基于加权LoG算子的红外弱小目标检测方法研究

发布时间：2018-03-11 09:53

本文选题：加权LoG算子　切入点：形态学运算　出处：《应用光学》2017年01期 　论文类型：期刊论文

【摘要】：提出一种基于加权LoG算子及形态学运算的方法来提高远距离红外弱目标检测效率。通过不同尺度的加权LoG算子对图像进行运算,提取响应值最大的特征图。对图像进行形态学运算去噪,并进行Otsu二值分割聚类。输出目标点在图像上的位置坐标。实验结果表明:该方法与传统滤波方法相比,信噪比增益为36.9,杂波抑制因子为4.7,均比传统滤波方法要好。
[Abstract]:A method based on weighted LoG operator and morphological operation is proposed to improve the efficiency of remote infrared weak target detection. The weighted LoG operator with different scales is used to calculate the image. The feature map with the largest response value is extracted, the image is de-noised by morphological operation, and the Otsu binary segmentation clustering is carried out. The position coordinates of the target point on the image are outputted. The experimental results show that the proposed method is compared with the traditional filtering method. The SNR gain is 36.9 and the clutter suppression factor is 4.7, which is better than the traditional filtering method.
【作者单位】：河南工程学院计算机学院;上海交通大学电子信息与电气工程学院;
【基金】：国家自然科学基金(U1304616)
【分类号】：TN215

【相似文献】