Authors: Pratiksha Chaudhari, Mark Cheng – University of Alabama
Title: Using Deep Learning for Detection of environmental samples of Microplastic in Real-Time
Abstract: Microplastics pose a significant threat to the environment and human health. Traditional methods for detecting microplastics in water are often slow and costly. However, new sensors leveraging computer vision and deep learning have been developed for this purpose. Recent studies utilized the YOLOv5 deep learning model for microplastic detection, demonstrating superior performance compared to earlier versions and other algorithms. This model is capable of running on devices with limited processing power and can identify a diverse range of microplastic shapes, sizes, and colors, including those that are small and irregularly shaped. Additionally, high-speed cameras were used to capture the swift movements of microplastics in water samples, enhancing detection accuracy and efficiency. YOLOv5 shows significant promise for real-time monitoring and management of environmental pollution.