Semiautomatic training and evaluation of a learning-based vehicle make and model recognition system
| Autor: | Peter H. N. de With, Rob G. J. Wijnhoven, Guido M. Y. E. Brouwers, Matthijs H. Zwemer |
|---|---|
| Přispěvatelé: | Video Coding & Architectures |
| Rok vydání: | 2018 |
| Předmět: |
Computer science
vehicle make and model ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION occlusion 02 engineering and technology Convolutional neural network Data modeling 020204 information systems Histogram 0202 electrical engineering electronic engineering information engineering Computer vision Electrical and Electronic Engineering traffic Contextual image classification business.industry Detector convolutional neural network classification Statistical model Atomic and Molecular Physics and Optics Computer Science Applications Visualization ComputingMethodologies_PATTERNRECOGNITION dataset generation 020201 artificial intelligence & image processing Artificial intelligence business Classifier (UML) |
| Zdroj: | Journal of Electronic Imaging, 27(5):051225. SPIE |
| ISSN: | 1017-9909 |
| DOI: | 10.1117/1.jei.27.5.051225 |
| Popis: | We describe a system for vehicle make and model recognition (MMR) that automatically detects and classifies the make and model of a car from a live camera mounted above the highway. Vehicles are detected using a histogram of oriented gradient detector and then classified by a convolutional neural network (CNN) incorporating the frontal view of the car. We propose a semiautomatic data-selection approach for the vehicle detector and the classifier, by using an automatic number plate recognition engine to minimize human effort. The resulting classification has a top-1 accuracy of 97.3% for 500 vehicle models. This paper presents a more extensive in-depth evaluation. We evaluate the effect of occlusion and have found that the most informative vehicle region is the grill at the front. Recognition remains accurate when the left or right part of vehicles is occluded. The small fraction of misclassifications mainly originates from errors in the dataset, or from insufficient visual information for specific vehicle models. Comparison of state-of-The-Art CNN architectures shows similar performance for the MMR problem, supporting our findings that the classification performance is dominated by the dataset quality. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|