Haoran Zhu
ABSTRACT
Aiming to address the low processing efficiency problem that arises with the massive image data generated in tobacco-related commercial terminals, a deep learning technology-based laser code text recognition solution for commercial purposes is developed. During identification, the YOLO deep learning network is used to initially determine the position of the laser code. Afterwards, the brightness of the light source is adjusted via the light adaptive algorithm. Finally, the laser code results undergo a filtering and interference process. Then, the RANSAC algorithm is utilized for classification, and post-processing is performed, including interaction with the UI interface. The final result is analyzed and displayed. Due to the multiple variations of laser codes, the article suggests utilizing code value adaptive mutation technology to train and distinguish diverse forms of identical characters. This process enhances the model's capacity to learn. Overall, the findings suggest varying levels of recognition among the tested cigarettes. Finally, the study examined various cigarette backgrounds, revealing that 63% of cigarettes had a recognition rate of 90-100%, while 22% had a recognition rate of 80-90%. Additionally, 10% of cigarettes with a recognition rate of 70-80% were accounted for, and cigarettes with a recognition rate of 60-70% represented 5%. The algorithm developed in this study was used in a cigarette laser code recognition system, resulting in a successful recognition outcome.
- Buta M, Neumann L, Matas J 2015 FASText: efficient unconstrained scene text detector Proc. of IEEE International Conference on Computer Vision Piscataway, NJ: IEEE Press pp 1206-1214.4.Google Scholar
- He P, Huang W L, Qiao Y, 2016 Reading scene text in deep convolutional sequences Proc. of the 30th Conference on Artificial Intelligence Palo Alto: AAAI Press pp 3501-3508.Google Scholar
- Huang W L, Qiao Y, Tang X O 2014 Robust scene text detection with convolution neural network induced MSER trees Proc. of the l3th European Conference on Computer Vision Berlin: Springer pp 497-511.4.Google Scholar
- Zhang Z, Zhang C Q, Shen W, 2016 Multi-oriented text detection with fully convolutional networks Proc. of the 29th IEEE Conference on Computer Vision and Pattern Recognition Piscataway, NJ: IEEE Press pp 497-511.Google Scholar
- Zhu Y, Chen J, Liang L, 2021 Fourier contour embedding for arbitrary-shaped text detection Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition 3123-3131.Google Scholar
- Hu R, Dollar P, He K, 2018 Learning to segment every thing Proc. of the 3lst IEEE Conference on Computer Vision and Pattern Recognition IEEE Press pp 4233-4241.4.Google Scholar
- Girshick R, Donahue J, Darrell T, 2014 Rich feature hierarchies for accurate object detection and semantic segmentation Proc. of the27th IEEE Conference on Computer Vision and Pattern Recognition IEEE Press pp 580-587.Google Scholar
- Girshick R 2015 Fast R-CNN Proc. of IEEE International Conference Computer Vision IEEE Press pp 1440-1448.4.Google Scholar
- Ren S Q, He K M, Girshick R, 2017 Faster R-CNN: towards real-time object detection with region proposal networks IEEE Trans on Pattern Analysis & Machine Intelligence. 39(6):1137-1149.Google Scholar
- Dai J F, Li Y, He K M, 2016 R-FCN: object detection via region-based fully convolutional networks Proc. of the 30th International Conference on Neural Information Processing Systems Curran Associates Inc. Press pp 379-387.Google Scholar
- Redmon J, Diyvala S, Girshick R, 2016 You only look once: unified, real time object detection Proc. of the 29th IEEE Conference on Computer Vision and Pattern Recognition IEEE Press pp 779-788.Google Scholar
- Liu W, Anguelov D, Erhan D, 2016 SSD: single shot multi-box detector Proc. of the 14th European Conference on Computer Vision Berlin: Springer pp 21-37.Google Scholar
- Wang K, Babenko B and Belongie S 2012 End-to-end scene text recognition Proc. of IEEE International Conference on Computer Vision IEEE Press pp 1457-1464.Google Scholar
- Bissacco A, Cummins M, Netzer Y, 2013 PhotoOCR: reading text in uncontrolled conditions Proc. of IEEE International Conference on Computer Vision IEEE Press pp 785-792.Google Scholar
- Jaderberg M, SimonyanK, VedaldiA, 2016 Reading text in the wild with convolutional neural networks International Journal of Computer Vision 116(1): 1-20.Google ScholarDigital Library
- Graves A 2009 A novel connectionist system for unconstrained handwriting recognition IEEE Trans on Pattern Analysis and Machine Intelligence 31(5): 855-868.Google Scholar
- Defferrard M, Bresson X, Vandergheynst P 2016 Convolutional neural networks on graphs with fast localized spectral filtering Advances in Neural Information Processing Systems 29Google Scholar
- Liu S, Qi L, Qin H F, 2018 Path aggregation network for instance segmentation Proc. of the IEEE Conference on Computer Vision and Pattern Recognition pp 8759-8768.Google Scholar
- Lin T Y, Dollar P, Girshick R, 2017 Feature pyramid networks for object detection Proc. of the IEEE Conference on Computer Vision and Pattern Recognition pp 936-944.Google Scholar
Index Terms
- Research on Intelligent Recognition Technology of Cigarette Laser Code Based on Deep-Learning
Recommendations
Face Recognition Based on Deep Learning
Human Centered ComputingAbstractAs one of the non-contact biometrics, face representation had been widely used in many circumstances. However conventional methods could no longer satisfy the demand at present, due to its low recognition accuracy and restrictions of many ...
Facial expression recognition via learning deep sparse autoencoders
Facial expression recognition is an important research issue in the pattern recognition field. In this paper, we intend to present a novel framework for facial expression recognition to automatically distinguish the expressions with high accuracy. ...
Research on facial feature-based gender intelligent recognition based on the Adaboost algorithm
In order to overcome the problem of poor facial recognition intelligence and weak gender judgment, a new method based on Adaboost algorithm for facial feature-based gender intelligence recognition is proposed in this paper. In this method, the three-...
Comments