Robust Recognition of Chinese Text from Cellphone-acquired Low-quality Identity Card Images Using Convolutional Recurrent Neural Network

Wang, Jianmei; Wu, Ruize; Zhang, Shaoming

doi:10.18494/sam.2021.2991

Cited by 6 publications

(5 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…(III) For text detection, the DB model trained on a public dataset with 1,670 images (17,548 annotated regions) at the 13 th International Conference on Document Analysis and Recognition (ICDAR) is used in our deep learning system since this model is the best among 44 methods summited to the ICDAR 2015 Robust Reading Competition and its detection accuracy reached to 83.79%. For text recognition, the CRNN model is used in our deep learning system, since it is suitable for multi-lingual text recognition and its average accuracy reached 98.57% for Chinese text recognition ( 47 ). These excellent methods suitable for processing our UPBMR images are used in our system to improve the accuracy of character recognition.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

DeepSSR: a deep learning system for structured recognition of text images from unstructured paper-based medical reports

et al. 2022

View full text Add to dashboard Cite

Background Complete electronic health records (EHRs) are not often available, because information barriers are caused by differences in the level of informatization and the type of the EHR system. Therefore, we aimed to develop a deep learning system [deep learning system for structured recognition of text images from unstructured paper-based medical reports (DeepSSR)] for structured recognition of text images from unstructured paper-based medical reports (UPBMRs) to help physicians solve the data-sharing problem. Methods UPBMR images were firstly preprocessed through binarization, image correction, and image segmentation. Next, the table area was detected with a lightweight network (i.e., the proposed YOLOv3-MobileNet model). In addition, the text of the table area was detected and recognized with the model based on differentiable binarization (DB) and convolutional recurrent neural network (CRNN). Finally, the recognized text was structured according to its row and column coordinates. DeepSSR was trained and validated on our dataset with 4,221 UPBMR images which were randomly split into training, validation, and testing sets in a ratio of 8:1:1. Results DeepSSR achieved a high accuracy of 91.10% and a speed of 0.668 s per image. In the system, the proposed YOLOv3-MobileNet model for table detection achieved a precision of 97.8% and a speed of 0.006 s per image. Conclusions DeepSSR has high accuracy and fast speed in structured recognition of text based on UPBMR images. This system may help solve the data-sharing problem due to information barriers between hospitals with different EHR systems.

show abstract

Section: Discussionmentioning

confidence: 99%

“…These key processing methods include image correction based on the inclination angle in the image pre-processing stage, the proposed YOLOv3-MobileNet model in the (47). These excellent methods suitable for processing our UPBMR images are used in our system to improve the accuracy of character recognition.…”

Section: Discussionmentioning

confidence: 99%

DeepSSR: a deep learning system for structured recognition of text images from unstructured paper-based medical reports

et al. 2022

View full text Add to dashboard Cite

show abstract

“…They achieved a character error rate of 6.81% on the ICDAR 2013 competition dataset. By adding DenseNet, Wang et al [16] upgraded a convolutional recurrent neural network and achieved an average character recognition accuracy of 98.57%. A data synthesizer based on conditional adversarial generative networks is also being developed to create synthetic ID card text line graphics.…”

Section: Deep Learning Chinese Character Recognitionmentioning

confidence: 99%

Large-Scale Printed Chinese Character Recognition for ID Cards Using Deep Learning and Few Samples Transfer Learning

Chang

Lin

2022

Applied Sciences

View full text Add to dashboard Cite

In the field of computer vision, large-scale image classification tasks are both important and highly challenging. With the ongoing advances in deep learning and optical character recognition (OCR) technologies, neural networks designed to perform large-scale classification play an essential role in facilitating OCR systems. In this study, we developed an automatic OCR system designed to identify up to 13,070 large-scale printed Chinese characters by using deep learning neural networks and fine-tuning techniques. The proposed framework comprises four components, including training dataset synthesis and background simulation, image preprocessing and data augmentation, the process of training the model, and transfer learning. The training data synthesis procedure is composed of a character font generation step and a background simulation process. Three background models are proposed to simulate the factors of the background noise patterns on ID cards. To expand the diversity of the synthesized training dataset, rotation and zooming data augmentation are applied. A massive dataset comprising more than 19.6 million images was thus created to accommodate the variations in the input images and improve the learning capacity of the CNN model. Subsequently, we modified the GoogLeNet neural architecture by replacing the fully connected layer with a global average pooling layer to avoid overfitting caused by a massive amount of training data. Consequently, the number of model parameters was reduced. Finally, we employed the transfer learning technique to further refine the CNN model using a small number of real data samples. Experimental results show that the overall recognition performance of the proposed approach is significantly better than that of prior methods and thus demonstrate the effectiveness of proposed framework, which exhibited a recognition accuracy as high as 99.39% on the constructed real ID card dataset.

show abstract

“…State Grid Corporation of China proposed the "Three-year Action Plan for Lean Operation of Measurement Assets" in 2021 to strengthen research on measurement data. With measurement data management as the focus, the action plan aims to achieve the processing, storage, transmission, exchange, and management activities of measurement data, to excel in the entire process control and management supervision of measurement data, to accelerate the transformation from traditional measurement models to information models, to provide timely and accurate feedback and verification mechanisms for measurement data, and to utilize the advantages of measurement data in testing and inspection (Wang et al, 2021b). Currently, State Grid Corporation of China has in place a three-level measurement system consisting of the State Grid Measurement Center, Provincial Marketing Service Center (Measurement Center), and county-level measurement institutions.…”

Section: Introductionmentioning

confidence: 99%

Data Recognition for Multi-Source Heterogeneous Experimental Detection in Cloud Edge Collaboratives

Yubo,

Jing,

Xiaomeng

et al. 2023

International Journal of Information Technologies and Systems Approach

View full text Add to dashboard Cite

This article proposes a multisource heterogeneous experimental data recognition method based on CRNN and DBNet in a cloud-edge collaborative environment in an attempt to address the issues of low efficiency and a high error rate that come with traditional manual data detection and recognition. Firstly, a recognition architecture for experimental detection data of intelligent measurement systems is designed based on a cloud-edge collaborative environment to improve the efficiency of data processing. Then, the improved DBNet network is used in the edge computing center to detect the text, and the correction module is used to correct the deviation of the detected text to ensure the standardization of the text. Finally, in the central cloud, the end-to-end indefinite length character recognition (CRNN) algorithm is used to analyze and identify the text order, rules, and other information of the image after the correction is completed, extract the test detection data, and convert the detection data into row data according to the two-dimensional table structure, and conduct structured storage and management through the relational database. An experimental analysis of the proposed method is conducted based on the deep learning framework, and results show that its accuracy rate and recall rate are close to 96% and 94%, respectively, with an average accuracy of 95.09%. This fully demonstrates the proposed method is effective and, therefore, applicable to power equipment experimental detection data recognition.

show abstract

Robust Recognition of Chinese Text from Cellphone-acquired Low-quality Identity Card Images Using Convolutional Recurrent Neural Network

Cited by 6 publications

References 18 publications

DeepSSR: a deep learning system for structured recognition of text images from unstructured paper-based medical reports

DeepSSR: a deep learning system for structured recognition of text images from unstructured paper-based medical reports

Large-Scale Printed Chinese Character Recognition for ID Cards Using Deep Learning and Few Samples Transfer Learning

Data Recognition for Multi-Source Heterogeneous Experimental Detection in Cloud Edge Collaboratives

Contact Info

Product

Resources

About