Abstract: Vehicle Classification has wide applications in intelligent transportation and smart cities. Classification is crucial for an intelligent transportation system (ITS). Vehicle make and model classification technique is very useful. Make and model is a fine-grained information that can help officers uncover cases of traffic violations when license plate information cannot be obtained. Faster-RCNN and YOLO are two algorithms which are used for Object Detection and Classification. These algorithms work very well for classifying different types of vehicles. The various vehicles are classified into different vehicle classes. Initially, the dataset is made ready using the videos or converted images format. The dataset needs to contain the data which is of the predefined classes types. Secondly, the images or videos not used during the training are then used for the evaluation of vehicle classifier model. The main parameters that are used in the evaluation of the model include training time, the testing time, the vehicle classification accuracy, as well as the performance of the specific deep learning methods. Many of the datasets are used in the classification of datasets, it makes the evaluation of the classes of datasets more simpler and easier. The previous data could be pre-processed to generate the new datasets that could be used for the evaluation of datasets.
Keywords: Vehicle classification; color classification; deep learning; convolutional neural network; Faster R-CNN, YOLO, VGG-16.
|
DOI:
10.17148/IJARCCE.2023.12324
