This paper presents a novel approach to garbage container detection and analytics utilizing YOLO (You Only Look Once) models, leveraging a newly developed dataset tailored for urban waste management. Traditional waste collection systems often suffer from inefficiencies, leading to increased operational costs and environmental impact. To address these challenges, we introduce a comprehensive dataset specifically designed for training and evaluating object detection models in the context of urban waste management. Our dataset includes diverse images of garbage containers in various environments and conditions. By employing four YOLO models, renowned for their real-time object detection capabilities, we achieve high-accuracy container detection with minimal computational overhead. The detected containers can be analyzed to provide actionable insights, such as fill level estimation and collection route optimization in next steps of garbage collection. This container-based analytics framework enables dynamic and efficient waste collection scheduling, reducing fuel consumption and operational costs while enhancing service quality. Our experimental results demonstrate the effectiveness of YOLO models in accurately detecting garbage containers, and the subsequent analytics highlight significant improvements in waste management operations. This work underscores the potential of advanced computer vision techniques in transforming urban waste management through enhanced data-driven decision-making processes.