Vehicle Make and Model Identification Using Vision System
Reconnaissance par Vision du Type d’un Véhicule Automobile
OPEN ACCESS
With the first strategy, the system correctly identifies 80,2% of 830 test samples. The mean of the recognition rates per class is 69,4%. The second strategy obtains better results (in mean, with 100 randomly different repartitions). In the first space, wf, we obtain 93,1% for the correctly identification rate (83,5% for the mean of the recognition rates per class). In a second space, PC A50 wf , we obtain 86,2% for the correctly identification rate (78,8% for the mean of the recognition rates per class). In the last space f, we obtain 90,6% for the correctly identification rate (86,4% for the mean of the recognition rates per class). Another test simulates the presence of a tollgate at four different locations. The better recognition are obtained if the virtual tollgate hides the upper part of the images: a lot of noise points are extracted from this part. These points perturb the recognition system. They are filtered if the number of images used in the model creation is sufficient (> 5). The results presented allow to conlude that the second strategy obtains better results, particularly if the fusion scores are used.
Résumé
Cet article présente un système de reconnaissance du type (constructeur, modèle) de véhicules par vision. À partir d’une vue de face avant d’un véhicule, limitée à sa calandre, nous en construisons une représentation à base de points de contour orientés. La classification est réalisée essentiellement en se fondant sur des algorithmes de votes. L’utilisation d’algorithmes de votes permet au système d’être robuste aux données manquantes ou erronées de la représentation. Nous avons donc construit une fonction de discrimination qui combine 3 votes et une distance, et agit comme une mesure de similarité entre chaque modèle et l’image de véhicule testée. Deux stratégies de décision ont été testées. La première associe à une image de calandre avant du véhicule, le modèle qui a obtenu la valeur la plus importante en sortie de la fonction. Une seconde stratégie regroupe toutes les sorties en un vecteur. La décision est alors prise via un algorithme de plus proche voisin dans un espace dit de votes. Avec la première stratégie, un taux de reconnaissance de 93 % est obtenu sur une base d’images prises en conditions réelles composée de 20 classes de type de véhicules. De plus, une caractérisation et une analyse du fonctionnement du système vis-à-vis de ses différents paramètres est proposée. Cependant ce taux chute à 80 % lorsque le nombre de modèles passe à 50 classes. Pour le même nombre de classes, la seconde stratégie permet d’obtenir un taux supérieur à 90 %.
Pattern Recognition, Vision, Multiclass recognition, Voting Algorithm.
Mots clés
Reconnaissance des formes, vision, classification multi-classes, méthode de votes.
[1] S. BELONGIE, J. MALIK, and J. PUZICHA, Shape matching and object recognition using shape contexts. IEEE Transactions on Pattern Analysis and Machine Intelligence, 24(4):509-522, 2002.
[2] O. CARMICHAEL and M. HEBERT, Object recognition by a cascade of edge probes. In British Machine Vision Conference, volume 1, pages 103-112, Cardiff, UK, September 2002. British Machine Vision Association.
[3] T. COOTES and C. TAYLOR, On representing edge structure for model matching. In Computer Vision and Pattern Recognition, volume 1, pages 1114-1119, Hawai, USA, December 2001.
[4] J. DOURET and R. BENOSMAN, A multi-cameras 3d volumetric method for outdoor scenes: a road traffic monitoring application. In International Conference on Pattern Recognition, pages III: 334-337, 2004.
[5] L. DLAGNEKOV, Video-based car surveillance : License plate make and model recognition. Master Thesis. University of California at San Diego, 2005.
[6] MP. DUBUISSON and A. JAIN, A modified hausdorff distance for object matching. In International Conference on Pattern Recognition, volume A, pages 566-569, 1994.
[7] MP. DUBUISSON-JOLLY, S. LAKSHMANAN, and A. JAIN, Vehicle segmentation and classification using deformable templates. IEEE Transactions on Pattern Analysis and Machine Intelligence, 18(3):293-308, 1996.
[8] J. M. FERRYMAN, A. D. WORRALL, G. D. SULLIVAN, and K. D. BAKER, A generic deformable model for vehicle recognition. In British Machine Vision Conference, pages 127-136, 1995.
[9] D.M. GRAVILA, A pedestrian detection from a moving vehicle. In European Conference on Computer Vision, pages 37-49, 2000.
[10] D. HAN, M.J. LEOTTA, D.B. COOPER, and J.L. MUNDY, Vehicle class recognition from video-based on 3d curve probes. In VS-PETS, pages 285-292, 2005.
[11] D. HOND and L. SPACEK, Distinctive descriptions for face processing. In Machine Vision Conference, University of Essex, UK, 1997.
[12] F. JURIE and C. SCHMID, Scale-invariant shape features for recognition of object categories. In Computer Vision and Pattern Recognition, Washington, DC, June-July 2004.
[13] F. M. KAZEMI, S. SAMADI, H. R. POOREZA and M. R. AKBARZADEH-T, Vehicle Recognition Based on Fourier, Wavelets and Curvelet Transforms – a Comparative study. In IEEE International Conference on Information Technology, INTNG’07, 2007.
[14] P. NEGRI, X. CLADY, M. MILGRAM and R. POULENARD, Système de reconnaissance multiclasses de type de véhicule. In Orasis. Obernai, France, 2007.
[15] P. A. NEGRI, X. CLADY, S. MUHAMMAD HANIF and L. PREVOST, A cascade of boosted generative and discriminative classifiers for vehicle detection. In EURASIP Journal on Advances in Signal Processing (Special Issue : Machine Learning in Image Processing), 12 Pages, 2008.
[16] E. NOWAK and F. JURIE, Learning visual similarity measures for comparing never seen objects. In IEEE Conference on Computer Vision and Pattern Recognition, 2007.
[17] D. G. LOWE, Distinctive image features from scale-invariant keypoints. In International Journal of Computer Vision, 60(2):91-110, 2004.
[18] C. F. OLSON and D. P. HUTTENLOCHER. Automatic target recognition by matching oriented edge pixels. IEEE Transactions on Image Processing, 6(1):103-113, 1997.
[19] V.S. PETROVIC and T.F. COOTES, Analysis of features for rigid structure vehicle type recognition. In British Machine Vision Conference, volume 2, pages 587-596, 2004.
[20] V.S. PETROVIC and T.F. COOTES, Vehicle Type Recognition with Match Refinement. In International Conference on Pattern Recognition, volume 3, pages 95-98, 2004.
[21] D.T. MUNROE and M.G. MADDEN, Multi-Class and Signle-Class Classification Approaches to Vehicle Model Recognition from Images. AICS, 2005.
[22] C. J. Van RIJSBERGEN, Information Retrieval. Butterworths, London, Second Edition, 1979.
[23] Z. SUN, G. BEBIS, and R. MILLER, On-road vehicle detection: A review. IEEE Transactions on Pattern Analysis and Machine Intelligence, 28(5):694-711, May 2006.
[24] D.A. TORRES, More Local Structure for Make-Model Recognition. Technical Rapport, University of California, San Diego, 2007.
[25] I. ZAFAR, E.A. EDIRISINGHE, B.S. ACAR and H.E. BEZ, Two Dimensional Statistical Linear Discriminant Analysis for Real-TimeRobust Vehicle Type Recognition. SPIE, vol.6496, 2007.
[26] I. ZAFAR, B.S. ACAR and E.A EDIRISINGHE, Vehicle Make and Model Identification using Scale Invariant Transforms. IASTED, 2007.
[27] W. ZHAO, R. CHELLAPPA, P. J. PHILLIPS, and A. ROSENFELD, Face recognition: A literature survey. ACM Computing Surveys, 35(4):399-458, 2003.