基于深度学习的气泡水平尺自动矫正系统研究

Automatic bubble level correct system based on deep learning

针对气泡水平尺大角度偏斜工件的自动矫正问题,对深度学习目标检测领域的模型进行了研究,设计了一种基于深度学习的自动矫正方案。在不同光照条件的气泡工件图片样本集上,采用k-means聚类方法,分析了不同anchor box数量对YOLO检测性能的影响,对比了YOLO与SSD两种模型的检测准确率和平均IOU;同时结合概率霍夫变换与最小二乘法,设计了两种气泡工件参考线边缘拟合方法,并对比了两种拟合方法计算工件偏斜角度的准确度;针对应用场景,采用Client/Server网络结构,服务端接收客户端采集的图像并将计算结果返回,客户端控制电机对偏斜工件进行了自动矫正。研究结果表明:该方案能对气泡大角度偏斜的工件进行检测,相比现有方案矫正的效率更高。

Aiming at the automatic correction of the bubble level with large angle of inclination, the models in the field of deep learning ob- ject detection were studied, and a method based on deep learning object detection models was designed. The YOLO and SSD models were trained on the data set of bubble level images with different light conditions. By adopting K-means clustering algorithm, the effect of different anchor box numbers on YOLO model＇s performance was analyzed. And two models＇ detection accuracy and average IOU werecompared. Two edge fitting methods combining Progressive Probabilistic Hough transfoml with least square method were designed and two methods＇ accuracy was compared. Aiming at this application, Client/Server network structure was used. The images were sent to the server and calculated. The calculation results were sent back to the client to control motor to correct the bubble level. The results indicate that the proposed method can detect the bubble level with large angle of inclination, and is more efficient than the conventional methods.