基于帧差法和自适应阈值区域生长的红外运动目标检测

Infrared Moving Target Detection Based on Frame Difference Method and Adaptive Threshold Region Growing

摘要:

针对具有背景干扰、信噪比低的红外图像,提出了一种基于帧差法和自适应区域生长的红外运动目标检测方法。首先对红外图像进行了高帽变换,以抑制大面积背景的干扰,相邻帧图像间做帧差,初步提取目标区域;其次分析了红外目标的特性,针对其特性提出了一种基于灰度等级的自适应阈值分割方法;最后以帧差法检测的目标质心为种子点,以自适应阈值为分割准则,在预处理后的图像中进行区域生长,最终实现了红外运动目标的检测。结果表明,所提算法可抑制大面积背景的干扰,实现单个和多个红外运动目标的完整提取和检测。?更多还原

Abstract:

Aiming at infrared images with background interference and low SNR, an infrared moving target detection method based on frame difference and adaptive threshold region growing was proposed. Firstly, the top-hat transform was adapted on the input infrared images in order to inhibit the large area of background interference. The algorithm used the method of frame difference to extract target regions initially between the adjacency images. Secondly, it analyzed the characteristics of infrared targets, and proposed an adaptive threshold segmentation algorithm based on image gradations. Finally, using the target centroid as the seed points and the adaptive threshold as the segmentation criteria, the region growing was applied on the pretreated images, and the infrared moving target detection was achieved. The results show that the algorithm in this paper can inhibit the large area of background interference and achieve the complete extraction and detection for one or more infrared moving targets.