PriceSnap: A Mobile Application Framework for Real-Time Grocery Identification and Price Comparison

PriceSnap is a proposed mobile application framework for real-time grocery item identification and price comparison across multiple retailers. Leveraging a ResNet-50 model optimized with TensorFlow Lite and a price aggregation module integrating retailer APIs and web scraping with MongoDB caching, PriceSnap achieves projected metrics of 92.3% accuracy, 91.1% F1-score, and 1.8s end-to-end latency (±0.2s). This paper presents a comprehensive design framework, including a simulated dataset of 15,000 images across 50 grocery items, system architecture, mathematical formulations, and a theoretical evaluation against state-of-the-art methods. The framework supports applications in meal planning, inventory management, and sustainable shopping, with robust considerations for privacy, accessibility, and scalability.