基于模糊数学模型的舰船红外成像目标智能识别方法

Intelligent recognition method of ship infrared imaging target based on fuzzy mathematical model

利用传统基于SVM和基于神经网络的方法对舰船红外成像目标进行智能识别,识别距离较短,导致识别范围受限。针对上述问题,提出基于模糊数学模型的舰船红外成像目标智能识别方法。该方法分为3步:1)对舰船红外图像进行预处理,包括图像滤波、图像增强、图像分割;2)利用基于几何特性方法提取处理后的图像特征;3)以图像特征作为模糊数学模型特征因子,构建模糊集合,并利用贴近度原则对被识别对象进行归属判决,完成目标识别。结果表明:与基于SVM和基于神经网络的方法相比,利用本方法进行舰船红外成像目标智能识别,识别距离延长10 m和20 m,识别范围扩大。

The traditional method based on SVM and neural network is used for intelligent recognition of ship infrared imaging target. The recognition distance is short and the recognition range is limited. Aiming at the above problems, an intelligent recognition method of infrared imaging target based on fuzzy mathematical model is proposed. The method consists of three steps: first, preprocessing infrared images of ships, including image filtering, image enhancement and image segmentation;second, extracting image features based on geometric characteristics;third, using image features as feature factors of the fuzzy mathematical model to construct a fuzzy set, and using the principle of proximity. The recognition object is judged by attribution, and the target recognition is completed. The results show that, compared with the methods based on SVM and neural network, this method can be used for intelligent recognition of ship infrared imaging targets. The recognition distance is extended by 10 and 20 m, and the recognition range is expanded.