Design and Method Research of Intelligent Detection System for Marine Microplastics Driven by Microfluidic Chip

Marine microplastics have become a global health threat due to their widespread pollution, difficulty in degradation, and access to the human body through contaminated seafood. To deal with such a problem, this paper proposes a structural model of intelligent detection system for marine microplastics based on microfluidic chip, which is combined with microfluidic chip fluid dynamics simulation to capture seawater microplastics samples. This paper also verifies the effectiveness of microfluidic chip in sample detection and data extraction. Then the public dataset is used with Raman spectroscopy to rapidly detect seawater microplastics. Besides, efficient classification and identification of microplastics through the convolutional neural network (CNN) model are achieved. The experimental results show that the system can capture and identify a variety of microplastic particles, with 93% as the recognition and classification accuracy rate, and 98±0. 02% as the average ROC area. The intelligent detection system provides an innovative microplastic detection solution and efficient technical support for future marine environmental monitoring.