基于改进YOLOv5s机械臂按键的自动校准流程研究OA

An automated calibration execution system that combines an electrical instruments keystroke detection algorithm and a robotic arm touch positioning deployed in embedded development board is proposed to achieve partial automation of the calibration process.By using the lightweight backbone network,modifying the bounding box regression loss function,adding an attention mechanism to the algo-rithm,the number of model parameters is reduced and the accuracy of the target detection is ensured.The test results show that the av-erage accuracy of the improved model on the test set reaches 93.3%,and the detection time of single frame is reduced by 23.1%after the number of parameters is reduced.A robotic arm automated keystroke calibration execution process of function generators is de-signed,the test completion success rate reached 92.5%,which can basically attain the automatic calibration execution process.