Review on Evolution of Vehicle Detection Methods in Computer Vision for Surveillance

One major area of computer vision is object detection research that tries to automatically detect and track items of interest in camera video streams. The goal of object detection using video is to provide real-time, accurate, and robust detection and tracking of objects, such as vehicles, people, animals etc., for effective monitoring and security solutions. The ability to detect and track objects in video streams can help to identify potential threats, prevent criminal activity, and improve situational awareness. Appearance-based and motion-based are two categories of object detection. Instead of sitting through endless hours of monitoring the live feed from security cameras, computer vision methods are deployed. There are many challenges associated with object detection including occlusion, cluttered backgrounds, and changes in lighting and weather conditions, avoid traffic on the roads and parking lots and to prevent accidents, a variety of image processing algorithms have been created with the goal of detecting and recognizing moving objects from cameras. A variety of algorithms and techniques are applied to solve these challenges such as neural network based approaches, feature-based approaches, and hybrid approaches that combine multiple techniques have been summarized.