iProViS: Intelligent Product Vision System
This project aims to recognize products in retail sectors such as supermarkets, clothing, food, electronics, and others using computer vision technology instead of barcodes or QR codes. The system is designed to develop a mobile application that allows users to access product price information, compare prices across different stores, obtain product content information with multilingual support, and gain detailed insights about products.
Additionally, a new framework has been proposed for recyclable product packaging. The recognized product is integrated into the intelligent recycling process by determining its type through the API of the iPRoVis system and incorporating information such as weight when empty or full. This feature enables users to save time by having their discarded packaging recognized and sorted by smart recycling machines during the recycling process. The developed system not only enhances consumer experience but also contributes significantly to sustainability and environmentally friendly approaches within green IT strategy.
Recognizing products in retail sectors such as supermarkets, clothing, food, electronics, and others using computer
By taking only one photo of the product, you will be able to access the stock and price information of the relevant product in other markets, including the stock information and other features of the product in that market.
The user can compare the prices of the same product across various stores and e-commerce platforms. The app displays the lowest price at the nearest store for the selected product.
Contributing significantly to sustainability and environmentally friendly approaches within the realm of green IT.
Store owners will be able to track the stock status and price information of other stores in the marketplaces.
- Product Image Detection: Insufficient ambient light may hinder the detection of product images.
- Internet Dependency: Devices without an internet connection may struggle to access the iProVis server.
- Computer Vision
- Software Development
- Li, Q., Peng, X., Cao, L., Du, W., Xing, H., Qiao, Y., & Peng, Q. (2020). Product image recognition with guidance learning and noisy supervision. Computer Vision and Image Understanding, 196, 102963.
- Bai, Y., Chen, Y., Yu, W., Wang, L., & Zhang, W. (2020). Products-10k: A large-scale product recognition dataset. arXiv preprint arXiv:2008.10545.
- Chen, S., Liu, D., Pu, Y., & Zhong, Y. (2022). Advances in deep learning-based image recognition of product packaging. Image and Vision Computing, 128, 104571.
- Programming Languages:
- Python
- Dart
- Frameworks:
- Flutter
- Yolo v8.0
- Libraries:
- Various Python libraries (Numpy, Opencv) for Computer Vision (e.g., TensorFlow Lite).
- Dart libraries for Computer Vision (e.g., TensorFlow Lite, Flutter Vision).
The product image may not be detected due to insufficient ambient light.
Alternative Solution: In case of any negative scenario that may occur in product recognition with the iProVis system, it will be ensured that it works in harmony with QR code or barcode components. The multilingual feature will also be integrated for these components.
The problem of not being able to access the iProVis server on a device without an internet connection.
Alternative Solution: In case of no internet connect
| Advisor |
|---|
| Gül Tokdemir |
| Sezer Uğuz |
| Team Member | Numbers |
|---|---|
| Nursena Bitirgen | 202011029 |
| Tamer Memiş | 201911210 |
| Furkan Yamaner | 202011211 |
| Boran Gülbaşar | 202011033 |