Handwritten Letter Recognition Research Articles

Handwritten Character Recognition using Deep Learning Algorithm with Machine Learning Classifier

Handwritten character recognition is a problem that has been worked on for many mainstream languages. Handwritten letter recognition has been proven to achieve promising results. Several studies using deep learning models have been conducted to achieve better accuracies. In this paper, the authors conducted two experiments on the EMNIST Letters dataset: Wavemix-Lite and CoAtNet. The Wavemix-Lite model uses Two-Dimensional Discrete Wavelet Transform Level 1 to reduce the parameters and speed up the runtime. The CoAtNet is a combined model of CNN and Visual Transformer where the image is broken down into fixed-size patches. The feature extraction part of the model is used to embed the input image into a feature vector. From those two models, the authors hooked the value of the features of the Global Average Pool layer using EMNIST Letters data. The features hooked from the training results of the two models, such as SVM, Random Forest, and XGBoost models, were used to train the machine learning classifier. The experiments conducted by the authors show that the best machine-learning model is the Random Forest, with 96.03% accuracy using the Wavemix-Lite model and 97.90% accuracy using the CoAtNet model. These results showcased the benefit of using a machine learning model for classifying image features that are extracted using a deep learning model.

Approaches to the recognition of handwritten letters of the Azerbaijani language and their analysis

Optical character recognition (OCR) computer systems are widely used for recognition of car license plates, for enabling visually impaired people to work with written information, etc. One of the current areas of OCR problems is the recognition of handwritten characters. This study proposes approaches to the construction of combined models using simple and ensemble learning for recognition of handwritten letters of the Azerbaijani language. The models built using the proposed approaches utilized several classes of features. Based on the results of computer experiments, a comparative analysis of the recognition accuracy of the constructed models and features using the proposed recognition methods is conducted.

Information technology for recognition of handwritten ukrainian letters and numbers using synthetic data sets

The paper considers several variants of the architecture of convolutional neural networks for recognizing isolated handwritten digits and Ukrainian letters. The results of recognizing different images containing letters and numbers were compared on models with different architectures. Several variants of rather complex architectures of neural networks were considered. Research was conducted with VGG16 and VGG19 architectures, ResNet or ResNetV2, MobileNet or MobileNetV2, DenseNet. The possibility of learning convolutional neural networks with the help of a synthetic data set built on the basis of handwritten or cursive fonts is shown. The size of the training data set significantly affects the reliability of character recognition. The data sets used in the work correspond in volume to the well-known EMNIST Letters dataset. The lower limit of the sample size, which provides acceptable recognition accuracy, was about 1500 characters per class. Reducing the sample by reducing the number of symbols per class leads to a significant decrease in recognition accuracy (from 90-100% accuracy of recognizing elements of real inscriptions to 40-60% with a 4-fold reduction in the sample size). Thus, when using deep neural networks to recognize letters or numbers, the reliability of recognizing elements of real inscriptions depended primarily on the size of the training data set. The accuracy of recognition of the test data set after training all variants of the models was quite high - 97-98% and higher. However, the generation of training data sets of small size — 300-500 images per class - practically did not provide reliable recognition. In general, when comparing the achieved recognition accuracy of real images and the model training speed, the best performance was provided by the DenseNet or ResNetV2 family models. Experiments with changing the optimization algorithm compared to Adam did not give any improvement in the accuracy and reliability of recognizing real samples. Increasing the number of model training epochs beyond the specified one also did not change the results.

An intelligent approach for Arabic handwritten letter recognition using convolutional neural network.

Currently, digital transformation has occurred in most countries in the world to varying degrees, but digitizing business processes are complex in terms of understanding the various aspects of manual documentation. The use of digital devices and intelligent systems is vital in the digital transformation of manual documentation from hardcopy to digital formats. The transformation of handwritten documents into electronic files is one of the principal aspects of digitization and represents a common need shared by today’s businesses. Generally, handwriting recognition poses a complex digitization challenge, and Arabic handwriting recognition, specifically, proves inordinately challenging due to the nature of Arabic scripts and the excessive diversity in human handwriting. This study presents an intelligent approach for recognizing handwritten Arabic letters. In this approach, a convolution neural network (CNN) model is proposed to recognize handwritten Arabic letters. The model is regularized using batch normalization and dropout operations. Moreover, the model was tested with and without dropout, resulting in a significant difference in the performance. Hence, the model overfitting has been prevented using dropout regularization. The proposed model was applied to the prominent, publicly-available Arabic handwritten characters (AHCD) dataset with 16,800 letters, and the performance was measured using several evaluation measures. The experimental results show the best fit of the proposed model in terms of higher accuracy results that reached 96.78%; additionally, other evaluation measures compared to popular domain-relevant approaches in the literature.

Comparative Analysis of the Application of Multilayer and Convolutional Neural Networks for Recognition of Handwritten Letters of the Azerbaijani Alphabet

Introduction. The implementation of information technologies in various spheres of public life dictates the creation of efficient and productive systems for entering information into computer systems. In such systems it is important to build an effective recognition module. At the moment, the most effective method for solving this problem is the use of artificial multilayer neural and convolutional networks. The purpose of the paper. This paper is devoted to a comparative analysis of the recognition results of handwritten characters of the Azerbaijani alphabet using neural and convolutional neural networks. Results. The analysis of the dependence of the recognition results on the following parameters is carried out: the architecture of neural networks, the size of the training base, the choice of the subsampling algorithm, the use of the feature extraction algorithm. To increase the training sample, the image augmentation technique was used. Based on the real base of 14000 characters, the bases of 28000, 42000 and 72000 characters were formed. The description of the feature extraction algorithm is given. Conclusions. Analysis of recognition results on the test sample showed: as expected, convolutional neural networks showed higher results than multilayer neural networks; the classical convolutional network LeNet-5 showed the highest results among all types of neural networks. However, the multi-layer 3-layer network, which was input by the feature extraction results; showed rather high results comparable with convolutional networks; there is no definite advantage in the choice of the method in the subsampling layer. The choice of the subsampling method (max-pooling or average-pooling) for a particular model can be selected experimentally; increasing the training database for this task did not give a tangible improvement in recognition results for convolutional networks and networks with preliminary feature extraction. However, for networks learning without feature extraction, an increase in the size of the database led to a noticeable improvement in performance. Keywords: neural networks, feature extraction, OCR.

English Feature Recognition Based on GA-BP Neural Network Algorithm and Data Mining.

With the development of society and the promotion of science and technology, English, as the largest universal language in the world, is used by more and more people. In the life around us, there is information in English all the time. However, because the process of manual recognition of English letters is very labor-intensive and inefficient, the demand for computer recognition of English letters is increasing. This paper studies the influence of the parameters of BP neural network and genetic algorithm on the whole network, including the input, output, and number of hidden layer nodes. Finally, it improves and determines the settings and values of the relevant parameters. On this basis, it shows the rationality of the selected parameters through experiments. The results show that only GA-BP neural network and feature data mining algorithm can complete feature extraction and become the main function of feature classification at the same time. After enough initial data sample analysis training, the GA-BP neural network was found to have good data fault tolerance and feature recognition. The experimental results show that the genetic algorithm can find the best weights and thresholds and the weights and thresholds are given to the BP neural network. After training, the recognition of handwritten letters can be realized. Finally, the convergence of the two algorithms is compared through experiments, which shows that the overall performance of the BP neural network algorithm is improved after genetic algorithm optimization. It can be seen that the genetic algorithm has a good effect in improving the BP neural network and this method has a broad prospect in English feature recognition.

A novel handwritten Turkish letter recognition model based on convolutional neural network

AbstractConvolutional neural networks have provided state‐of‐the‐art solutions for many subfields of computer vision. While there exist many studies in the literature for several languages, studies for handwritten Turkish character recognition lack in the research field. To this end, we propose a novel handwritten Turkish letter recognition model based on a convolutional neural network. Since, to the best of our knowledge, there do not exist any publicly available handwritten Turkish letters datasets, we constructed a handwritten Turkish letters dataset that consists of 25,875 samples. To compare the performance of the proposed model with the related work, three state‐of‐the‐art models, namely, VGG19, InceptionV3, and Xception, were utilized through the transfer learning technique. When these models were evaluated on the handwritten Turkish letter dataset, the proposed model's accuracy was calculated as high as 96.07% which was higher than the benchmark models. To measure the generalization ability of the proposed model, it was evaluated on a gold standard dataset, namely, EMNIST, and has achieved an accuracy of 80.54% which was higher than the benchmark models. Finally, the proposed model was trained and evaluated on the EMNIST dataset and it has achieved an accuracy of 94.61% which outperformed the related work.

Pengenalan Pola Tulisan Tangan Aksara Arab Menggunakan Ekstraksi Fitur Discrete Cosine Transform Dan Klasifikasi Backpropagation Artificial Neural Network

Arabic is a language that is spoken as the first or native language of more than 280 million people, most of whom live in the Middle East and North Africa. Apart from the Middle East and North Africa, Arabic is also familiar and often used in Indonesia because of the majority of Indonesia's population is Muslim and Arabic is the language of worship in Islam. The recognition of Arabic handwritten letters is one of the studies that has been done before, where the accuracy results obtained vary according to the research and methods used. This study aims to determine the accuracy resulting from the recognition of Arabic script handwriting patterns using a combination of the DCT(Discrete Cosine Transform) feature extraction method and the ANN Backpropagation classification method. The data used for this study were data from handwritten sources on A4 HVS paper using markers with categories of age from 7-13 years old and 18-23 years old with 15 respondents in each group and a total dataset image of 8400. Testing the best model model obtained on all images produces an accuracy of 80.79%, using the images of age range 17-23 years produces 87.27% accuracy, and the images of age range 7-13 produces an accuracy of 72.84%.
 Keywords: pengenalan pola, tulisan tangan, aksara, DCT, backpropagation

Application of Backpropagation Algorithm for Handwriting Recognition

Handwritten letter recognition is one form of pattern recognition. The introduction of letters looks simple to humans, but it becomes a tough task for computer programs to complete. In recognizing someone’s handwriting, a computer program trained first. This study discusses how a computer recognizes a digital image pattern in the form of handwritten letters using the backpropagation algorithm. The backpropagation algorithm used in making this system is backpropagation with momentum. The network architecture used consists of an input screen with 64 neurons, a hidden screen consisting of 46 neurons, and an output screen with five neurons. Before the recognition process, input image with image format (jpg) processed first, which includes scaling, grayscale, and binarization. The trial results show that the system can recognize handwriting with an accuracy of up to 65%.

Recognition of handwritten letters and numbers using deep learning neural networks

Recognition of handwritten letters and numbers using deep learning neural networks

Graph Neural Networks Using Local Descriptions in Attributed Graphs: An Application to Symbol Recognition and Hand Written Character Recognition

Graph-based methods have been widely used by the document image analysis and recognition community, as the different objects and the content in document images is best represented by this powerful structural representation. Designing of novel computation tools for processing these graph-based structural representations has always remained a hot topic of research. Recently, Graph Neural Network (GNN) have been used for solving different problems in the domain of document image analysis and recognition. In this article we take forward the state of the art by presenting a new approach to gather the symbolic and numeric information from the nodes and edges of a graph. We use this information to learn a Graph Neural Network (GNN). The experimentation on the recognition of handwritten letters and graphical symbols shows that the proposed approach is an interesting contribution to the growing set of GNN-based methods for document image analysis and recognition.

Turkish handwriting recognition system using multi-layer perceptron

Recently, handwriting recognition has found many application areas along with technological advances. Handwriting recognition systems can greatly simplify human life by reading tax returns, forwarding mail, reading bank checks, and so on. On the other hand, these systems can reduce the need for human interaction. Therefore, academic and commercial studies of handwriting characters have recently become an important research topic in pattern recognition. In this study, Turkish handwritten letter recognition system from A to Z was developed in C++ environment by using Artificial Neural Networks (ANNs). After the feature data were extracted, handwriting images were presented to the network, the training process of ANN was completed, and different handwriting images were classified with trained ANN. In this study, MLP (Multi-Layered Perceptron: MLP) type ANN and back-propagation learning algorithm were used. The ANN used has 35 inputs and 23 outputs. In the hidden layer, ANNs with different numbers of artificial neural cells (neurons) were evaluated and the most appropriate neural number ANN was selected. As a result, ANN with 24 neurons was selected in the hidden layer and handwriting images was classified with an accuracy rate of 94.90 %.

Depth for Curve Data and Applications

ABSTRACT In 1975, John W. Tukey defined statistical data depth as a function that determines the centrality of an arbitrary point with respect to a data cloud or to a probability measure. During the last decades, this seminal idea of data depth evolved into a powerful tool proving to be useful in various fields of science. Recently, extending the notion of data depth to the functional setting attracted a lot of attention among theoretical and applied statisticians. We go further and suggest a notion of data depth suitable for data represented as curves, or trajectories, which is independent of the parameterization. We show that our curve depth satisfies theoretical requirements of general depth functions that are meaningful for trajectories. We apply our methodology to diffusion tensor brain images and also to pattern recognition of handwritten digits and letters. Supplementary materials for this article are available online.

PrivEdge: From Local to Distributed Private Training and Prediction

Machine Learning as a Service (MLaaS) operators provide model training and prediction on the cloud. MLaaS applications often rely on centralised collection and aggregation of user data, which could lead to significant privacy concerns when dealing with sensitive personal data. To address this problem, we propose PrivEdge, a technique for privacy-preserving MLaaS that safeguards the privacy of users who provide their data for training, as well as users who use the prediction service. With PrivEdge, each user independently uses their private data to locally train a one-class reconstructive adversarial network that succinctly represents their training data. As sending the model parameters to the service provider in the clear would reveal private information, PrivEdge secret-shares the parameters among two non-colluding MLaaS providers, to then provide cryptographically private prediction services through secure multi-party computation techniques. We quantify the benefits of PrivEdge and compare its performance with state-of-the-art centralised architectures on three privacy-sensitive image-based tasks: individual identification, writer identification, and handwritten letter recognition. Experimental results show that PrivEdge has high precision and recall in preserving privacy, as well as in distinguishing between private and non-private images. Moreover, we show the robustness of PrivEdge to image compression and biased training data. The source code is available at https://github.com/smartcameras/PrivEdge .

MCSM-Wri: A Small-Scale Motion RecognitionMethod Using WiFi Based on Multi-ScaleConvolutional Neural Network.

Small-scale motion recognition has received wide attention recently with the development of environmental perception technology based on WiFi, and some state-of-the-art techniques have emerged. The wide application of small-scale motion recognition has aroused people’s concern. Handwritten letter is a kind of small scale motion, and the recognition for small-scale motion based on WiFi has two characteristics. Small-scale action has little impact on WiFi signals changes in the environment. The writing trajectories of certain uppercase letters are the same as the writing trajectories of their corresponding lowercase letters, but they are different in size. These characteristics bring challenges to small-scale motion recognition. The system for recognizing small-scale motion in multiple classes with high accuracy urgently needs to be studied. Therefore, we propose MCSM-Wri, a device-free handwritten letter recognition system using WiFi, which leverages channel state information (CSI) values extracted from WiFi packets to recognize handwritten letters, including uppercase letters and lowercase letters. Firstly, we conducted data preproccessing to provide more abundant information for recognition. Secondly, we proposed a ten-layers convolutional neural network (CNN) to solve the problem of the poor recognition due to small impact of small-scale actions on environmental changes, and it also can solve the problem of identifying actions with the same trajectory and different sizes by virtue of its multi-scale characteristics. Finally, we collected 6240 instances for 52 kinds of handwritten letters from 6 volunteers. There are 3120 instances from the lab and 3120 instances are from the utility room. Using 10-fold cross-validation, the accuracy of MCSM-Wri is 95.31%, 96.68%, and 97.70% for the lab, the utility room, and the lab+utility room, respectively. Compared with Wi-Wri and SignFi, we increased the accuracy from 8.96% to 18.13% for recognizing handwritten letters.

Automatic Learning Algorithms for Local Support Vector Machines

In this paper, we propose automatic learning algorithms (called auto-kSVM-1p, autokSVM-dp) being possible to automatically tune hyper-parameters of k local support vector machines (SVM) for classifying large data sets. The autokSVM algorithms are able to determine the number of clusters k to partition the large training data, followed which they auto-learn the non-linear SVM model in each cluster to classify the data locally in the parallel way on multi-core computers. The autokSVM algorithms combine the grid search, the .632 bootstrap estimator, the hill climbing heuristic to optimize hyper-parameters in the local non-linear SVM training. The numerical test results on four data sets from UCI repository and three benchmarks of handwritten letters recognition showed that our autokSVM algorithms give very competitive classification results compared to the standard LibSVM and the original kSVM. An example of autokSVM-1p’s effectiveness is given with an accuracy of 96.74% obtained in the classification of Forest cover-type data set having 581,012 datapoints in 54-dimensional input space and 7 classes in 334.45 seconds using a PC Intel(R) Core i7-4790 CPU, 3.6 GHz, 4 cores.

Handwritten capital letter recognition based on OpenCV

Handwriting capitalization recognition is a function of distinguishing handwritten capital letters by means of machine or computer intelligence, which is classified into the field of optical character recognition. Given that capital letters are widely used around the world, identification and analysis are often used as the main components of some control systems. Therefore, the research on handwritten capital letter recognition is also very practical and has important practical significance. The key part of the research contained in this paper is the image preprocessing and the optimal selection of feature vectors, and finally completes the design of handwritten digit recognition system. In this paper, the Fourier and Bayesian commonly used are compared, and eventually the Fourier transform feature is applied to the system classification identification. After completing the test on the relevant experimental data, the results show that the handwritten capital recognition system established in this paper has a high recognition accuracy for handwritten capital letters after repeated training.

A Genetic-LVQ neural networks approach for handwritten Arabic character recognition

Handwritten recognition systems are a dynamic field of research in areas of artificial intelligence. Many smart devices available in the market such as pen-based computers, tablets, mobiles with handwritten recognition technology need to rely on efficient handwritten recognition systems. In this paper we present a novel Arabic character handwritten recognition system based on a hybrid method consisting of a genetic algorithm and a Learning vector quantization (LVQ) neural network. Sixty different handwritten Arabic character datasets are used for training the neural network. Each character dataset contains 28 letters written twice with 15 distinct shaped alphabets, and each handwritten Arabic letter is represented by a binary matrix that is used as an input to a genetic algorithm for feature selection and dimension reduction to include only the most effective features to be fed to the LVQ classifier. The recognition process in the system involves several essential steps such as: handwritten letter acquisition, dataset preparation, feature selection, training, and recognition. Comparing our results to those acquired by the whole feature dataset without selection, and to the results using other classification algorithms confirms the effectiveness of our proposed handwritten recognition system with an accuracy of 95.4%, hence, showing a promising potential for improving future handwritten Arabic recognition devices in the market.

Isolated Arabic Hand Written Letters Recognition Based on Contour Matching and Neural Network

Complexity of Arabic writing language makes its handwritten recognition very complex in terms of computer algorithms comparing with other languages such as English and French. The Arabic handwritten recognition has high importance in modern applications. The contour analysis of word image can extract special contour features that discriminate one character from another by the mean of vector features. This paper aims to implements a set of pre-processing functions over a handwritten Arabic character, with contour analysis, to enter the contour vector to neural network to recognize it. For training part, the neural network architecture was trained using many patterns regardless of the Arabic font style building a rigid recognition model. The presented algorithm structure got recognition ratio about 97%.

Handwriting Trajectory Reconstruction Using Low-Cost IMU

In this paper, we propose a trajectory reconstruction method based on low-cost Inertial Measurement Unit (IMU) in smartphones. The IMU used in our work consists of a three-axis accelerometer and a three-axis gyroscope, which can record information of acceleration and rotation, respectively. Since intrinsic bias and random noise usually cause unreliable IMU signals, filtering methods are utilized to reduce high- or low-frequency noises of the signals. In addition, to more accurately detect whether the smartphone is moving or not, we extract multiple features from IMU signals and train a movement detection model based on linear discriminant analysis (LDA). Then, a “reset switch” mechanism is applied when the smartphone is detected as in a static state. The “reset switch” mechanism can effectively restrain the accumulated error of displacement calculation. Finally, the trajectory reconstruction results are applied to handwritten letter recognition for English alphabet, and the experimental results show that our proposed trajectory reconstruction method is reliable.