强磁环境下室内设备的轮式机器人自动巡检方法

Automatic Inspection Method of Wheeled Robot for Indoor Equipment in Strong Magnetic Environment

强磁环境下轮式机器人难以准确定位仪表盘,导致仪表图像识别误差较大。为此,设计强磁环境下室内设备的轮式机器人自动巡检方法。建立表盘重点区域定位模型,通过二维卷积操作特征定位摄影机中影像重点区域;提取仪表图像特征,建立回归方程获取目标区域的特征点,得到排除强磁环境干扰的图像特征提取方法。设计轮式机器人巡检仪表自动识别算法,通过双向匹配法控制巡检结果的误差。实验结果表明方法对圆周指针型仪表的识别误差控制在0.5以下,可准确识别电子数字型仪表,提高仪表图像识别精度。

In a strong magnetic environment,the wheeled robot is difficult to accurately locate the instrument panel,resulting in a large instrument image recognition error.To this end,this paper designs a wheeled robot automatic inspection method for indoor equipment in a strong magnetic environment.It establishes positioning model of dial key area,locates the image key area in the camera through two-dimensional convolutional operation;extract instrument image features,establishes the regression equation to obtain target area,obtains image feature extraction method to exclude strong magnetic interference of the environment;designs automatic identification algorithm to control the error of inspection result by two-way matching method.The experimental results show that the identification error of circular pointer instrument is below 0.5 to accurately identify electronic digital instrument and can improve instrument image recognition accuracy.