论文部分内容阅读
介绍了一种新的基于Ｇａｂｏｒ特征向量相似函数的复杂背景下红外（ＩＲ）图像目标自动检测方法。它利用２－ＤＧａｂｏｒ函数所提供的最佳联合空间－空间频率分辨率，及其形态符合人类视觉神经元感受野的特性，通过求解简单目标模板与原ＩＲ图像之间的相似度函数，自动检测出复杂背景下红外图像中的人造目标。大量的实际检测结果表明，本文介绍的自动目标检测方法不但对背景噪声和交错干扰具有良好的抑制能力，而且对目标的对比度、大小、形状及姿态具有一定程度的不变性。
A new automatic target detection method for infrared (IR) images with complex background based on Gabor eigenvector similarity function is introduced. It uses the best joint spatial-spatial frequency resolution provided by the 2-DGabor function and its morphology conforms to the characteristics of the human visual neuron receptive field. It automatically detects by solving the similarity function between the simple target template and the original IR image Artificial objects in infrared images under complex backgrounds. A large number of actual test results show that the automatic target detection method presented in this paper not only has a good ability to suppress background noise and staggered interference, but also has some degree of invariance to the target’s contrast ratio, size, shape and attitude.