Vehicle re-identification for Visual Surveillance

Intelligent video surveillance systems are becoming an increasingly important major area of artificial intelligence as the use of motor vehicles in today’s transportation networks grows. In recent times re-identifying vehicles over a surveillance camera network have been proven to be the most effective method of efficiently controlling traffic, upholding the law, gathering information, and programming the traffic. Vehicle Re-ID seeks to recognize a target vehicle in several, non-overlapping camera perspectives. Compared to the common person re-ID issue, it has gotten far less attention in the computer vision community. The absence of pertinent research data and the unique 3D structure of a vehicle are two potential causes of this delayed progression. 

This research presents PAK Vehicle Re-ID Dataset (PV-ReID) with arbitrary viewpoints including unique vehicle identities based on Asian regions like Pakistan and India. In addition, a generalized pipeline for vehicle re-identification systems is presented.

In an Intelligent traffic monitoring system vehicle re-id plays an import role as vehicle re-id is a process in which we are going to detect a vehicle and then re-identify the vehicle from our database of vehicles. This can play an important role in the field of live monitoring or tracking the vehicles or doing forensic analysis like finding patterns of different vehicles etc. This requires intelligent and efficient algorithms and to solve this problem, we will be using the deep learning which is state of the art for many problems these days. Vehicle re-identification helps to determine whether a given vehicle has appeared in another camera. The technology is useful to automatically locate a certain vehicle in a network of cameras, mainly for surveillance applications. In practice, issues like variations in illumination changes, camera view angles, vehicle positions, etc. are the typical challenges that need to be solved by the vehicle re-identification technology.

In real-world practice, humans can treat this task in a progressive manner. For instance, if the security officer needs to find a suspect car in a city with large-scale video surveillance networks, appearance attributes such as models,types, and colors can be initially used to find similar vehicles and reduce the search field. Then, they can identify the targets precisely from the filtered vehicles by matching the license plates, which can reduce the enormous workload. Meanwhile,they will search videos recorded by cameras from near to far positions and from close to distant time range. Therefore,the contextual information such as spatio temporal cues thus can decidedly assist in the search process.