An intelligent garbage sorting and recycling method and terminal based on audio-visual integration

Abstract

The invention discloses an intelligent garbage sorting and recycling method and terminal based on audio-visual integration. Collect the image number data of the complete images of garbage of different materials, and establish and train the preprocessing convolutional neural network and the first sub-classification model; collect the sound signals of the impact sound of garbage of different materials, and establish the second sub-classification model and train it; first Input to the preprocessing convolutional neural network to judge whether it is a single garbage or a mixed garbage: then input the image and sound data into the first and second sub-classification models for processing to obtain the garbage type results of the two models, and combine the weights to get the final result . The intelligent garbage sorting terminal includes a garbage bin body, an automatic sensor, a sorting mechanism, a garbage storage box and a control display. The invention can form the digital management and control of the entire waste treatment production process, and improve the waste treatment efficiency.

Description

technical field [0001] The present invention relates to a garbage sorting and processing method and a terminal, in particular to an audio-visual-based intelligent garbage sorting and recycling method and a terminal. Background technique [0002] With the explosive growth of global garbage production year after year, garbage recycling has become an extremely important part of municipal solid waste management. There are many problems in relying on residents to consciously sort garbage when throwing out garbage, such as high education costs, difficult supervision, and difficulty in ensuring the accuracy of classification, etc., which restrict the implementation of more refined garbage classification schemes. Existing intelligent garbage classification methods are usually based on a single signal such as vision, which requires extremely high sample size and cannot meet the needs of real scenarios. In response to these problems, the present invention explores a new intelligent g...

AI Technical Summary

Problems solved by technology

There are many problems in relying on residents to consciously sort garbage when throwing out garbage, such as high education costs, difficult supervision, and difficult to guarantee the accuracy of classification, etc., which restrict the implementation of more refined garbage classification schemes

Existing intelligent garbage classification methods are usually based on a single signal such as vision, which requires extremely high sample size and is difficult to meet the needs of real scenarios

Images