With the exponential growth of digital transactions, individuals and organizations generate a massive volume of financial records on a daily basis. Despite this growth, expense tracking is still largely performed manually using physical receipts and spreadsheet-based methods. These approaches are inefficient, error-prone, and unsuitable for long-term financial analysis. This paper presents a Smart Receipt OCR and Expense Tracker Using Machine Learning, an intelligent system that automates the extraction, categorization, storage, and analysis of expense data from receipt images. The system integrates Optical Character Recognition (OCR) for text extraction, Natural Language Processing (NLP) for text cleaning and structuring, and supervised Machine Learning algorithms for expense classification. A relational database ensures secure and scalable data storage, while interactive dashboards provide analytical insights such as category- wise expenditure, monthly trends, and budget utilization. Extensive experimentation demonstrates improved accuracy, reduced manual effort, and enhanced usability, making the proposed solution suitable for large-scale real- world deployment.
Sawant et al. (Tue,) studied this question.