Abstract:

In the field of environmental protection, recycling of resources and social livelihoods, wasteclassification was always a crucial subject. A deep learning automated waste classification approach is introduced to enhance the efficiency of the front-end waste collection. With the fast increase in global production levels, the problem of garbage disposal is growing severe. Trash classification is an important step towards waste reduction, harmlessness and resource utilization. Increasing trash types and quantities implies that traditional scrap classification algorithms can no longer comply with accurate identification requirements. This study offers a VGG16 neural network model based on the process of attention for classifying recyclable waste. The attention module is introduced to the model after convolution so that the essential information in the feature map may be given greater attention. The algorithm can automatically extract categorization features such as organic, recyclable and non-recyclable waste. Experimental findings reveal that 84 per cent of the algorithm in the recyclable trash classification can effectively categories the garbage.

Keywords: