利用小波变换与Gabor滤波检测红外小目标

将小波变换应用于红外图像,提出了一种新的弱小目标检测方法。该算法首先对图像进行小波变换。为了获取图像的多方向性分解,使用Gabor滤波器与高频信号做卷积运算得到24个方向的高频信息。计算各点的局部能量以及方向离散值,将以上特征融合,得到图像的多特征统计值。其次采用Renyi信息熵分割完成目标检测。最后,由于卷积运算,将会造成弱小目标点的扩大。因此,利用检测结果中小目标具有较大面积的特点,有效地排除了虚警。实验结果表明,该算法参数较少,能较精确地检测出红外弱小目标。

Detection algorithm for infrared small and weak targets based on wavelet transform and Gabor filter

The problem of detecting small and weak targets in infrared image sequences was addressed based on wavelet transform in this paper.Firstly,stationary wavelet multi-scale analysis was used for infrared image.To obtain the multidirectional decomposition,convolution operation was performed between Gabor filter and high frequency band of the signal.After running Gabor filter,24 directional high frequency signals can be got.The features were extracted and the multi-feature statistics of infrared image were obtai...