Paper Details

Name: Murali D, P Mohammed Sulaiman, Y Prudhvi Raj, S Obulesu & G Lokesh
Country: India

Year: 2026
Volume: Volume No: 13, January, Year: 2026 (Special Issue)
Page Number: 522-528
DOI: https://doi.org/10.5281/zenodo.19063657

ABSTRACT —
Maintaining urban cleanliness and optimizing waste management are critical for sustainable city development. Traditional waste management systems rely on manual reporting, which often results in inefficiencies, delayed response times, and unorganized waste disposal. These limitations contribute to environmental degradation, public health risks, and ineffective resource utilization. To address these challenges, this paper presents an advanced, real- time waste management system that leverages Bootstrapped Language-Image Pre-training (BLIP) alongside modern web technologies to enhance automation in waste detection and reporting. The proposed system allows users to capture and upload images of waste accumulation in public areas. These images are processed using a fine-tuned BLIP-2 deep learning model, which automatically generates detailed textual descriptions of the waste conditions. This AI- powered description is then used to notify municipal authorities, enabling them to take timely action. Unlike traditional manual approaches, the system provides accurate and real-time waste monitoring, reducing delays in waste collection and improving urban cleanliness. To ensure seamless operation, the system incorporates live location
tracking, allowing authorities to pinpoint the exact site of waste accumulation. The web-based interface is developed using Angular for the frontend and ASP.NET for backend processing, ensuring a responsive and scalable architecture. By integrating deep learning and modern web technologies, this system introduces a structured and automated approach to waste management, improving efficiency, reducing environmental hazards, and promoting a cleaner urban environment. This paper discusses the system’s design, implementation, and impact, demonstrating its potential to revolutionize waste management strategies in smart cities.

Keywords— Machine Learning, Reverse Geocoding API, Deep Learning, BLIP-based model.