• NURUKURTHI MANOJ SAI
• Dr. PRIYANKA K BHANSALI

Microplastics have become a growing concern due to their harmful impact on marine life and water quality [5]. Identifying these tiny plastic particles manually is time-consuming and often inaccurate. In this study, we propose an image classification system that uses deep learning—specifically the efficient net-b0 model—to automatically detect microplastics in water samples [2],[4],[1]. To improve accuracy, we trained separate models on five different colour space versions of the same images: RGB, HSL, YCbCr, Grayscale, and FFT. By combining the outputs through an ensemble method, we were able to achieve more reliable predictions. Among all colour spaces, HSL performed the best individually, but the ensemble approach offered balanced and stable results across all cases [14],[15]. The model was evaluated using accuracy, precision, recall, F1-score, and AUC, showing strong performance overall. This system can help researchers and environmental agencies monitor water bodies more efficiently and support early detection efforts to address plastic pollution [1],[3].

Microplastic detection, Image classification, Deep learning, Color space fusion, efficient net-b0, Ensemble learning, Environmental monitoring, CNN, HSL, FFT, YCbCr, Grayscale, RGB, Marine pollution

"IMAGE CLASSIAFICATION OF MICROPLASTICS", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.12, Issue 7, page no.g128-g137, July-2025, Available :http://www.jetir.org/papers/JETIR2507625.pdf

"IMAGE CLASSIAFICATION OF MICROPLASTICS", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.12, Issue 7, page no. ppg128-g137, July-2025, Available at : http://www.jetir.org/papers/JETIR2507625.pdf