ScanCart

Ever had trouble checking if an item is offered at another retailer for a lower price? With ScanCart, finding the best deal is as simple as snapping a photo, compared with other retailers like Amazon

Comment

Inspiration

The inspiration behind ScanCart came from the frustration many of us feel when shopping online: trying to find the best price for a product across multiple retailers. With so many platforms to compare, it can be time-consuming and difficult to ensure you're getting the best deal. We wanted to make this process simpler and more efficient for consumers. The idea was to create an app that could scan a product image and automatically show the best prices across different retailers—saving time and money for users.

What it does

ScanCart is a full-stack retail application that allows users to scan a product image using either file upload or webcam capture. It leverages the Google Cloud Vision API to recognize and extract product information, such as labels and brands. The app then uses a web scraping API to compare prices of the same product from retailers like Amazon and Target. Users can instantly see the product details, AI-generated descriptions, and price comparisons to help them make informed purchasing decisions.

How we built it

Frontend: React for building a dynamic and responsive user interface. Backend: Flask to serve as the backend, which handles requests for uploading images and processing the price comparison logic. Google Cloud Vision API for product image recognition and label extraction. Web scraping (OxyLabs) to fetch product pricing from multiple online retailers.

Challenges we ran into

Handling image Processing: One of the major challenges was getting the image to be properly captured and recognized in a 64 base format.

Web Scraping Limitations: Using the Web Scraping API proved to be difficult because of the IP restriction we could not figure out for a certain amount of time, resorting to try and building our own web scraper.

What's next for ScanCart

We want to integrate a machine learning model to improve product recognition and match the scanned products more accurately to retailers' listings. This will reduce dependency on generic image recognition and make the tool smarter.

Built With

Share this project:

Updates