基于最优可免域神经免疫网络的深度模糊红外目标提取算法

深度模糊是模糊红外图像的一类表现特征，准确提取红外图像的深度模糊区域是提取模糊红外目标的基础。基于生物免疫系统在抗原检测、提取和消除上表现出识别、学习、记忆、耐受和协调配合等优异特性，结合生物免疫中神经系统与免疫系统相互作用的关系，提出了一种基于最优可免域神经免疫网络的深度模糊红外目标提取算法。该算法通过设计神经网络能给进行模糊红外图像目标与背景分类的免疫网络以指导作用。依靠独立于免疫系统神经网络先验知识的作用，设计了最优可免域神经免疫网络，实现了针对深度模糊红外目标的准确提取。实验结果证明，相对于其他传统目标提取算法，该算法能更有效和更准确地提取模糊红外目标图像中的目标。

Deeply Blurred Infrared Target Extraction Based on Optimal Immune Field Neural Immune Network

Deep blurring is a kind of expression feature of blurred infrared images. The accurate extraction of the deeply blurred region in infrared images is the foundation of extracting blurry infrared targets. On the basis of the excellent characteristics of recognition, learning, memory, tolerance and coordination exhibited by biological immune systems in antigen detection, extraction and elimination, a deeply blurred infrared target extraction algorithm based on optimal immune field neural immune network is proposed by combining the relationship between the nervous system and the immune system in biological immunity. The algorithm can provide a guiding role for the immune network in target and background classification of blurred infrared images by designing a neural network. By relying on the function of prior knowledge of neural network independent of the immune system, an optimal immune field neural immune network is designed and accurate extraction of blurred infrared targets is implemented. The experimental results show that the algorithm can extract targets in blurred infrared target images more effectively and accurately than other traditional target extraction algorithms for blurred infrared target images.