论文部分内容阅读
随着目标自动跟踪、识别技术的发展,对目标特征提取技术的要求变得愈来愈高,基于目标形状的跟踪、识别系统的发展非常迅速。本文针对目标红外图像的特点,提出了一种新的目标形状分类、识别系统,该系统包括目标图像预处理、目标分割提取、目标边界探测以及目标形状分类、识别。系统中,我们采用一种均值——中值混合滤波方法来消除噪音的影响,利用峰谷法分割提取目标,进而提取目标的边界,在目标边界提取的基础上,提出了边距离分布特征的概念用于目标的分类、识别。
With the development of automatic tracking and recognition technology, the requirements of the target feature extraction technology become more and more high. Based on the tracking of the target shape, the recognition system has been developed rapidly. In this paper, a new target shape classification and recognition system is proposed according to the characteristics of target infrared images. The system includes target image preprocessing, target segmentation and extraction, target boundary detection and object shape classification and recognition. In the system, we use an average-median mixed filtering method to eliminate the influence of noise, and use the peak-valley method to extract the target and extract the boundary of the target. Based on the extraction of the target boundary, we propose the feature of the edge distance distribution Concept for the classification of the target, identification.