SIFT-BASED MEASUREMENTS FOR VEHICLE MODEL RECOGNITION

A SIFT-based Vehicle Manufacturer and Model Recognition (VMMR) method was utilized to tackle the problem of vehicle security. Distinctive parts of the vehicle frontal view such as the headlights, grill and logo area were segmented. A series of experiments were conducted in a variety of outdoor conditions, where a query image that was rotated, scaled, shifted or set in different lighting conditions, matched against a database of model images. In this work, is shown that image processing functions based on Scale Invariant Feature Transform ( SIFT) measurements can be used to obtain high performance object features recognition, creating a key-point fingerprint ( pattern) for each image class. In the majority of the cases, SIFT method performs very well, in terms of efficiency and robustness.