基于OpenMV的自动追踪识别仪表小车设计

Design of a Car Based on OpenMV Automatically Tracking and Identifying Instruments

针对电气仪表在测量过程中面临的人工成本高昂及传统方法在工业环境下识别准确率不高的问题,设计了一种基于OpenMV的自动追踪识别仪表小车.首先,将OpenMV固定在二维云台上,用于被测物体自动追踪,同时与主控制器连接,进行数据传输;然后,使用测距算法测量小车与目标物之间的距离,在室内铺设AprilTag标识物用于定位和导航,当小车检测到仪表盘时,判断其类型;最后,实验验证巡检小车在行进过程中,准确识别到仪表盘位置,分辨仪表盘类型,并自动记录仪表盘图像,具有较高的识别准确率.所设计的OpenMV自动追踪识别仪表小车造价便宜,同时小车硬件设计和算法方面都具备小型化和轻量化的特点,大大提高了测量的智能化水平.实验结果表明,在损失较少精度的情况下,提高了检测速度,最高为22帧/s,在电力巡检中具有一定的实际应用价值.

Aiming at the problem of high labor cost and low identification accuracy of traditional methods in industrial environment,a car based on OpenMV that can automatically track and identify instruments was designed.First of all,OpenMV is fixed on the 2D holder in this design,which is used for automatic tracking of the object under test.Meanwhile,it is connected with the main controller for data transmission.Then,distance measurement algorithm is used to measure the distance between the car and the target,and AprilTag markers are laid indoors for positioning and navigation.When the car detects the dashboard,judge its type.Finally,the experiment verifies that the patrol car can accurately identify the location of the instrument panel,distinguish the type of the instrument panel,and automatically record the image of the instrument panel in the process of moving,which has a high recognition accuracy.The designed OpenMV automatic tracking and identification instrument car is cheap in cost and has the characteristics of miniaturization and lightweight in hardware design and algorithm,which greatly improves the intelligent level of measurement.The experimental results show that the detection speed is improved with a maximum of 22 frames/s at the expense of less precision,which has certain practical application value in electric power inspection.