论文部分内容阅读
为提高复杂环境下红外小目标的检测效率,将图像分为平坦区域、边缘区域和小目标区域三种区域,并针对三种成分的特点,提出基于拉普拉斯金字塔的非线性局部滤波检测方法.首先将图像进行高斯金字塔分解,将高斯低通金字塔与原图像尺寸匹配后,相减并进行阈值操作,抑制平坦区域;其次将标记像素灰度值与其周围环域均值的最小差作为指标,滤除边界区域;最后将非线性局部滤波结果生成相应的拉普拉斯金字塔各层系数,重构得到高对比度的检测图像,利用邻域特点剔除孤立噪声点并通过简单阈值标记红外小目标.实验结果表明:与现有其他算法相比,该检测算法能够对复杂背景有效抑制,检测速度快.
In order to improve the detection efficiency of small infrared targets in complex environment, the image is divided into three regions: flat area, edge area and small target area. According to the characteristics of the three components, a non-linear local filter based on Laplacian pyramid Firstly, the image was decomposed by Gaussian pyramid, then the Gaussian low-pass pyramid was matched with the original image size, then subtracted and thresholded to suppress the flat region. Secondly, the minimum difference between the gray value of the marked pixel and the mean value of its surrounding region was taken as the index , Filter out the boundary area; Finally, the non-linear local filtering results generated corresponding Laplacian pyramid coefficient of each layer, reconstructed to obtain a high contrast detection image, the use of the characteristics of the neighborhood to remove isolated noise points and marked by a simple threshold infrared small target The experimental results show that compared with other algorithms, this algorithm can effectively suppress complex background and detect fast.