ChakmaNumerals: A Comprehensive Handwritten Numerals Dataset in Chakma Language

The ChakmaNumerals dataset consists of single-page handwritten samples collected from 600 individuals of diverse ages and personalities. Each page is annotated with bounding boxes around every numeral,The dataset includes 600 samples per class. All bounding boxes and labels were manually generated to ensure accuracy. Digit recognition plays a crucial role in numerous real-world applications, such as vehicle license plate identification, postal mail sorting, bank check processing, and automated form data entry. Moreover, digit recognition from images remains a fundamental research area in optical character recognition (OCR) and pattern classification.