论文部分内容阅读
激光成像雷达的空间分辨率较高,能成四维像(强度像+三维距离像),适合作目标识别探测器.支持向量机(SVM)是一种能在小样本学习的情况下,仍有较高识别正确率的目标识别方法.通过优化支持向量机算法,将它嵌入到激光成像雷达系统的数字信号处理器(DSP)芯片内,实现目标识别的功能,有很高的现实意义.首先用真实激光成像雷达强度像做实验,测试56个样本,共耗时31.97μs,证明嵌入到数字信号处理器的支持向量机算法能满足实时性要求,识别正确率为98.2%;再用仿真激光成像雷达距离像验证支持向量机的推广能力,证明支持向量机在实时性和识别性能两方面都能满足激光成像雷达的识别要求.
Laser imaging radar has a high spatial resolution and can be used as a target recognition detector in the form of four-dimensional image (intensity image + three-dimensional distance image). Support Vector Machine (SVM) High recognition accuracy of the target recognition method.It is of high practical significance by optimizing the support vector machine algorithm, embedding it into the digital signal processor (DSP) chip of the laser imaging radar system, to achieve the target recognition function.First, Real laser imaging radar intensity as an experiment, testing 56 samples, a total time-consuming 31.97μs, support for embedded digital signal processor support vector machine algorithm to meet the real-time requirements, recognition accuracy was 98.2%; then the simulation laser imaging Radar distance as the verification of support vector machine to promote the ability to prove support vector machine in real-time and recognition performance to meet both laser imaging radar recognition requirements.