal of Ordnance Equipment Engineering
关键词
红外图像
目标提取
连通性检验
红外强弱目标
灰度值
infrared image
target extraction
connectivity test
infrared strong target
gray value g
