首页> 外文期刊>International journal of computational intelligence research >Study on Vehicle Accident Avoidance System Using Kinect Depth Sensor Along with Notifications on Mobile Devices
【24h】

Study on Vehicle Accident Avoidance System Using Kinect Depth Sensor Along with Notifications on Mobile Devices

机译:基于Kinect深度传感器的移动事故预防系统与移动设备通知的研究

获取原文
获取原文并翻译 | 示例
           

摘要

In this paper, Kinect depth sensor, a low-cost range sensor released by Microsoft is used as vision sensor to monitor vehicle's surroundings by detecting objects in front of it and measuring how far they are from the sensor. Therefore, our monitoring system comes with computer vision techniques which imitate natural human vision system letting unmanned (robots) or manned vehicles to understand the environment around them by extracting and analyzing useful information from a single or a sequence of images taken by cameras, here Kinect camera. Several technologies have been used to deal with computer vision applications such stereovision, LIDAR and RADAR but each of them still has advantages and disadvantages in term of limitations and price. Firstly, experiments have been conducted with Kinect depth sensor by detecting the obstacles in front of the sensor and measuring the distance from them to depth sensor in accuracy way by checking if the detected object is real or not to reduce processing time ignoring pixels which are not a part of real object in such real-time applications. Results are promising and also Kinect accuracy have been tested on distance measurement function based on the location of detected object, different environments (brightness and darkness), physical measurements and system outputs. In our case, we suppose a vehicle equipped with Kinect camera as imaging sensor in order to take more precautions to avoid accident by helping the driver to be warned with simple message or an audio about some objects in front of the vehicle or get proof in case of accident that the driver didn't take care or not by saving color images and distances in web server whenever there is a warning. Also, a mobile application on android platform is developed to get notification whenever there is a warning about the presence of objects and at the same time distances and color images are recorded in web server for further analysis. Notification system can be used in many scenarios in case the driver is not the vehicle owner and also in parking area when the owner is notified in real time when other vehicle comes closer at a certain distance and can view the color image of the scene simultaneously when there is a warning.
机译:在本文中,由Microsoft发布的Kinect深度传感器(一种低成本的距离传感器)用作视觉传感器,通过检测前方的物体并测量它们与传感器之间的距离来监视车辆周围的环境。因此,我们的监控系统配备了计算机视觉技术,该技术模仿自然的人类视觉系统,使无人驾驶(机器人)或载人车辆通过从相机拍摄的单个图像或一系列图像中提取和分析有用信息来了解周围的环境。相机。已经使用了多种技术来处理计算机视觉应用,如立体视觉,LIDAR和RADAR,但每种技术在局限性和价格方面仍具有优缺点。首先,通过Kinect深度传感器进行实验,方法是检测传感器前面的障碍物,并通过检查检测到的物体是否真实来准确测量从障碍物到深度传感器的距离,以减少处理时间,而忽略不存在的像素。在此类实时应用程序中,真实对象的一部分。结果是有希望的,并且Kinect准确性已经基于检测到的物体的位置,不同的环境(明暗),物理测量和系统输出在距离测量功能上进行了测试。在我们的案例中,我们假设装有Kinect摄像头作为图像传感器的车辆是为了通过采取简单的消息或有关车辆前部某些物体的音频警告驾驶员,以防万一,以防发生事故,以防万一发生警告时,通过将彩色图像和距离保存在Web服务器中来避免驾驶员不小心的情况。此外,开发了Android平台上的移动应用程序,以便在有关于对象存在的警告时发出通知,同时将距离和彩色图像记录在Web服务器中以进行进一步分析。通知系统可用于许多情况下,如果驾驶员不是车主,也可以在停车区,当其他车辆在一定距离处更近时实时通知车主,并且在停车时可以同时查看场景的彩色图像有警告。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号