Vehicle Classification Using Infrared Image Analysis

Abstract

This study applies image‐analysis techniques to classify vehicles on highways. The procedure of image analysis includes four stages: image acquisition, data processing, feature extraction, and object classification. An infrared image device system is used to record vehicle images on highways. A modified edge‐sharpening technique is derived in this study to filter out noise and to locate edges of bridge clusters simultaneously. A true classifier, including a modified kNN (k nearest neighbor) method, is also established for the classification of vehicles. The time complexity of the proposed algorithm is linear. The accuracy of the algorithm is examined for the classification of a single vehicle per image. The possibility of real‐time analysis for infrared image analysis is also discussed. The results of this study show that the image analysis has potential in the area of automatic traffic‐monitoring systems.