Skip to main content

Overview

What is FertiScan?​

FertiScan is a web application usable on browsers (computer and mobile), currently under development, which leverages Artificial Intelligence (AI) technology. This technology includes the use of Optical Character Recognition (OCR) and advanced language models, enabling high-precision text processing. In short, FertiScan can "read" OCR results and reformulating this data into a structured form with predefined fields. The app aims to simplify and streamline the process of capturing, extracting, organizing, and storing the information present on fertilizer labels, thereby increasing the efficiency and accuracy of inspectors from the Canadian Food Inspection Agency (CFIA) in their tasks.

Key Features​

  • AI technology and OCR capabilities: Enables scanning of fertilizer labels and extracting essential data.

  • Data storage: Integrates an organized database system for easy access and analysis of collected information.

  • Flexibility: Offers integration with other software platforms and customization through tailored AI models.

Why are we doing this?​

The project aims to:

  • Increase precision and efficiency: The application ensures the elimination of errors and reduces the time spent on the manual data entry process of label information.

  • Improve organization: The information currently collected is dispersed within the RDIMS (Records Document and Information Management System) system. FertiScan proposes a centralized and structured storage solution, addressing one of the main current challenges.

  • Facilitate access to and analysis of data: Decision-making is faster and based on updated and accurate data.

  • Resource conservation: To save time and efforts of the stakeholders involved in the process.

  • Regulatory compliance: To reinforce adherence to current standards and regulations in the sector.

  • Usability: To increase user satisfaction and operational efficiency through a practical and easy-to-use application.

Who is FertiScan for?​

FertiScan is primarily aimed at CFIA inspectors and potentially other stakeholders in the agricultural sector, such as fertilizer distributors and manufacturers, as well as regulatory agencies that could benefit from easier access to fertilizing product data.

How does FertiScan work?​

The FertiScan application would operate as follows:

  1. Capture: An inspector uses the app to take photos of fertilizer labels.

  2. Extraction: The AI analyzes the image and extracts the relevant textual information.

  3. Organization: The extracted data is sorted by labels and validated.

  4. Storage: The extracted data is uploaded into a centralized database.

Additionally, we are considering the development of a simple interface providing an overview of the labels stored in the database to facilitate verification and data updates.