Abstract:Microplastics (MPs) pollution has become an important issue in the global environmental field. In particular, their persistent accumulation in soil environments could pose significant threats to ecosystems and human health over time. Consequently, the development of effective detection and identification techniques for MPs is essential for environmental monitoring and remediation efforts. While advances have been made in the identification of MPs using Raman spectroscopy coupled with machine learning, a significant challenge remains: the requirement for extensive datasets, the collection of which can be both time-consuming and costly. Thus, this study aims to explore a Raman imaging technique that combines hyperspectral unmixing methods with machine learning for the identification and visualization of MPs in soil. The primary advantage of this approach lies in its ability to achieve high-precision identification of target samples without reliance on large datasets, making it both cost-effective and efficient.