基于局部特征统计的小目标检测算法

Dim Targets Detection Based on Local Character Information Measurement

本文提出了一种新的基于局部特征统计的红外图像小目标检测方法。首先对图像的局部区域灰度概率进行统计,通过阈值分割,去除图像中缓慢变化的背景和弱的边缘,得到包含强的边缘点、噪声点和目标点的残留图像,然后利用残留图像内各点的局部方向信息测度的差异,进一步剔除强的边缘点。最后通过多帧累加判决的方式将真实目标从噪声点中检测出来。实验表明该方法能够极大地减少候选目标点数,准确有效地检测复杂自然背景中的红外运动弱小目标,适合于实时和多目标的检测。

A novel method based on local grey-level probability analysis and orientation information measurement is presented to detect small targets in IR images. Firstly, local grey-level probability is computed and used to remove the slowly-changing background and weak edges on the bases of threshold segmentation, as a result the remaining image composed of strong edge pointsl noise points and targets is obtained, Secondly we utilize the difference of the orientation information at the points of the remaining images to remove the strong edge points, Finally true targets are separated from noise points by multi-frames cumulating, Simulation result indicates that this approach can reduce the number of possible targets and detect small moving IR targets in complex natural background with high efficiency and veracity, Furthermore, it can be used in real-time and multi-targets detection,.