融合改进YOLOv5算法的图像全站仪全自动测量方法

Fully automated measurement method of image total station based on the improved YOLOv5 algorithm

针对图像全站仪在无棱镜合作工作模式下无法实现目标点全自动测量的问题,提出一种融合改进YOLOv5算法的图像全站仪全自动测量方法。应用融合卷积注意力机制模块的YOLOv5算法,实现了反射片靶标的广角镜头识别与检测;应用目标自动照准算法,实现了反射片靶标中心的长焦镜头精确照准,进而实现目标点位置坐标的全自动测量。借助自研的图像全站仪开展了反射片靶标的识别与检测实验和目标点全自动测量实验。实验结果表明,利用改进的YOLOv5算法对反射片靶标的识别与检测的准确率可达98.65%;目标点全自动测量方法具有与人工照准测量方法相当的测量精度且测量效率较后者提高了1.5倍。所提方法具有较高的测量精度和测量效率,可广泛应用于无人值守的全自动测量工作场合。

The image total station cannot achieve the fully automated measurement of the target point in the prism-free cooperative working mode.To address this issue,a fully automated measurement method of image total station based on the improved YOLOv5 algorithm is proposed.The YOLOv5 algorithm fused with the convolutional block attention module is used to realize the wide-angle camera identification and detection of the reflector.And the target automatic aiming algorithm is applied to realize the accurate aiming of the telephoto camera at the center of the reflector,which realizes the fully automated measurement of the position coordinates of the target point.With the help of the self-developed image total station,the identification and detection experiment of the reflector and the fully automated measurement experiment of the target point are carried out.Experimental results show that the accuracy of identifying and detecting reflector targets by the improved YOLOv5 algorithm can reach 98.65%.Compared with manual photometric measurement method,the fully automated measurement method of target point has comparable measurement accuracy and increases the measurement efficiency by 1.5 times.The proposed method has high measurement accuracy and measurement efficiency,which can be widely used in the unattended and fully automated measurement work occasions.