Handwritten Marathi character recognition using R -HOG Feature

There are several supervised machine learning methods used for the application of automated morphological classification of galaxies; however, there has not yet been a clear comparison of these different methods using imaging data, or a investigation for maximising their effectiveness. We carry out a comparison between several common machine learning methods for galaxy classification (Convolutional Neural Network (CNN), K-nearest neighbour, Logistic Regression, Support Vector Machine, Random Forest, and Neural Networks) by using Dark Energy Survey (DES) data combined with visual classifications from the Galaxy Zoo 1 project (GZ1). Our goal is to determine the optimal machine learning methods when using imaging data for galaxy classification. We show that CNN is the most successful method of these ten methods in our study. Using a sample of $\sim$2,800 galaxies with visual classification from GZ1, we reach an accuracy of $\sim$0.99 for the morphological classification of Ellipticals and Spirals. The further investigation of the galaxies that have a different ML and visual classification but with high predicted probabilities in our CNN usually reveals an the incorrect classification provided by GZ1. We further find the galaxies having a low probability of being either spirals or ellipticals are visually Lenticulars (S0), demonstrating that supervised learning is able to rediscover that this class of galaxy is distinct from both Es and Spirals. We confirm that $\sim$2.5\% galaxies are misclassified by GZ1 in our study. After correcting these galaxies' labels, we improve our CNN performance to an average accuracy of over 0.99 (accuracy of 0.994 is our best result).

Optimizing automatic morphological classification of galaxies with machine learning and deep learning using Dark Energy Survey imaging

As an emerging and applicable method, deep learning (DL) has attracted much attention in recent years With the development of DL and the massive of publications and researches in this direction, a comprehensive analysis of DL is necessary In this paper, from the perspective of bibliometrics, a comprehensive analysis of publications of DL is deployed from 2007 to 2019 (the first publication with keywords “deep learning” and “machine learning” was published in 2007) By preprocessing, 5722 publications are exported from Web of Science and they are imported into the professional science mapping tools: VOS viewer and Cite Space Firstly, the publication structures are analyzed based on annual publications, and the publication of the most productive countries/regions, institutions and authors Secondly, by the use of VOS viewer, the co-citation networks of countries/regions, institutions, authors and papers are depicted The citation structure of them and the most influential of them are further analyzed Thirdly, the cooperation networks of countries/regions, institutions and authors are illustrated by VOS viewer Time-line review and citation burst detection of keywords are exported from Cite Space to detect the hotspots and research trend Finally, some conclusions of this paper are given This paper provides a preliminary knowledge of DL for researchers who are interested in this area, and also makes a conclusive and comprehensive analysis of DL for these who want to do further research on this area

A bibliometric analysis on deep learning during 2007–2019

The offline character recognition is very useful software in the field of research. Authoritative field of research has made by character recognition due to its need in various fields of research as in banks, post offices to fulfill all recognition requirements. This paper is an exploration on the different scripts including Mathematical digits, Hindi consonants and vowels, Gurumukhi characters as well as numerals. Different handwritten samples of all these language scripts are taken from different persons in a good environment. First of all preprocessing including different operations for the reduction of noise is operated. The pre-processed documents are then segmented for the feature extraction. In this technique, different features like lines, corners etc. are to be extracted. Number of hidden layers taken are 15. Then neural network is trained for the testing of all these handwritten samples. Out of 100% samples, 90% are used for training, 5% are for testing and the remaining of the samples are used for validation. Comparison of recognition rates achieved on different scripts is accessible. Out of all samples taken, highest accuracy is achieved on the Gurumukhi script which is 99.9%.

/pdf/offline-handwritten-character-recognition-using-neural-2tt1uc2hp1.pdf

Offline Handwritten Character Recognition using Neural Network

Handwritten character recognition has been profoundly studied for many years in the field of pattern recognition. Due to its vast practical applications and financial implications, handwritten character recognition is still an important research area. In this research, the Handwritten Ethiopian Character Recognition (HECR) dataset has been prepared to train the model. The images in the HECR dataset were organized with more than one color pen RGB main spaces that have been size normalized to 28 × 28 pixels. The dataset is a combination of scripts (Fidel in Ethiopia), numerical representations, punctuations, tonal symbols, combining symbols, and special characters. These scripts have been used to write ancient histories, science, and arts of Ethiopia and Eritrea. In this study, a hybrid model of two super classifiers: Convolutional Neural Network (CNN) and eXtreme Gradient Boosting (XGBoost) is proposed for classification. In this integrated model, CNN works as a trainable automatic feature extractor from the raw images and XGBoost takes the extracted features as an input for recognition and classification. The output error rates of the hybrid model and CNN with a fully connected layer are compared. A 0.4630 and 0.1612 error rates are achieved in classifying the handwritten testing dataset images, respectively. Thus XGBoost as a classifier performs a better result than the traditional fully connected layer.

/pdf/a-new-hybrid-convolutional-neural-network-and-extreme-3vnuhsb7ww.pdf

A New Hybrid Convolutional Neural Network and eXtreme Gradient Boosting Classifier for Recognizing Handwritten Ethiopian Characters

Marathi is one of the ancient Indian languages majorly spoken in the state of Maharashtra. Marathi is one of the Devanagari script and the literals and numerals are almost similar to Hindi. Recognition of handwritten Marathi numerals is quite challenging task because people have the practice of writing these numerals in variant ways. In this work we have presented a method to recognize the handwritten Marathi numerals using multilayer feed-forward neural network. The scanned document image is pre-processed to eliminate the noise and care is taken to link the broken characters. Each numeral is segmented from the document and it is resized to 7 × 5 pixels using cubic interpolation. While resizing a technique is used to provide better representation for every pixel in segmented numeral. This resized numeral is converted into a vector with 35 values before inputting it to the neural network. We have used 100 sets containing 1000 numerals for this experimentation, of which 50 sets are used for training the network and 50 sets for the testing purpose. The overall recognition rate of the proposed method is 97%.

Recognition of Marathi Handwritten Numerals Using Multi-Layer Feed-Forward Neural Network

Handwritten character recognition of the Marathi language is a challenging task because characters are complex in structure. 31320 samples of characters from different writers have been collected and database is prepared. Noise is removed by using morphological and thresholding operation. Skewed scanned pages and segmented characters are corrected using Hough Transformation. The characters are segmented from scanned pages by using bounding box techniques. Size variation of each handwritten Marathi characters are normalized in 40 X 40 pixel size. Here we propose feature extraction from handwritten Marathi characters using connected pixel based features like area, perimeter, eccentricity, orientation and Euler number. The k-nearest neighbor (KNN) algorithm with five fold validation has been used for result preparation. The accuracy of proposed method is 85.88 % obtained.

Geometrical Features Extraction and KNN Based Classification of Handwritten Marathi Characters

In this work we show that the sparse autoencoder based feature extraction method in combination with deep neural network based classifier can produce enhanced results when applied to Marathi character recognition. However, in pattern recognition tasks, such as the optical character recognition problem, it is difficult to directly convert handwritten character document image into its constituent character data. Here, we proposed classification of handwritten Marathi character using Deep Neural Network (DNN), which is based on two key concepts: autoencoder features and deep neural network architectures. The proposed model extracts the valid patches from character zones using voting schemes. These extracted patches are arranged in row vector and further fed into DNN for training and testing. We evaluated our approach on handwritten numeral and character of two cross datasets. The proposed classification technique has shown higher performance results over both
datasets.

Deep neural network for handwritten Marathi character recognition

Robust handwritten Marathi character recognition is essential to the proper function in document analysis field. Many researches in OCR have been dealing with the complex challenges of the high variation in character shape, structure and document noise. In proposed system, noise is removed by using morphological and thresholding operation. Skewed scanned pages and segmented characters are corrected using Hough Transformation. The characters are segmented from scanned pages by using bounding box techniques. Size variation of each handwritten Marathi characters are normalized in 40 $\times $ 40 pixel size. Here we propose feature extraction from handwritten Marathi characters using connected pixel based features like area, perimeter, eccentricity, orientation and Euler number. The modified k-nearest neighbor (KNN) and SVM algorithm with five fold validation has been used for result preparation. The comparative accuracy of proposed methods are recorded. In this experiment modified SVM obtained high accuracy as compared with KNN classifier.

Parshuram M. Kamble

Papers

Handwritten Marathi character recognition using R -HOG Feature

Recognition of Marathi Handwritten Numerals Using Multi-Layer Feed-Forward Neural Network

Geometrical Features Extraction and KNN Based Classification of Handwritten Marathi Characters

Deep neural network for handwritten Marathi character recognition

Comparative Study of Handwritten Marathi Characters Recognition Based on KNN and SVM Classifier