Journal

In recent times, the presence of microplastic emerged as a serious environmental threat rendering ecological risk and human health hazard. Contemporary research indicates that microplastic is omnipresent in the environment including terrestrial, aquatic, aerial, and even biological environments i.e., within organisms as well as the human body. Therefore, detection and analysis of microplastics in the environmental matrix is a decisive task, which is further a necessity for the prevention and removal of microplastic pollution. However, microplastics arise from diverse sources and are of diverse types, and need to be detected for a wide variety of environmental matrices. Thus, to understand the level of microplastic pollution in the environmental matrix, the complexity associated with microplastic detection and analysis which include qualitative and quantitative detection followed by classification of microplastic according to the type of polymer, size, and shape, structure types (fiber, fragment, film), etc. Microplastic pollution in the environmental matrix is assessed either by microscopy and visual sorting or by spectroscopy. Many researchers have developed methods of visual detection using microscopes which are generally easy to apply but require a lot of human work time and are likely to reveal misleading results with a lack of further information on types of microplastics. While spectroscopy is a simple method to apply to a large number of samples, further complexity is associated with classifying microplastic. To solve these problems, scientists have resorted to the application of artificial intelligence (AI) for better detection and classification of the different types of microplastics accumulated in samples taken from various ecosystems during the last decades. Integration of AI with microscopic or spectroscopic detection of microplastic can be a forensic tool for microplastic detection to reduce the complexity associated with detection and identification. Machine learning or Artificial Neural networks can be a powerful tool for processing the images obtained by spectroscopy or microscopy for automatic and fast screening/classification of microplastics. AI-based detection of microplastic pollution in environmental matrix opens a new scope for big data processing with interpretability to provide reliable results and prediction. This study review methods for detecting microplastics in the environmental matrix using AI developed by researchers to automate and accurate categorization of microplastics in the environment.