Vehicle logo recognition using histograms of oriented gradient descriptor and sparsity score

Abstract

Vehicle logo images are first characterized by Histograms of Oriented Gradient descriptors and the final features vector are then applied feature selection method to reduce the irrelevant information. [...]we release a new benchmark dataset for vehicle logo recognition and retrieval task namely, VLR-40. Traditional vehicle recognition systems identify vehicle based on manual human observations via license plate or model of vehicles. [...]automatic vehicle identification is a key problem in intelligent transportation system. Examples of feature extraction methods include principal component analysis (PCA) [11], locality preserving projections (LPP) [12]. Recently, feature selection has gained increasing interest in the field of machine learning [13-16], data analysis [17-19], and successfully applied in computer vision such as information retrieval [20-22] or visual object tracking [23-25].

Headnote

Article Info

Article history:

Received Mar 24, 2020

Revised May 24, 2020

Accepted Jun 25, 2020

Keywords: Feature selection HOG descriptor Image classification Sparsity score Vehicle logo recognition

ABSTRACT

Most of vehicle have the similar structures and designs. It is extremely complicated and difficult to identify and classify vehicle brands based on their structure and shape. As we require a quick and reliable response, so vehicle logos are an alternative method of determining the type of a vehicle. In this paper, we propose a method for vehicle logo recognition based on feature selection method in a hybrid way. Vehicle logo images are first characterized by Histograms of Oriented Gradient descriptors and the final features vector are then applied feature selection method to reduce the irrelevant information. Moreover, we release a new benchmark dataset for vehicle logo recognition and retrieval task namely, VLR-40. The experimental results are evaluated on this database which show the efficiency of the proposed approach.

This is an open access article under the CC BY-SA license.

1. INTRODUCTION

How to identify a brand and distinguish with the others visually? Each brand has its own logo that represent a trademark, contain a certain meaning symbolizing that brand and its manufacturer. To create an impressive logo, specific characteristics need to pay attention to many details including: layout, colors, lines, angles, and all information must be arranged in coherent and harmony way. Traditional vehicle recognition systems identify vehicle based on manual human observations via license plate or model of vehicles. Thus, automatic vehicle identification is a key problem in intelligent transportation system. Each vehicle has a unique license plate, but it is difficult to track and identify since these images are in very low quality in in smart surveillance systems.

Various works have been proposed for vehicle logo recognition in the past. We briefly review several works in this field. For example, Llorca et al. [1] apply histograms of oriented gradients and support vector machines (SVM) classifier for vehicle manufacturer recognition. A region of interest is applied before extracting histograms of oriented gradient (HOG) features from logo images. Huang et al. [2] propose a system for car logo segmentation and recognition based on an efficient pre-trained convolutional neural network (CNN) model. Huan et al. [3] present a new algorithm based on Hough transform and Deep Learning is for vehicle logo retrieval task which combine shape detection and deep belief networks. Pan et al. [4] use scale invariant feature transform (SIFT) descriptor and CNN as the feature extraction method from vehicle logo images. An enhanced logo-recognition system is presented by Psyllos et al. [5] on the Medialab license plate recognition (LPR) dataset. Huang et al. [6] apply Faster-RCNN model with two different CNNs (VGG-16 and ResNet-50) for vehicle logo recognition. Sotheeswaran and Ramanan [7] present a study focuses on local features that describe structural characteristics of the logo of a car using a coarse-to-fine strategy. Nie et al. [8] present a new VLR method based on foreground-background pixel-pair feature. Different hand-crafted descriptors (HOG, LBP and SIFT) are applied to extract features. Cyganek and M. Wo zniak [9] use ensemble classifiers based on higher-order singular value decomposition to classify vehicle logo. More recently, Zhao and Wang [10] introduce a modified version of HU invariant moment to represent car logo images.

Indeed, many machine learning problems in computer vision and several related domains need to deal with very high dimensional data. Many of these features may not be relevant for the final prediction task and degrade the classification performance. Multiple studies have shown that the classification performance can be improved by eliminating these features. These issues can be solved by the method of the dimensionality reduction. For this purpose, the dimensionality reduction can be achieved either by feature extraction or feature selection to a low dimensional space. Feature extraction refers to the methods that create a set of new features based on the linear or non-linear combinations of the original features. Further analysis is problematic since we cannot get the physical meanings of these features in the transformed space. Examples of feature extraction methods include principal component analysis (PCA) [11], locality preserving projections (LPP) [12].

In contrast, the feature selection methods aim at finding adequate subsets of features by keeping some original features and therefore maintains the physical meanings of the features. The use of both methods has the advantage of improving performance of classification and increasing computational efficiency. Recently, feature selection has gained increasing interest in the field of machine learning [13-16], data analysis [17-19], and successfully applied in computer vision such as information retrieval [20-22] or visual object tracking [23-25]. In this work, we focus on the application of feature selection methods to vehicle logo images classification by sparsity score. This paper is organized and structured as follows. Section 2 introduces the feature extracting methods based on three local image descriptors. Section 2 and 3 present proposed approach and experimental results. Finally, the conclusion is discussed in section 4.

2. THE FEATURE EXTRACTION AND SELECTION

2.1. Histograms of oriented gradient descriptor

Histograms of oriented gradient (HOG) descriptor is applied for different problems in machine vision [26-32]. HOG feature is extracted by counting the occurrences of gradient orientation base on the gradient angle and the gradient magnitude of local patches of an image. The gradient angle and magnitude at each pixel are computed in an 8 x 8 pixels patch. Next, 64 gradient feature vectors are divided into 9 angular bins 0-180° (20°each). The gradient magnitude T and angle K at each position (k, h) from an image J are computed as follows:

2.2. Feature selection

Based on the availability of supervised information (i.e. labels), feature selection techniques can be grouped into two large categories: supervised and unsupervised context [33]. Additionally, different strategies of feature selection are proposed based on evaluation process such as filter, wrapper, and hybrid methods [34]. Hybrid approaches incorporate both filter and wrapper into a single structure, to give an effective solution for dimensionality reduction [35].

Liu et al. extend the unsupervised sparsity score to supervised context by utilizing the class label information [36, 37]. Let f^sub ri^^sup c^ denotes the rthfeature of ith instance in class c, Sy is the element of sparse similarity matrix Sewhich is constructed within the class c, ec is a N- dimensional vector with ee=l, if It belongs to the class c and 0 otherwise. The two proposed supervised sparsity score of the rth feature, denoted SparseScorer, which should be minimized, are defined as follows:

After calculating the score for each feature, they are sorted in the ascending order of SparseScore? to select the relevant ones. In the classification experiments, Liu et al. have demonstrated that this score outperforms other methods in most cases, especially for multi-class problems [36].

3. EXPERIMENTAL RESULTS

3.1. Experimental setup

Despite the vehicle logo recognition problem has been studied for many years, a few publicly available is available for the computer vision community. There are a few datasets is applied for logo detection such as vehicle logo dataset with 30 classes, namely VLD-30 [38]. Table 1 analyses the existing vehicle logo datasets. The first column indicates the dataset name and its reference. The second column show the availability of the predefined training and testing set. This information is important for compare the classification results because other researchers can use this decomposition to train and test their models instead of using cross-validation methods. The next columns represent the number of logos, total images, and its resolution. The last column shows the availability of the corresponding dataset.

The drawback of these datasets is the lack of pre-defined decomposition for training and testing set and mages are resized to the same resolution. To this end, we collected and organized a large-scale and comprehensive image database called VLR-40. This dataset contains images that were taken by different users, in unconstrained condition. The data was gathered by crawling from web pages and cropped semi-automatically. We keep the original resolution from cropped logo. It can be downloaded publicly at: https://data.mendeley.com/datasets/dr233ns3g6/3 which contains total 4,000 color images of vehicle logo (see several example images from this dataset in 1). The sizes of each image are totally different. There is a total of 40 classes of logo in this original dataset. The data is split in half to be used as training and testing sets for classification task. Vehicle logo images are cropped from photo of car images semi-automatically. They also contain background and foreground and make this dataset is more challenging than the others. Figure 1 illustrates the two-vehicle logos are cropped from an original image.

Figure 2 illustrates the cropped images from original image crawled from the web. We can observe that the two cropped images of the Ferrari logo are different visually by the resolution, rotation, and color. Under the current technology condition, these problems make this dataset is more challenging than others. Table 2 presents example images of 40 different logo from VLR-40 dataset.

3.2. Results

There are many color spaces are proposed in the literature for different applications. The HOG descriptor is applied to extract features from vehicle logo image of each color component. The final feature is obtained by fusing all features from three components. Three color spaces (RGB, HSV, YCbCr) are considered to encode image since these spaces are widely used for pattern recognition application. The training set is used to compute sparse score by (5) and (6). The value of these scores are then applied to rank features of training and testing set. Here, we use the cut-off ratio is 1% number of features to determine the optimum dimension. Table 2 presents the classification results on the VLR-40 dataset. The first column indicates the color space used to encode vehicle logo images. The second column shows the accuracy achieved of each space and its dimension when no selection method is applied. The number of features is 11,532x3 = 34,596 features. We see that the accuracy varies on different color space. The second and third column present the classification results by using sparse score 1 and sparse score 2, respectively. The sparse score 1 clearly outperforms other methods by giving the best accuracy (75.25%) by using 77% (26,638 features) number of features. The sparse score 2 give the accuracy close to the results when no selection method is applied. However, it largely reduces number of features comparing with sparse score 1. For example, sparse score 2 only uses 20% number of features on HSV space while giving better performance. By observing this table, we see that feature selection method gives the accuracy as performing when no selection method, but it allows to reduce the dimension space.

Additionally, Figure 3 compares the performance of two sparse score 1 and 2 on three different color spaces. The combination of HSV color space and sparse core 1 give the worst performance compared with other methods. The RGB space and sparse score 2 give a good performance at early stage since it only need fewer than 10% number of features to reach an accuracy more than 70%. In contrast, the YCbCr and HSV spaces combined with sparse score 1 achieve a very low accuracy at the beginning when number of selected features is fewer than 55%. So, experimental results show that it should be interesting to find a suitable color space to encode vehicle logo images and an appropriate feature selection method to remove irrelevant features.

4. CONCLUSION

This paper presents a vehicle logo recognition based on HOG descriptor and feature selection via two sparse score. We also release a new benchmark vehicle logo image (VLR-40) dataset for research community. The experimental results show that the sparse score 1 gives the best accuracy on the RGB color space and largely reduce number of features. This study is now extended to compare the performance and find a suitable color space for encoding vehicle logo images.

ACKNOWLEDGEMENTS

This work was supported by Suan Sunandha Rajabhat University, Thailand.

References

REFERENCES

[1] D. F. Llorca, R. Arroyo, and M. A. Sotelo, "Vehicle logo recognition in traffic images using HOG features and SVM," In 16th International IEEE Conference on Intelligent Transportation Systems (ITSC 2013), pp. 2229-2234, 2013.

[2] Yue Huang, Ruiwen Wu, Ye Sun, Wei Wang, and Xinghao Ding, "Vehicle Logo Recognition System Based on Convolutional Neural Networks with a Pretraining Strategy," IEEE Transactions on Intelligent Transportation Systems, vol. 16, no. 4, pp. 1951-1960, 2015.

[3] Li Huan, Wang Li, and Qin Yujian, "Vehicle Logo Retrieval Based on Hough Transform and Deep Learning," International Conference on Computer Vision Workshops (ICCVW), pp. 967-973, 2017.

[4] Chun Pan, Mingxia Sun, Zhiguo Yan, Jie Shao, Xiaoming Xu, and Di Wu, "Vehicle logo recognition based on deep learning architecture in video surveillance for intelligent traffic system," IET International Conference on Smart and Sustainable City 2013 (ICSSC 2013), pp. 132-135, 2013.

[5] Apostolos P. Psyllos, Christos-Nikolaos E. Anagnostopoulos, and Eleftherios Kayafas, "Vehicle Logo Recognition Using a SIFT-Based Enhanced Matching Scheme," IEEE Transactions on Intelligent Transportation Systems, vol. 11, no. 2, pp. 322-328, 2010.

[6] Z. Huang, M. Fu, K. Ni, H. Sun, and S. Sun, "Recognition of Vehicle Logo Based on Faster-RCNN," Songlin Sun, editor, Signal and Information Processing, Networking and Computers, vol. 494, pp. 75-83, 2019.

[7] S. Sotheeswaran, A. Ramanan, "A Coarse-to-Fine Strategy for Vehicle Logo Recognition from Frontal-View Car Images," Pattern Recognition and Image Analysis, vol. 28, no. 1, pp. 142-154, 2018.

[8] Z. Nie, Y. Yu, and Q. Jin, "A Vehicle Logo Recognition Approach Based on Foreground-Background Pixel-Pair Feature," Transactions on Edutainment XIII, pp. 204-214, 2017.

[9] B. Cyganek and M. Wo zniak, "Vehicle Logo Recognition with an Ensemble of Classifiers," Bootstrapping and Rule-Based Model for Recognizing Vietnamese Named Entity," pp.117-126, 2014.

[10] J. Zhao and X. Wang, "Vehicle-logo recognition based on modified HU invariant moments and SVM," Multimedia Tools and Applications, vol. 78, no. 1, pp. 75-97, 2019.

[11] K. Fukunaga, "Introduction to statistical pattern recognition. Computer science and scientific computing," Academic Press, Boston, 2nd ed edition, 1990.

[12] X. He and P. Niyogi, "Locality preserving projections," Advances in neural information processing systems, pp. 153-160, 2004.

[13] S. K. Shevade and S. S. Keerthi, "A simple and efficient algorithm for gene selection using sparse logistic regression," Bioinformatics, vol. 19, no. 17, pp. 2246-2253, 2003.

[14] T. Li, C. Zhang, and M. Ogihara, "A comparative study of feature selection and multiclass classification methods for tissue classification based on gene expression," Bioinformatics, vol. 20, no. 15, pp. 2429-2437, 2004.

[15] M. Yang, F. Wang, and P. Yang, "A novel feature selection algorithm based on hypothesis margin," Journal of Computers, vol. 3, no. 12, pp. 27-34, 2008.

[16] W. Megchelenbrik, "Relief-Based feature selection in bioinformatics: detecting functional specificity residues from multiple sequence alignments," Radboud University, Nijmegen, 2010.

[17] M. Dash and H. Liu, "Feature selection for classification," Intelligent data analysis, vol. 1, no. 3, pp. 131-156, 1997.

[18] Y. Sun, S. Todorovic, and S. Goodison, "Local-learning-based feature selection for high-dimensional data analysis," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 32, no. 9, pp. 1610-1626, 2010.

[19] D. Dernoncourt, B. Hanczar, and J. D. Zucker, "Analysis of feature selection stability on high dimension and small sample data," Computational Statistics & Data Analysis, vol. 71, pp. 681-693, 2014.

[20] G. Roffo, C. Segalin, A. Vinciarelli, V. Murino, and M. Cristani, "Reading between the turns: statistical modeling for identity recognition and verification in chats," Proceedings of the 10th IEEE International Conference on Advanced Video and Signal Based Surveillance, pp. 99-104, 2013.

[21] G. Roffo, C. Giorgetta, R. Ferrario, W. Riviera, and M. Cristani, "Statistical analysis of personality and identity in chats using a keylogging platform," Proceedings of the 16th International Conference on Multimodal Interaction, pp. 224-231, 2014.

[22] G. Roffo, M. Cristani, L. Bazzani, H. Q. Minh, and V. Murino, "Trusting skype: Learning the Way People Chat for Fast User Recognition and Verification," Proceedings of the IEEE International Conference on Computer Vision Workshops, pp. 748-754, 2013.

[23] J. Yuan and F. B. Bastani, "Robust object tracking via online informative feature selection," Proceedings of the IEEE International Conference on Image Processing, pp. 471-475, 2014.

[24] K. Zhang, L. Zhang, and M. H. Yang, "Real-Time Object Tracking Via Online Discriminative Feature Selection," IEEE Transactions on Image Processing, vol. 22, no. 12, pp. 4664-4677, 2013.

[25] O. Deniz, G. Bueno, J. Salido, and F. De la Torre, "Face recognition using histograms of oriented gradients," Pattern Recognition Letters, vol. 32, no. 12, pp. 1598-1603, 2011.

[26] D. P. V. Hoai, T. Surinwarangkoon, V. T. Hoang, H. T. Duong, and K. Meethongjan, "A comparative study of rice variety classification based on deep learning and hand-crafted features," ECTI Transactions on Computer and Information Technology (ECTI-CIT), vol. 14, no. 1, pp. 1-10, 2020.

[27] H. T. M. Nhat and V. T. Hoang, "Feature fusion by using LBP, HOG, GIST descriptors and Canonical Correlation Analysis for face recognition," 26th International Conference on Telecommunications (ICT), pp. 371-375, 2019.

[28] T. N. Van and V. T. Hoang, "Kinship Verification based on Local Binary Pattern features coding in different color space," 26th International Conference on Telecommunications (ICT), pp. 376-380, 2019.

[29] K. Meethongjan, M. Dzulkifli, P. K. Ree, and M. Y. Nam, "Fusion affine moment invariants and wavelet packet features selection for face verification," Journal of Theoretical and Applied Information Technology, vol. 64, no. 3, pp. 606-615, 2014.

[30] T. Surinwarangkoon, S. Nitsuwat, and J. Elvin, "A traffic sign detection and recognition system," International Journal of Circuits, Systems and Signal Processing, vol. 7, no. 1, pp. 58-65, 2013.

[31] N. Dalal and B. Triggs, "Histograms of oriented gradients for human detection, "2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05)," vol. 1, pp. 886-893, 2005.

[32] K. Benabdeslem and M. Hindawi, "Constrained Laplacian score for semi-supervised feature selection," Machine Learning and Knowledge Discovery in Databases, pp. 204-218, 2011.

[33] I. Guyon and A. Elisseeff, "An introduction to variable and feature selection," The Journal of Machine Learning Research, vol. 3, no. 1157-1182, 2003.

[34] J. Cai, J. Luo, S. Wang, and S. Yang, "Feature selection in machine learning: A new perspective," Neurocomputing, vol. 300, pp. 70-9, 2018.

[35] M. Liu and D. Zhang, "Sparsity score: a novel graph-preserving feature selection method," International Journal of Pattern Recognition and Artificial Intelligence, vol. 28, no. 04, 2014.

[36] C. Zu, L. Zhu, and D. Zhang, "Iterative sparsity score for feature selection and its extension for multimodal data," Neurocomputing, vol. 259, pp. 146, 53, 2017.

[37] S. Yang, J. Zhang, C. Bo, M. Wang, and L. Chen, "Fast vehicle logo detection in complex scenes," Optics & Laser Technology, vol. 110, pp. 196-201, 2018.

[38] R. Liu, Q. Han, W. Min, L. Zhou, and J. Xu, "Vehicle Logo Recognition Based on Enhanced Matching for Small Objects, Constrained Region and SSFPD Network," Sensors, vol. 19, no. 20, pp. 1-18, 2019.

[39] F. Tafazzoli, H. Frigui, and K. Nishiyama, "A Large and Diverse Dataset for Improved Vehicle Make and Model Recognition," 2017 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), pp. 874-881, 2017.

[40] Y. Yu, J. Wang, J. Lu, Y. Xie, and Z. Nie, "Vehicle logo recognition based on overlapping enhanced patterns of oriented edge magnitudes," Computers & Electrical Engineering, vol. 71, pp. 273-283, 2018.

[41] Y. Xia, J. Feng, and B. Zhang, "Vehicle Logo Recognition and attributes prediction by multi-task learning with CNN," 2016 12th International Conference on Natural Computation, Fuzzy Systems and Knowledge Discovery (ICNC-FSKD), pp. 668-672, 2016.

[42] Xia Y, Feng J, Zhang B. Vehicle logo recognition and attributes prediction by multi-task learning with CNN. Proceedings of 12th IEEE international conference on natural computation, fuzzy systems and knowledge discovery (ICNC-FSKD). 2016. p. 668-72.

[43] V. T. Hoang, "HGM-4: A new multi-cameras dataset for hand gesture recognition," vol. 30, 2020.

BIOGRAPHIES OF AUTHORS

Kittikhun Meethongjan is a full lecturer in the Computer Science Program, Head of Apply Science Department, Faculty of Science and Technology, Suan Sunandha Rajabhat University (SSRU), Thailand. He received his B.Sc. Degree in Computer Science from SSRU in 1990, a master's degree in computer science from King Mongkut's University of Technology Thonburi (KMUTT) in 2000 and a PhD degree in Computer Graphic from the University of Technology Malaysia in 2013. His research interest is Computer Graphic, Image Processing, Artificial Intelligent, Biometrics, Pattern Recognition, Soft Computing techniques and application. Currently, he is an advising of Master and PhD student in Forensic Science of SSRU.

Thongchai Surinwarangkoon was born in Suratthani, Thailand in 1972. He obtained B.Sc. degree in mathematics from Chiang Mai University, Thailand in 1995. He received M.Sc. degree in management of information technology from Walailak University, Thailand in 2005 and Ph.D. in information technology from King Mongkut's University of Technology North Bangkok, Thailand in 2014. He is now a Lecturer in Department of Business Computer, Suan Sunandha Rajabhat University, Bangkok, Thailand. His research interests include digital image processing, machine learning, artificial intelligence, business intelligence, and mobile application development for business.

Vinh Truong Hoang received his master's degree from the University of Montpellier in 2009 and his PhD degree in computer science from the University of the Littoral Opal Coast, France. He is currently an assistant professor and Head of Image Processing and Computer Graphics Department at the Ho Chi Minh City Open University, Vietnam. His research interests include image analysis and feature selection.