| 摘要: |
| 为提升霍尔测试仪的测试效率，本文提出了一种综合应用运动控制、物体检测和机器视觉等技术的智能辅助操控系统，以外接于霍尔测试仪的键盘按钮作为检测目标，并以优化的you only look once （YOLO） v5算法损失函数和检测头的结构改进测试系统，提升检测性能.通过改进机械臂的运动模式，优化运动轨迹，以保证机械臂平稳准确点击目标按钮，完成自动化测试任务.通过以上改进，相机目标检测精度提升了10.8%，达到87.3%；机械臂点击正确率提升了10.0%，达到95.0%；系统辅助霍尔测试仪完成测试成功率提升了20.0%，达到80.0%. |
| 关键词: 目标检测 自动化 人工智能 键盘识别 |
| DOI：10.20192/j.cnki.JSHNU（NS）.2025.02.014 |
| 分类号:TP391.4 |
| 基金项目: |
|
| A hall effect measurement assisted manipulation system based on improved YOLOv5 |
|
HE Qingqing, WANG Yang, YANG Li, WANG Yihan, GUO Zhiqiang, JIANG Chenghan, WU Yifan
|
|
College of Information, Mechanical and Electrical Engineering, Shanghai Normal University, Shanghai 201418, China
|
| Abstract: |
| To enhance the testing efficiency of the Hall effect measurement， an intelligent assistive control system that integrated technologies such as motion control， object detection， and machine vision was constructed. The external keyboard buttons on the Hall tester set as the detection targets were used in the system. The detection algorithm was improved through the optimization of the you only look once （YOLO） v5 loss function and the structure of the detection head， in order to promote detection performance. Additionally， by changing the robotic arm motion mode ， the motion trajectory was optimized to ensure stable and accurate button clicks for automated testing tasks. With the above improvements， the system object detection accuracy was increased by 10.8%， reaching 87.3%； the robotic arm correct click rate was increased by 10.0%， reaching 95.0%，and the completion rate for the system assisted by Hall tester was increased by 20.0%， reaching 80.0%. |
| Key words: object detection automation artificial intelligence keyboard recognition |
