As urban transportation systems become increasingly complex, efficient vehicle traffic analysis is essential for optimizing traffic flow and improving road management. Traditional traffic monitoring methods, such as inductive loops and radar, provide limited contextual insights compared to computer vision-based techniques, which make use of visual data for a more comprehensive understanding of traffic dynamics. This systematic literature review explores the role of computer vision in vehicle traffic analysis, focusing on key applications such as vehicle detection, counting, classification, automatic number plate recognition (ANPR), traffic incident detection, violation detection, and traffic signal optimization. An emphasis is placed on vehicle detection and tracking, as these foundational steps enable accurate traffic flow analysis. We examine various implementations, including YOLO and R-CNN for object detection and Deep SORT for object tracking, highlighting the algorithms that enhance detection and tracking efficiency. Additionally, we discuss the advantages and limitations of these methods, addressing challenges such as object occlusion, computational constraints, and the trade-offs between model accuracy and performance which all have to be addressed when implementing a real-time solution for traffic analysis using computer vision.