论文部分内容阅读
激光成像雷达目标识别是当前国际上的研究热点,为了提高激光成像雷达目标的识别率,提出一种滤波器与神经网络的激光成像雷达目标识别方法。首先对感兴趣雷达目标进行检测和定位,收集相应的信息,然后采用滤波器对信息进行处理和提取特征向量,然后采用极限学习机神经网络构建激光成像雷达目标识别的分类器,实现激光成像雷达目标的识别,采用实验数据对4类目标进行仿真测试。实验结果表明,本文方法的激光成像雷达平均识别率超过了95%,识别实时性远优于传统方法,可以满足激光成像雷达的识别要求,具有较好的应用推广价值。
Laser imaging radar target recognition is the current international research hot spot. In order to improve the recognition rate of laser imaging radar target, a method of laser imaging radar target recognition based on filter and neural network is proposed. Firstly, the radar target of interest is detected and located, the corresponding information is collected, and then the filter is used to process the information and extract the eigenvector. Then, an extreme learning machine neural network is used to construct the classifier of laser imaging radar target recognition, and the laser imaging radar Target identification, the use of experimental data on the four objectives of the simulation test. The experimental results show that the average recognition rate of the laser imaging radar in this method is more than 95%, and the real-time recognition is much better than the traditional method, which can meet the recognition requirements of laser imaging radar and has a good application and popularization value.