This paper focuses on solving a deep learning problem involving the detection of multiple objects in a live streaming public camera, which is prone to various challenges like blurring, noise, rotation jitter, and other factors. These issues arise due to high-speed object captures, varying lighting conditions (too bright or dim), foggy or dusty atmospheres, and occlusion by other objects, leading to significant data loss. The primary goal of this paper is to leverage the You Only Look Once (YOLO) method for object detection. YOLO excels in detecting objects faster than other algorithms and comprehensively examines the entire image, unlike methods such as Convolutional Neural Network (CNN) or Fast-Convolutional Neural Network (F-CNN) which may not fully explore the image. The model developed in this project represents an enhanced version of traditional object detection as it demonstrates the capability to function effectively with low image quality, achieves high accuracy, and requires fewer object features for detection.