On Optical Character Recognition Research Articles

Innovative AI methods for monitoring front-of-package information: A case study on infant foods.

Front-of-package (FOP) is one of the most direct communication channels connecting manufacturers and consumers, as it displays crucial information such as certification, nutrition, and health. Traditional methods for obtaining information from FOPs often involved manual collection and analysis. To overcome these labor-intensive characteristics, new methods using two artificial intelligence (AI) approaches were applied for information monitoring of FOPs. In order to provide practical implementations, a case study was conducted on infant food products. First, FOP images were collected from Amazon.com. Then, from the FOP images, 1) the certification usage status of the infant food group was obtained by recognizing the certification marks using object detection. Moreover, 2) the nutrition and health-related texts written on the images were automatically extracted based on optical character recognition (OCR), and the associations between health-related texts were identified by network analysis. The model attained a 94.9% accuracy in identifying certification marks, unveiling prevalent certifications like Kosher. Frequency and network analysis revealed common nutrients and health associations, providing valuable insights into consumer perception. These methods enable fast and efficient monitoring capabilities, which can significantly benefit various food industries. Moreover, the AI-based approaches used in the study are believed to offer insights for related industries regarding the swift transformations in product information status.

Robust OCR Pipeline for Automated Digitization of Mother and Child Protection Cards in India

The Universal Immunization Programme in India has a mandate to fully vaccinate all of India’s 27 million children born annually. The vaccination doses are recorded by frontline health workers on standardized paper-based Mother and Child Protection (MCP) cards, which are manually digitized by data entry operators, resulting in poor data quality, delays, and significant time and resources. In our article, we focus on Optical Character Recognition– (OCR) based automated digitization of MCP card images captured through a smartphone application developed by us. By utilizing a standardized template for the MCP cards, which is available a priori , we register the card images and perform OCR on the extracted region of interest (ROIs). Since the cards with curvature or torn edges had poor ROIs, we built a global–local alignment technique that first approximates the ROI using global homography and then refines using a local homography resulting in improved accuracy. Our pipeline gives a character level accuracy of 98.73% on our dataset against 75.02% by Google Cloud Vision and 79.26% by Azure OCR. We also describe our field testing experience, where the digitized MCP card images were used to provide useful features on the smartphone application for health workers to conduct vaccination sessions.

Smart Parking System Based on Image Processing

Abstract: Successful implementation of smart parking systems can effectively reduce a lot of problems related to traffic congestion in urban areas. Wastage of fuel and time in search of a vacant parking space will be significantly reduced and the complete utilization of the available parking areas can make our cities really smart. Parking slot detection and user notification are the two major sections of a smart parking system. The empty parking space detection was initially done by deploying a number of sensors in the parking lot. It is highly expensive and complicated to install. But the advancement of image processing has enabled us to use images of the parking area to find out empty spaces. In this paper, a comparative study of the various parking space identification techniques has been done. Also, image processing-based system models have been presented as a replacement for sensor-based systems. A very efficient and simple technique for parking slot identification based on optical character recognition (OCR) has been introduced in this paper. The camera installed in the parking lot captures the image of the parking area with specially numbered parking spaces. The OCR system recognizes numbers that are not hidden by a vehicle parked over it and this information is used for identifying the empty spaces.

Smart Parking System Based On Image Processing

Abstract: Successful implementation of smart parking systems can effectively reduce a lot of problems related to traffic congestion in urban areas. Wastage of fuel and time in search of a vacant parking space will be significantly reduced and the complete utilization of the available parking areas can make our cities really smart. Parking slot detection and user notification are the two major sections of a smart parking system. The empty parking space detection was initially done by deploying a number of sensors in the parking lot. It is highly expensive and complicated to install. But the advancement of image processing has enabled us to use images of the parking area to find out empty spaces. In this paper, a comparative study of the various parking space identification techniques has been done. Also, image processing-based system models have been presented as a replacement for sensor-based systems. A very efficient and simple technique for parking slot identification based on optical character recognition (OCR) has been introduced in this paper. The camera installed in the parking lot captures the image of the parking area with specially numbered parking spaces. The OCR system recognizes numbers that are not hidden by a vehicle parked over it and this information is used for identifying the empty spaces.

One datum and many values for sustainable Industry 4.0: a prognostic and health management use case

<span>Industrial context of today, driven by the Industry 4.0 paradigm, is overwhelmed by data. Decreasing cost of innovative technologies, and recent market dynamics have pushed and pulled respectively for those architectures and practices in which data are the masters. While advancing, we have to take care of waste, even though intangibility of data makes them hardly connected to waste. In this paper we are going to reflect on data intensive context of today, focusing on the industrial sector. A smart approach for fully exploiting data collecting infrastructures is proposed, and its declination in a prognostic and health management (PHM) use case set inside an automatic painting system is presented. The contributions of this papers are mainly two: first of all, the general conceptual take-away of "data re-use" is presented and discussed. Moreover, a PHM solution for painting system's number plates, based on optical character recognition (OCR), is proposed and tested as a proof-of-concept for the "data re-use" concept. Summarizing, the already-in-use data sharing principle for achieving transparency and integration inside Industry 4.0, is presented as complementary with the proposed "data re-use", in order to develop a really sustainable shift toward the future.</span>

Switching Text-Based Image Encoders for Captioning Images With Text

Visual understanding, such as image caption generation, has received extensive attention. Describing images with textual information is one way to help people achieve barrier-free visibility. This study focuses on the text-based image captioning (TextCaps) task. The TextCaps task is more complex than the traditional image captioning task because it depends on optical character recognition (OCR) and the textual information that appears in the image. It also requires consideration of the relationship between recognized objects and OCR’s linguistic part in the image. In this study, we propose maximizing the use of multiple modalities in an image to improve TextCaps performance. We enrich the image and OCR linguistic features using pre-trained Contrastive Language-Image Pre-training (CLIP) models. We then introduce using two additional attention models in a transformer architecture to strengthen the representation of the image modality. The experimental results demonstrate that our proposed method, which introduces a multimodal transformer with four image-related modalities, outperforms existing methods for the TextCaps dataset.

Designing a Real-Time-Based Optical Character Recognition to Detect ID Cards

This research 0aims to design a Real-time ID card detection based on Optical Character Recognition (OCR). OCR detects and records information into CSV files using a camera. Hopefully, it can become one of the administrative solutions in Indonesia by using existing identity cards using OCR in real time. This research method was carried out independently in August 2021 using ID cards as objects. The tool involved was a 320x320 pixel webcam camera on an HP Intel Core i5 7th Gen notebook. The software used by Easy OCR was Pytorch-based. ID cards were detected using an algorithm by TensorFlow object detection with SSD MobileNet V2 FPNLite 320x320 as the pre-trained model of Tensorflow. The researchers collected ID card images using a webcam with various light conditions and orientations and label them using labeling. The researchers trained it with only 20 photos. After 3000 training steps, the researchers obtained about 0.17 loss and 0.95. Thus, the ID card detection tool using OCR runs well.

Handwritten Bangla Numerical Digit Recognition Using Fine Regulated Deep Neural Network

The recognition of handwritten Bangla digit is providing significant progress on optical character recognition (OCR). It is a very critical task due to the similar pattern and alignment of handwriting digits. With the progress of modern research on optical character recognition, it is reducing the complexity of the classification task by several methods, a few problems encounter during recognition and wait to be solved with simpler methods. The modern emerging field of artificial intelligence is the Deep Neural Network, which promises a solid solution to these few handwritten recognition problems. This paper proposed a fine regulated deep neural network (FRDNN) for the handwritten numeric character recognition problem that uses convolutional neural network (CNN) models with regularization parameters which makes the model generalized by preventing the overfitting. This paper applied Traditional Deep Neural Network (TDNN) and Fine regulated deep neural network (FRDNN) models with a similar layer experienced on BanglaLekha-Isolated databases and the classification accuracies for the two models were 96.25% and 96.99%, respectively over 100 epochs. The network performance of the FRDNN model on the BanglaLekha-Isolated digit dataset was more robust and accurate than the TDNN model and depend on experimentation. Our proposed method is obtained a good recognition accuracy compared with other existing available methods.

Building an efficient OCR system for historical documents with little training data

As the number of digitized historical documents has increased rapidly during the last a few decades, it is necessary to provide efficient methods of information retrieval and knowledge extraction to make the data accessible. Such methods are dependent on optical character recognition (OCR) which converts the document images into textual representations. Nowadays, OCR methods are often not adapted to the historical domain; moreover, they usually need a significant amount of annotated documents. Therefore, this paper introduces a set of methods that allows performing an OCR on historical document images using only a small amount of real, manually annotated training data. The presented complete OCR system includes two main tasks: page layout analysis including text block and line segmentation and OCR. Our segmentation methods are based on fully convolutional networks, and the OCR approach utilizes recurrent neural networks. Both approaches are state of the art in the relevant fields. We have created a novel real dataset for OCR from Porta fontium portal. This corpus is freely available for research, and all proposed methods are evaluated on these data. We show that both the segmentation and OCR tasks are feasible with only a few annotated real data samples. The experiments aim at determining the best way how to achieve good performance with the given small set of data. We also demonstrate that obtained scores are comparable or even better than the scores of several state-of-the-art systems. To sum up, this paper shows a way how to create an efficient OCR system for historical documents with a need for only a little annotated training data.

A Real-Time Automatic Plate Recognition System Based on Optical Character Recognition and Wireless Sensor Networks for ITS.

Automatic License Plate Recognition has been a recurrent research topic due to the increasing number of cameras available in cities, where most of them, if not all, are connected to the Internet. The video traffic generated by the cameras can be analyzed to provide useful insights for the transportation segment. This paper presents the development of an intelligent vehicle identification system based on optical character recognition (OCR) method to be used on intelligent transportation systems. The proposed system makes use of an intelligent parking system named Smart Parking Service (SPANS), which is used to manage public or private spaces. Using computer vision techniques, the SPANS system is used to detect if the parking slots are available or not. The proposed system makes use of SPANS framework to capture images of the parking spaces and identifies the license plate number of the vehicles that are moving around the parking as well as parked in the parking slots. The recognition of the license plate is made in real-time, and the performance of the proposed system is evaluated in real-time.

English

English

Compressing CNN-DBLSTM models for OCR with teacher-student learning and Tucker decomposition

Integrated convolutional neural network (CNN) and deep bidirectional long short-term memory (DBLSTM) based character models have achieved excellent recognition accuracies on optical character recognition (OCR) tasks, along with large amount of model parameters and massive computation cost. To deploy CNN-DBLSTM model in products with CPU server, there is an urgent need to compress and accelerate it as much as possible, especially the CNN part, which dominates both parameters and computation. In this paper, we study teacher-student learning and Tucker decomposition methods to reduce model size and runtime latency for CNN-DBLSTM based character model for OCR. We use teacher-student learning to transfer the knowledge of a large-size teacher model to a small-size compact student model, followed by Tucker decomposition to further compress the student model. For teacher-student learning, we design a novel learning criterion to bring in the guidance of succeeding LSTM layer when matching the CNN-extracted feature sequences of the large teacher and small student models. Experimental results on large scale handwritten and printed OCR tasks show that, using teacher-student learning alone achieves 9.90 × footprint reduction and 15.23 × inference speedup yet without degrading recognition accuracy. Combined with Tucker decomposition method, we can compress and accelerate the model further. The decomposed model achieves 11.89 × footprint reduction and 22.16 × inference speedup while suffering no or only a small recognition accuracy degradation against the large-size baseline model.

Research on automatic recognition system for code of wheel hub model based on OCR

The composition, operating principle and image processing algorithm of automatic recognition system for code of wheel hub model based on optical character recognition (OCR)are introduced in this paper. In addition, characteristics of character information about wheel hub model are analyzed, algorithm of OCR based on gray-scale template matching is expounded and thus wheel hub model can be effectively identified. The system can meet requirements of mass production and online quick sorting for wheel hub production line, which can overcome disadvantages of conventional manual sorting, improve productivity and alleviate labor intensity.

Word Detection and Translation

The project basically focuses on Optical Character Recognition (OCR) and then translating the recognized character into the language understandable to the user. When traveling to foreign countries one finds it difficult to read and understand the signs, conventions or characters. To make it easy this application simply recognizes the character which the user wants to translate and then the user can simply translate that character into the language understandable to him/her. The main aim to develop this project is to help travelers, tourists or even the students to translate a character which is not understandable to him/her. All of the three users will use the product for three different purposes. In general any user will take a picture of the character or sign which he/she cannot understand using the camera of his/her phone. The processing of the character will start at the backed where each character will be broken down into nine sections and accordingly IP and CNN algorithms will be applied on that character and once the algorithms are applied, the character is matched with the training set which is already stored in the database. Once a match is found the translated character is shown back to the user.

Pattern Recognition Of Javanese Letter Using Template Matching Correlation Method

In this modern age, the impact of globalization is increasingly entering and expanding into most societies. One impact of globalization makes people prefer to learn the language and use a foreign language rather than the local language, especially the Java language. It is very influential on the knowledge of the community about the existence or the existence of Javanese Letter, especially in the field of education. In this study, In this research will be made an application to recognize the writing of Javanese Letter based on Optical Character Recognition (OCR). Matching templates correlation can be used as pattern recognition methods. How the Template Matching Algorithm works by matching the template image with the test image after going through the Pre-processing and segmentation process. From the research that has been done by using 10 character template and 20 data testing get accuracy equal to 93.44% and error rate 6.56%. So the Matching Template Algorithm can well recognize the Javanese Letter pattern.

Framework based on Mobile Augmented Reality for Translating Food Menu in Thai Language to Malay Language

Augmented reality (AR) technology is a technique that combines the real world and the virtual world digitally using mobile devices. Mobile AR technology is expected to help Malaysian tourists who have difficulties to understand the Thai language when visiting the country. Hence, a prototype called ARThaiMalay translator was developed to translate printed Thai food menu to Malay language. The objective of this study is to design a food menu translation framework from Thai to Malay language based on mobile AR, develop a translator application and to test the effectiveness of the translator application. The prototype consists of three main components which are translation based on optical character recognition (OCR) technology, dictionary development using SQLite database and display data from the local database. Evaluation of the developed application shows its effectiveness to perform translation of Thai text with certain features to Malay language.

Gujarati handwritten numeral recognition through fusion of features and machine learning techniques

Languages have played a major role in Indian history and they continue to influence the lives of Indians till date. Plentiful research on optical character recognition (OCR) techniques for Indian languages such as Hindi, Tamil, Bangla, Kannada, Gurumukhi, Malayalam and Marathi has already been carried out. Research efforts on Gujarati character recognition are few and yet to gain momentum. This paper intends to bring Gujarati character recognition in attention. Methods based on artificial neural network (ANN), support vector machine (SVM) and naive Bayes (NB) classifier are exercised for handwritten Gujarati numerals recognition. Experiments are carried out on two large datasets using three different kinds of features and their fusion. Zone-based, projection profiles-based and chain code-based features are employed as individual features. The paper proposes to use a fusion of these features for learning prediction models. Experimental results show significant improvement over state-of-the-art and validate our proposals.

Gujarati handwritten numeral recognition through fusion of features and machine learning techniques

Languages have played a major role in Indian history and they continue to influence the lives of Indians till date. Plentiful research on optical character recognition (OCR) techniques for Indian languages such as Hindi, Tamil, Bangla, Kannada, Gurumukhi, Malayalam and Marathi has already been carried out. Research efforts on Gujarati character recognition are few and yet to gain momentum. This paper intends to bring Gujarati character recognition in attention. Methods based on artificial neural network (ANN), support vector machine (SVM) and naive Bayes (NB) classifier are exercised for handwritten Gujarati numerals recognition. Experiments are carried out on two large datasets using three different kinds of features and their fusion. Zone-based, projection profiles-based and chain code-based features are employed as individual features. The paper proposes to use a fusion of these features for learning prediction models. Experimental results show significant improvement over state-of-the-art and validate our proposals.

Penerapan Histogram Equalization pada Optical Character Recognition Preprocessing

In this paper, we will discuss about the implementatoion of Histogram Equalization for images contrast enhancement, in the preprocessing step, on Optical Character Recognition(OCR). OCR has several steps, including preprocessing step. Implementing images contrast enhancement algorithm will make it easier. It is important for images to have high level contrast. It makes those images clear. Changing the histogram will make the colors of images also change. The output will be taken to the next step processing and we will get more accurate recognition.

A survey on optical character recognition for Bangla and Devanagari scripts

The past few decades have witnessed an intensive research on optical character recognition (OCR) for Roman, Chinese, and Japanese scripts. A lot of work has been also reported on OCR efforts for various Indian scripts, like Devanagari, Bangla, Oriya, Tamil, Telugu, Malayalam, Kannada, Gurmukhi, Gujarati, etc. In this paper, we present a review of OCR work on Indian scripts, mainly on Bangla and Devanagari—the two most popular scripts in India. We have summarized most of the published papers on this topic and have also analysed the various methodologies and their reported results. Future directions of research in OCR for Indian scripts have been also given.