Vehicle Color Recognition Based on License Plate Color

Abstract

As a significant feature of vehicle, the color feature plays an important role in the intelligent transportation systems. However, the color feature is easily affected by the variations of the lighting condition. In this paper, we present a new method for vehicle color recognition, which is based on license plate color. The color of license plate is recognized by the prior knowledge and the recognition result of the license plate, which is not sensitive to the variations of lighting condition. We select the vehicle ROI (region of interest) near the license plate, and convert the color space of the plate image and the vehicle ROI image from RGB to HSV. The vehicle color is identified by the relative location of the ROI color and the plate color in the spectrum. We verify the feasibility of our approach through a comparison experiment. Experiment results show that the color of license plate is helpful to recognize the vehicle color.

Improved license plate localisation algorithm based on morphological operations

“…A drawback of HLS is that it is compromised with environmental conditions, such as an image being taken in the dark or a reflection on a license plate. HSV methods are better able to deal with the problem of illumination conditions and identifies the colour features of a license plate even when the letter is inclined and deformed [29], [30]. When the colour information is extracted by the license plate localization system, the average accuracy rate is at about 75%, a value which is improved with comparison to the 69% accuracy when no plate information is used.…”

“…A drawback of HLS is that it is compromised with environmental conditions, such as an image being taken in the dark or a reflection on a number plate. HSV methods are better able to deal with the problem of illumination conditions and identifies the colour features of a license plate even when the letter is inclined and deformed [20]. Similar to other colour methods, HSV is limited by environmental conditions and also requires powerful processing systems.…”

Vehicle License Plate Detection for Parking Offenders Using Automatic License-Plate Recognition

“…In this study, ANPR was applied to recognize the license plates of parked vehicles; this system already exists in several countries. Many studies have been implemented to evaluate this new monitoring system, such as for vehicle detection in parking areas [1] and machine vision for moving object detection [2], [3], [4], vehicle license-plate detection [5], [6], [7], face recognition to open or close a parking gate [8], smart parking [9], and management and monitoring of parking lots [10]; however, few studies have discussed restricted parking zones [11], [12]. The ANPR system for license plate recognition has not yet been used for vehicles in Indonesia, so the license plate data used is standardized for foreign plates.…”

License Plate Location Using Feedback Control Technology in Complex Environment

“…Therefore, it is theoretically and economically to establish, improve and develop an efficient LPR system. [7][8][9] is using the properties of color feature, mainly using the color edge or the gray edge with the color features to locate the license plate. The third method [10] is machine learning where the characteristics and excellent training methods are needed.…”