Abstract
This work focuses on training difficulty and high computational cost caused by high-resolution input image when car plate character recognition based on support vector machine is applied. The use of singular value decomposition technology is proposed with improved optimal threshold analysis strategy to select retained features. The original data set is mapped in the low-dimensional feature space and the model is trained. The same test set is used to evaluate the models trained under different feature dimensions, and the optimal feature dimension is determined to 32 dimensional features. Results showed that the algorithm can accurately classify and detect license plate characters. The proposed model achieves 99.69% accuracy of the original data set, demonstrates excellent advantages, and remarkably simplifies the model design when only 96 dimensional features are used compared with traditional linear and convolutional neural network classifiers. New methods reduce the algorithm run time by 200%, and the number of parameter storage required by the algorithm to 25%. Hence, the proposed method can significantly improve the license plate recognition system usage in resource-constrained embedded devices.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
Chen, S.-L., et al.: Simultaneous end-to-end vehicle and license plate detection with multi-branch attention neural network. IEEE Trans. Intell. Transp. Syst. 21(9), 3686–3695 (2019)
Wang, D., et al.: LPR-Net: recognizing Chinese license plate in complex environments. Pattern Recogn. Lett. 130, 148–156 (2020)
Chenyang, H.: Research on the Method of License Plate Recognition Based on Image Processing (2018)
Izidio, D.M., et al.: An embedded automatic license plate recognition system using deep learning. Design Automation for Embedded Systems, 1–21 (2019)
Sutaji, D., Husenti, N.: Digital image processing for character detection of captcha login internet banking image using matching template. J. Phys. Conf. Ser. (2019)
Wenfeng, Y., et al.: Research on vehicle license plate character segmentation and recognition technology. Mod. Electron. Technol. 43(19), 65–69 (2020)
Meng Xianghuan, R.S., Yuzu, Z., Ya, C., Sitao, C.: License plate character recognition method based on TensorFlow. J. Shanghai Univ. Eng. Sci. 34(03), 42–47 (2020)
Lei, X., Pan, H., Huang, X.: A dilated CNN model for image classification. IEEE Access PP(99), 1–1 (2019)
Niu, X.X., Suen, C.Y.: A novel hybrid CNN–SVM classifier for recognizing handwritten digits. Pattern Recogn. 45(4), 1318–1325 (2012)
Peng, Z.: Research on license plate recognition system based on SVM (2018)
Zeng, J., et al.: Generalization ability of online pairwise support vector machine. J. Math. Anal. Appl. 497(2), 124914 (2021)
Peng, S., et al.: Robust semi-supervised nonnegative matrix factorization for image clustering. Pattern Recogn. 111, 107683 (2021)
Xu, Z. et al.: Towards end-to-end license plate detection and recognition: a large dataset and baseline. In: Proceedings of the European conference on computer vision (ECCV) (2018)
Yamashita, Y., Wakahara, T.: k-NN classification of handwritten characters using a new distortion-tolerant matching measure. In: International conference on pattern recognition (2014)
Jiajia, Z., et al.: Application of SVD and SVM superposition algorithm in image recognition of glass thermometer. Electron. Meas. Technol. 38(1), 47–50 (2015)
Lok, U.-W., et al.: Real time SVD-based clutter filtering using randomized singular value decomposition and spatial downsampling for micro-vessel imaging on a Verasonics ultrasound system. Ultrasonics. 107, 106163(2020)
Atif, S.M., Qazi, S., Gillis, N.: Improved SVD-based initialization for nonnegative matrix factorization using low-rank correction. Pattern Recogn. Lett. 122, 53–59 (2019)
Chan, T.H., et al.: PCANet: a simple deep learning baseline for image classification? IEEE Trans. Image Process. 24(12), 5017–5032 (2015)
Chunikhina, E., Raich, R., Nguyen, T.: Performance analysis for matrix completion via iterative hard-thresholded SVD. In Statistical Signal Processing (2014)
Gavish, M., Donoho, D.L.: The optimal hard threshold for singular values is 4/sqrt(3). IEEE Trans Inf Theory 60(8), 5040–5053 (2014)
Donoho, D.L.: G.M. Code supplement to "the optimal hard threshold for singular values is 4/sqrt(3)". https://purl.stanford.edu/vg705qn9070 (2014)
Bondarev, N.: Artificial neural network and multiple linear regression for prediction and classification of sustainability of sodium and potassium coronates. Russ. J. Gen. Chem. 89(7), 1438–1446 (2019)
Mu, Y., et al.: Weighted tensor nuclear norm minimization for tensor completion using tensor-SVD. Pattern Recogn. Lett. 130, 4–11 (2020)
Wang, A., Jin, Z., Tang, G.: Robust tensor decomposition via t-SVD: near-optimal statistical guarantee and scalable algorithms. Signal Process. 167, 107319 (2020)
Funding
Funding was provided by National vocational education teachers innovation team project of China (CN) (Grant No. YB2020020103), National Natural Science Foundation of China (CN) (Grant Nos. 61701267, 61971251) and Postdoctoral Foundation of China (CN) (Grant No. 2019M663474).
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Zhou, K., Ge, Q., Wei, C. et al. Character classification algorithm based on the low-dimensional feature-optimized model. SIViP 16, 543–550 (2022). https://doi.org/10.1007/s11760-021-01997-0
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11760-021-01997-0