采用形态神经网络背景自适应预测的图像弱小目标检测*

提出一种有效的背景杂波预测形态神经网络模型,用于检测图像数据中的弱小目标.目标被假设为只有很小的空域扩展度,而且淹没于强背景杂波干扰中.通过形态神经网络,杂波背景被准确地估计并从输入数据中去除,只剩下残留噪声和目标信号.采用扩展输入层数据的办法修正了传统的形态开、闭运算三层前馈BP网络模型.为了跟踪包含不同子结构的复杂背景,原始图像被划分为多个子块,并在相应的子块中选择训练样本对结构元进行优化.对真实图像数据的计算机仿真表明该算法在性能上优于其他传统算法.

Detecting Dim Small Targets in Image Data Using Morphological Neural Networks of Background Adaptive Prediction

An effective morphological neural network of background clutter prediction for detecting dim small targets in image data was proposed.The target of interest was assumed to have a very small spatial spread,and was obscured by heavy background clutter.The clutter was predicted exactly by morphological neural networks and subtracted from the input signal,leaving components of the target signal in the residual noise.The traditional 3-layer feed forward BP network modal of morphological opening and closing operation was modified by extending the input layer data.For tracking complex background including different sub-structures,the raw image was partitioned to some sub-blocks,in which the training samples were chosen for optimizing the weights of structuring element in the corresponding block.Computer simulations of real image data show better performance compared with other traditional methods.