摘要
行人检测是机器视觉、图像处理等领域中很多有关应用的关键技术之一,在安防监控、无人汽车驾驶、机器人以及多媒体分析等方面具有广泛的应用前景和潜在的经济价值,因此成为了近年来的研究热点。本文介绍了基于opencv的行人检测方法,利用开源的跨平台计算机视觉库opencv在Visual Studio进行开发,调试及应用。通过提取行人的梯度方向直方图(HOG)特征,并利用事先准备的样本训练出来线性支持向量机(SVM)分类器对提取的行人特征进行分类,建立一种基于梯度方向直方图特征的行人检测系统。实验结果表明本设计符合要求,能够有效地检测出行人。
Pedestrian detection is one of the key techniques of many related applications in the field of computer vision and image pro-cessing, which has wide application prospect and potential economic value in the security monitoring, unmanned driving cars, robotsand multimedia analysis aspects. Thus, it has become research hot spot in recent years. This paper introduces the pedestrian detectionmethod based on opener, using open source computer vision library developed, debugged and applied in Visual Studio. The histogramof oriented gradient extracted from pedestrian will he classified by linear SVM classifier trained with samples prepared in advance. Apedestrian detection system based on HOG is constructed. Experimental results show that the design accords with the requirements andit can effectively detect the pedestrian.
出处
《惠州学院学报》
2016年第3期55-57,共3页
Journal of Huizhou University
基金
惠州市科技计划项目(2015B010002008)
广东省普通高校青年创新人才项目(2015KQNCX156)
