对行人摔倒进行自动检测,可以为行人的生命安全提供保障。设计基于行为视觉的行人摔倒自动检测系统,系统硬件包括视频采集操作模块和视频轮廓采集模块,采用TMS320F2812芯片作为系统的核心控制单元,通过TMS320F2812芯片和SAA7111芯片构建视频采集处理模块,将视频采集模块采集的图像进行轮廓的描述,获取行人不同姿态的运动目标的特征,实现行人摔倒自动检测。仿真实验结果表明,所设计系统对复杂环境的适应能力较强,行人发生摔倒等事件检测正确率较高,可以应用于智能交通监控系统中。
The automatic detection for pedestrians tumble can provide the safeguard for the life safety of pedestrians. The pedestrian tumble automatic detection system based on behavior vision was designed. The system hardware includes the video acquisition and operation module, and video contour acquisition module. The TMS320F2812 chip is taken as the core control unit of the system. The TMS320F2812 chip and SAA7111 chip are used to construct the video collection and processing module. The contour of the image acquired by video acquisition module is described to obtain the moving target characteristics of pedestri- ans' different poses, and detect the pedestrian tumble automatically. The results of simulation experiment show that the designed system has good adaptive ability to the complex environment, can improve the detection accuracy of pedestrian tumble incidents, and can be used in the intelligent traffic monitoring system.