Handwritten Marathi Character Dataset in Devanagari Script
Description
This dataset consists of handwritten Marathi characters written in the Devanagari script. All samples were collected manually and scanned at 300 DPI resolution. Each image is preprocessed and converted into a 28×28 grayscale format for consistency and compatibility with deep learning models. The dataset is designed to support research in Optical Character Recognition (OCR), character recognition, and pattern classification tasks. It provides a valuable resource for training and evaluating models such as Convolutional Neural Networks (CNNs) and Quantum Neural Networks (QNNs). This dataset aims to contribute to the development of efficient recognition systems for the Marathi language and promote research in regional language processing using the Devanagari script.
Files
Steps to reproduce
The dataset was created using handwritten samples of Marathi characters written in the Devanagari script. Four contributors participated in collecting and preparing the dataset. Each participant wrote all characters on white A4 sheets using black ink pens under consistent lighting conditions. The handwritten sheets were then scanned at 300 DPI resolution using a flatbed scanner to ensure high-quality image capture. After scanning, the images were cropped, cleaned, and converted into 28×28 grayscale images using Python and OpenCV for preprocessing. give in point format
Institutions
- Annasaheb Dange College of Engineering and Technology