摘要
卷烟条码是烟草局对卷烟是否串货销售的主要判断依据,针对当前人工录码方式操作烦琐、效率低、成本高的问题,提出一种基于深度神经网络的烟码智能识别方法.首先通过迁移学习技术构建区域检测模型,实现对烟码区域的准确定位;然后采用基于角点检测的切割算法将烟码区域切分为待识别的小块;再构建字符识别模型,对小块进行多字符识别;最后按顺序拼接各小块的识别结果输出完整烟码.实验结果表明,该方法准确率高、运行速度快,能够替代人工录码方式,满足实际应用需求.
Cigarette identification code is the basis of discrimination of illegal retailing for tobacco boards, yet it's artificial transcription was quite costly and inefficient. In this paper, we proposed a high-efficient and accurate cigar-code identification method based on Deep Neural Network(DNN). First, it utilized Transfer Learning technology for constructing regional detection model to locate the cigar-code region precisely. Then, it divided the region into small blocks by a cutting algorithm based on Corner Detection. Afterwards, it constructed a character recognition model for multi-character recognition of the small blocks. At last, it reordered the recognition results to achieve a full cigar-code. Results show that our DNN-based cigar-code identification method achieves high accuracy and is far more efficient than artificial transcription, which meets the practical application requirements.
引文
[1]Wang Q Q,Lu Y,Sun S L.Text detection in nature scene images using two-stage nontext filtering[C]//Proceedings of the13th International Conference on Document Analysis and Recognition.Los Alamitos:IEEE Computer Society Press,2015:106-110
[2]Gouk H G R,Blake A M.Fast sliding window classification with convolutional neural networks[C]//Proceedings of the29th International Conference on Image and Vision Computing.New York:ACM Press,2014:114-118
[3]Ye Q X,Doermann D.Text detection and recognition in imagery:a survey[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2015,37(7):1480-1500
[4]Sedaghat A,Mokhtarzade M,Ebadi H.Uniform robust scale-invariant feature matching for optical remote sensing images[J].IEEE Transactions on Geoscience and Remote Sensing,2011,49(11):4516-4527
[5]Bay H,Tuytelaars T,van Gool L.SURF:speeded up robust features[C]//Proceedings of the 9th European Conference on Computer Vision.Heidelberg:Springer,2006,PartⅠ:404-417
[6]Zhu Q,Yeh M C,Cheng K T,et al.Fast human detection using a cascade of histograms of oriented gradients[C]//Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition.Los Alamitos:IEEE Computer Society Press,2006,2:1491-1498
[7]Yoo H J.Deep convolution neural networks in computer vision[J].IEIE Transactions on Smart Processing and Computing,2015,4(1):35-43
[8]Li Zongmin,Zhou Chenchen,Gong Yanhe,et al.Saliency object detection based on domain transform and contour detection[J].Journal of Computer-Aided Design&Computer Graphics,2018,30(8):1457-1465(in Chinese)(李宗民,周晨晨,宫延河,等.结合域变换和轮廓检测的显著性目标检测[J].计算机辅助设计与图形学学报,2018,30(8):1457-1465)
[9]Girshick R,Donahue J,Darrell T,et al.Rich feature hierarchies for accurate object detection and semantic segmentation[C]//Proceedings of the IEEE International Conference on Computer Vision and Pattern Recognition.Los Alamitos:IEEEComputer Society Press,2014:580-587
[10]Girshick R.Fast R-CNN[C]//Proceedings of the IEEE International Conference on Computer Vision.Los Alamitos:IEEEComputer Society Press,2015:1440-1448
[11]He K M,Zhang X Y,Ren S Q,et al.Spatial pyramid pooling in deep convolutional networks for visual recognition[J].IEEETransactions on Pattern Analysis and Machine Intelligence,2015,37(9):1904-1916
[12]Ren S Q,He K M,Girshick R,et al.Faster R-CNN:towards real-time object detection with region proposal networks[C]//Proceedings of the 28th International Conference on Neural Information Processing Systems.Cambridge:MIT Press,2015:91-99
[13]Zhang Shun,Gong Yihong,Wang Jinjun.The development of deep convolution neural network and its applications on computer vision[J].Chinese Journal of Computers,2017,40(144):1-30(in Chinese)(张顺,龚怡宏,王进军.深度卷积神经网络的发展及其在计算机视觉领域的应用[J].计算机学报,2017,40(144):1-30)
[14]Lecun Y,Bottou L,Bengio Y,et al.Gradient-based learning applied to document recognition[J].Proceedings of the IEEE,1998,86(11):2278-2324
[15]Sermanet P,Chintala S,LeCun Y.Convolutional neural networks applied to house numbers digit classification[C]//Proceedings of the 21st International Conference on Pattern Recognition.Los Alamitos:IEEE Computer Society Press,2012:3288-3291
[16]Krizhevsky A,Sutskever I,Hinton G E.ImageNet classification with deep convolutional neural networks[C]//Proceedings of the 25th International Conference on Neural Information Processing Systems.Cambridge:MIT Press,2012:1097-1105
[17]Simonyan K,Zisserman A.Very deep convolutional networks for large-scale image recognition[OL].[2018-06-14].https://arxiv.org/abs/1409.1556
[18]Szegedy C,Liu W,Jia Y,et al.Going deeper with convolutions[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Los Alamitos:IEEE Computer Society Press,2015:1-9
[19]Liu W Y,Wen Y D,Yu Z D,et al.Large-margin Softmax loss for convolutional neural networks[C]//Proceedings of the 33rd International Conference on International Conference on Machine Learning.New York:ACM Press,2016:507-516
[20]Ioffe S,Szegedy C.Batch normalization:accelerating deep network training by reducing internal covariate shift[C]//Proceedings of the 32nd International Conference on International Conference on Machine Learning.New York:ACMPress,2015:448-456
[21]Nair V,Hinton G E.Rectified linear units improve restricted Boltzmann machines[C]//Proceedings of the 27th International Conference on International Conference on Machine Learning.New York:ACM Press,2010:807-814