基于梯度关系和局部连通性的红外小目标检测方法

Infrared Dim Target Detection Method Based on Gradient Relation and Local Connectivity

红外图像中的弱小目标检测技术在精确制导等领域有着广泛的应用,是提升现代军事实力的关键环节。传统的目标检测算法主要基于背景抑制和目标显著性两个方面,但往往不适用于色差较大、干扰较多的复杂环境。本文提出一种基于梯度检测和局部连通性判断的弱小目标检测算法,利用弱小目标在局部区域的梯度特性和连通特性识别可疑目标,并采用区域生长算法进一步剔除检测到的干扰点。同时采用自适应策略确定相关阈值,提升算法的灵活性与应用范围。本文算法在反映多种检测环境的图片和视频序列中进行检测,并与传统方法比较。检测结果表明:该方法准确率较高,可保证较低的虚警率,同时处理速度较快。

The dim target detection technology in infrared images has been widely used in fields such as precision guidance.It is a necessary part for modern military strength.Traditional target detection algorithms are mainly based on background suppression and target saliency.They are often not suitable for complex environments such as large color difference and much interference.In this paper,a dim target detection algorithm based on gradient relation and local connectivity is proposed.The gradient and connectivity characteristics of the dim targets in local areas are used to identify the suspicious targets,and the region growing algorithm is used to further eliminate the detected interference points.At the same time,an adaptive strategy is used to determine the relevant thresholds so as to improve the flexibility and application scope of the algorithm.The proposed algorithm is used to detect the images and video sequences reflecting multiple detection environments.The results show that,compared with the traditional ones,the proposed algorithm has higher accuracy,lower false alarm rate,and faster processing speed.