Recognition of ethylene plasma image based on dual residual with attention mechanism network

Abstract

Ethylene discharge can be used for particle formation in complex plasma, industrial plasma process, environmental protection, and agricultural process. Ethylene discharge characteristics strongly depends on discharge parameters. Accurate and efficient recognition of discharge parameters is significant for the diagnosis of complex plasma, and industrial and agricultural practical applications. In this paper, we proposed a deep convolution neural network based on dual residual with attention mechanism (DRAM) to recognize discharge parameter through the image fusion of discharge glow and particles during ethylene discharge. It shows that the proposed model can effectively recognize the ethylene discharge parameters with all the four evaluation indicators of accuracy, precision, recall, and F1_Score of higher than 98.8%, respectively. Compared with other six classical recognition models, our model exhibits the best recognition performance. This method provides an effective technical support for the diagnosis and practical application of ethylene plasma.

Graphic abstract