Automatic control model training and target object recovery method and device based on machine vision

Abstract

The invention provides an automatic control model training and target object recovery method and device based on machine vision, and relates to the technical field of machine vision, and the method comprises the steps: obtaining target point positioning information and mechanical arm motion parameters, wherein the mechanical arm motion parameters comprise the included angles of a mechanical arm ina plurality of axial directions; determining the value of the loss function of the automatic control model according to the target point positioning information and the mechanical arm motion parameters; and adjusting the parameters of the automatic control model according to the value of the loss function until convergence conditions are met, and completing training of the automatic control model. According to the method, the positioning information of the target point of the grabbed object and the motion parameters of the mechanical arm are acquired as the data set, and the model is trainedso that the trained automatic control model has relatively high positioning calculation precision, and the trained model does not depend on the specific size of the claw, the precision can be prevented from being influenced by the claw with the fixed specification, and higher flexibility is achieved.

Description

technical field [0001] The invention relates to the technical field of machine vision, in particular to a method and device for automatic control model training and target recovery based on machine vision. Background technique [0002] With the continuous development of science and technology and the continuous improvement of people's living standards, various automation devices have emerged in large numbers, providing various conveniences for people's production and life with their high efficiency and fast features. For example, more and more large public restaurants are now equipped with tableware recovery devices for automatic tableware recovery, reducing the workload of staff in recycling tableware. [0003] Most of the existing tableware recycling devices are equipped with robotic arms for grabbing the tableware to be recycled to a designated location for subsequent processing. However, due to force majeure factors such as wear and tear of various parts and overall dis...

AI Technical Summary

Problems solved by technology

However, due to irresistible factors such as wear and tear of various parts and overall displacement of traditional low-cost robotic arm products, the accuracy of coordinate calculations drops rapidly

And the product must be delivered by high-end logistics services to ensure the safety of the transportation process, and the installation and commissioning must also be accurately calibrated by the professionals of the developer

After completing about 500 times of grasping work, mechanical parts will inevitably have problems of inaccurate grasping and reduced work quality, which seriously affects the application and promotion of tableware recycling devices

Images