基于行动感测的实时坑洞侦测系统与方法

A Pothole Detection System and Method Based on Mobile Sensing Techniques

提出一个以行动感测和云端运算为基础的实时坑洞侦测方法,通过分析内嵌于手机内的加速度传感器数据,并利用欧拉角进行手机加速度正规化的动作,改善搜集加速度数据时需要固定手机姿态角的限制。结合坑洞侦测算法进行坑洞判定,以及使用空间内插法改善全球定位系统(GPS)定位误差,将坑洞信息回传至云端服务器。再将此坑洞信息实时传输至每个使用者的行动式手持装置中,并搭配扩增实境(AR)导航功能,让用户可以避开坑洞路段。实验表明,所提出的基于行动感测的实时坑洞侦测方法将可以准确判断出坑洞。

A real-time pothole detection approach is proposed based on mobile sensing and cloud computing. The Euler angles are used to normalize the accelerometer data of mobile device. A pothole detection algorithm is applied to find pothole, and the spatial interpolation method is used to improve accuracy of global positioning system （GPS）. The pothole information is sent to a cloud server. Finally, the pothole information can be transferred to user＇s mobile device in real time, and augmented reality （AR） navigation service is combined in the mobile device for pothole avoidance. According to practical experiments, the results indicate that this proposed method for pothole detection based on mobile sensing techniques can precisely detect potholes without any false-positives.