Recycling bottle classification method based on deep learning model

Abstract

The invention discloses a recycling bottle classification method based on a deep learning model, and the method comprises the following steps: S01, obtaining recycling bottle images, carrying out themarking and category definition of each recycling bottle image, and forming a training image data set, a verification image data set and a test image data set; S02, establishing a deep learning model,and training the deep learning model by adopting the training image data set; S03, verifying and testing the trained deep learning model by adopting the verification image data set and the test imagedata set, and returning to the step S02 to continue deep learning model training if a verification or test result does not reach a set value; if the verification or test result reaches a set value, obtaining a prediction model; and S04, loading the prediction model in a recovery device, and enabling the output result of the prediction model to control the manipulator to classify the recovered bottles. The invention provides a recovery bottle classification method based on a deep learning model, which is used for improving the identification and recovery efficiency of recovery bottles.

Description

technical field [0001] The invention relates to the field of computer vision, in particular to a recycling bottle classification method based on a deep learning model. Background technique [0002] As my country's emphasis on ecological protection and environmental sanitation continues to increase, the recycling of garbage is becoming more and more standardized and industrialized. In the process of garbage recycling, it is hoped that the efficiency of garbage recycling will be improved. Therefore, the intelligent garbage disposal and recycling industry is rapidly developing. develop. Use automated equipment to share part of the work of manual sorting of garbage, so as to improve recycling efficiency and quality. [0003] Among them, automation equipment involves the fields of deep learning, computer vision and image processing. At present, due to the huge amount of garbage and the small number of people in the garbage recycling industry, most of them are not young and middl...

AI Technical Summary

Problems solved by technology

At present, due to the huge amount of garbage, and the small number of people in the garbage recycling industry, most of them are not young and middle-aged, the recycling environment is poor, manual sorting and processing, the efficiency is low, and the false detection rate is high

Images