基于分形理论和神经网络的红外图像分割算法

针对实现红外图像的分割,快速准确地检测出红外序列中的人造目标的需要,结合分形特征和Ko-honen神经网络的特点,提出了一种基于分形技术的图像分割算法。该算法利用自然背景和人造目标的不同分形特征检测目标,提出了包括分形维数在内的7个红外图像特征,结合神经网络的自组织学习能力来进行图像分割。给出了算法实现的具体步骤。仿真试验结果表明,该算法能有效地实现红外图像的分割。

IR image segmentation method based on fractal theory and neural network

A segmentation algorithm based on the characters of fractal and neural network is proposed for detecting IR target in natural background.In this algorithm the intrinsic difference in fractal characters between man-made objects and natural background is used and seven characters of IR image are proposed for identifying man-made object,and neural network is used to segment the image.The process of the algorithm is given and the results show that this method can implement IR image segmentation effectively.