This project is designed to address real-world challenges in detection and tracking a very large number of objects in video . By integrating YOLOv10 and Deep SORT, it combines speed and accuracy for video processing. Our use of filters, grid-based processing, and NMS demonstrates innovation and enhancing detection performance in complex scenarios up to 1.75 from normal YOLOv10. Beyond practical impact, this project offers a valuable learning experience, strengthens expertise in computer vision, and provides a scalable foundation for applications like crowd analysis, wildlife monitoring, and robotics. It’s a step toward impactful, intelligent systems while advancing personal and professional growth