An Efficient Method for Bangla Handwritten Digit Recognition Using Convolutional Neural Network Indronil Bhattacharjee Department of Computer Science New Mexico State University indronil@nmsu.edu Abstract. Handwritten digit recognition is a fundamental problem in the field of computer vision and pattern recognition. This paper presents a Convolutional Neural Network (CNN) approach for recognizing handwritten Bangla digits. The proposed method utilizes a dataset of handwritten Bangla digit images and trains a CNN model to classify these digits accurately. The dataset is preprocessed to enhance the quality of the images and make them suitable for training the CNN model. The trained model is then tested on a separate test dataset to evaluate its performance in terms of accuracy. With the Ekush: Bangla Handwritten Data - Numerals dataset, we tested our CNN implementation to determine the precision of handwritten characters. According to the test results, 25% of the images using a training set of more than 150,000 images from Ekush dataset had an accuracy of 98.3%. Keywords. Handwritten character recognition, Bangla number, Computer vision, Deep learning, Convolutional neural networks, Classification, Image processing 1. Introduction Handwritten digit recognition is a significant research area in the fields of computer vision and pattern recognition. It plays a crucial role in various applications including optical character recognition, document analysis, and postal automation systems. With the increasing availability of digital devices and the need for efficient digit recognition systems, the development of accurate and reliable methods for recognizing handwritten digits has gained immense importance. In this paper, we focus on the specific task of recognizing handwritten Bangla digits, which presents unique challenges owing to the complexity and distinctiveness of the Bangla script. Bangla, also known as Bengali, is the official language of Bangladesh and is widely spoken in the Indian states of West Bengal, Tripura, and Assam. It has its own distinct script that consists of a set of complex characters and symbols. Recognizing handwritten Bangla digits is particularly important for applications in various domains, including digit recognition systems in educational institutions, automated form processing, and digital- based document analysis. Despite the increasing demand for accurate Bangla digit recognition systems, existing methods often face difficulties in accurately recognizing handwritten Bangla digits because of their varying writing styles, distortions, and overlapping strokes. In this study, we propose a Convolutional Neural Network (CNN) approach for recognizing handwritten Bangla digits. CNNs have proven highly effective in various computer vision tasks, including image classification, object detection, and digit recognition. By leveraging the power of deep learning and 65 Technium Vol. 18, pp.65-74 (2023) ISSN: 2668-778X www.techniumscience.com