基于深度学习的智能分类垃圾桶

AbstractWith the advancement of human science and technology and theimprovement of quality of life,more and more domestic waste is generated.Therefore,how to effectively recycle and treat domestic waste has becomethe focus of people's attention.The investigation and study found that themethod of sorting waste at the source is the most efficient and the mostthorough in the entire waste sorting process.One of the current classification rules is that it is difficult to unify allregions,on the other hand,there are many types of garbage classification,which are difficult to distinguish,which will have a certain impact onpeople's quality of life.At present,there are a variety of"smart trash cans"on the market,butbasically only the automatic opening and closing function is realized,and theability to sort and identify trash is not available.The current artificialintelligence technology is developing rapidly,and the direction of imagerecognition in the field of deep learning has made great progress,making itpossible to use image recognition technology to classify garbage.Choosing the correct image recognition classification algorithm is thetop priority of this project.Therefore,this article introduces and summarizesthe algorithms that have made outstanding contributions in the past in imagerecognition,grasping the development context,and thus explaining thereasons for choosing the MobileNetV2 network.In this project,the system can be divided into a garbage recognitionmodule and a garbage classification delivery module,namely a recognitionmodule and a control module.After the garbage is put into the garbage bin,the recognition module classifies the garbage,and then sends theclassification result to the control module,and the control module puts it intothe corresponding garbage bin.After actual testing,the classificationaccuracy rate of garbage can reach 90%,and the recognition andclassification speed is after filtering the classification results,and the averagetime for successfully identifying a garbage is 4 seconds.This proves that theproject is still very promising in the field of intelligent waste classification.