| 摘要: |
| 为了在自行车转向或变道时，实现无需左右回头观察即可了解后方路况，设计了一套便于骑行者感知车后具体情况的自行车后视系统.该系统可分为车尾探测装置与终端反应手套，主要由USB摄像头与执行模块构成，并采用了you only look once（YOLO）v4目标检测算法.通过在自行车车尾安装检测设备，与骑手的特制手套进行无线通信，从而传递车后方探测到的有关安全状况的紧急信息.实验结果表明：该系统能够对后方的行进车辆进行多目标检测及警告优先级目标，并在足够的安全距离内识别率较高. |
| 关键词: 自行车 安全系统 手套 目标检测 单目测距 单片机 树莓派 |
| DOI：10.3969/J.ISSN.1000-5137.2023.02.011 |
| 分类号:TP18 |
| 基金项目: |
|
| A rear visual system based on tactile sensation for bicycle-riders |
|
YUAN Bo, LIU Xiangpeng
|
|
College of Information, Mechanical and Electrical Engineering, Shanghai Normal University, Shanghai 201418, China
|
| Abstract: |
| In order to enable the cyclist to keep abreast of the road conditions behind him without looking back when changing lanes or turning at an intersection, a bicycle rear-view system for riders was designed to sense the potential rear dangerous factors, such as the high-speed cycle which was going to overtake the rider on the left or right, to improve safety. This system could be divided into two parts:the detection system and the reaction gloves. The detection system described in this paper was constructed with USB camera as the main data acquisition sensor and the executive mechanism, as well as you only look once(YOLO) v4 as the main vision recognition algorithm. By installing the camera at the rear of the bicycle and transmitting the information wirelessly to the special gloves, the rider could be effectively informed with the emergency detected, and then take precautions immediately. The experimental results showed that the constructed system was able to detect the moving rear-vehicles with considerable accuracy within a safe distance, select the most threatening targets, and display on the gloves. |
| Key words: bicycle safety system gloves object detection monocular ranging one-chip computer Raspberry Pi |
