Non-Destructive Quality Testing of Potatoes using NIR HyperspectralImaging and Machine Learning

How will this research impact Alberta’s agriculture industry?

Canada produced 5.8 million tonnes of potatoes and Alberta accounted for 21.5 percent of the total production in 2018 with over $900 million value. The Canadian potatoes generate $1.17 billion in farm cash receipts and have an export value of $1.72 billion. The major quality issues associated with potatoes are internal defects, greening, specific gravity and sugar content.

Most of these quality parameters associated with potatoes are currently detected using destructive methods, need significant manual work and are subjective and slow. This project aims to use an innovative and non-destructive near infrared (NIR) hyperspectral imaging system and machine learning techniques to detect quality parameters associated with potatoes, which will reduce the time and cost associated with quality testing.

The project will identify most significant wavelengths required to detect the quality parameters associated with potatoes. The NIR hyperspectral imaging system will be tested at the speeds simulating commercial scanning speed and eventually design a prototype for commercial application of the system based on extensive testing and analysis.

Why did RDAR invest in this research project?

Most of the quality testing in potato industry is done manually and involves destructive wet chemical analysis. Detecting internal defects is big challenge in commercial applications. Large potato company , ‘The Little Potato Company’ has indicated their concern about greening and slow specific gravity tests. This project is expected to result in significant digitation of QA in potato industry. Automated and non-destructive analytical tools will help potato processing industries in further improving processing efficiency and end-product quality. Since growers are paid on the quality specifications mutually agreed on the contract, rapid and accurate measurements will be beneficial for growers when they deliver the potatoes for processing.

The quality assurance technology developed from this research project can be also applied to other agricultural products such as sugar beets.

How will research knowledge be transferred and shared with producers?

The following knowledge transfer activities will be conducted to share learnings from this project:

• A promotional video explaining the system will be produced with the help from Lethbridge College media and marketing team and shared with potato industry and growers.
• The technology will be showcased at The International Potato Technology Expo takes place biennially in Charlottetown (PEI) for a large potato industry audience.
• The technology will be showcased in major Food Expo happening in Alberta and Canada.
• Project investigators will provide content for PGA’s newsletter based on the progress of project and results.
• Research work will be published in scientific journals and results will be presented at the national/international technical conferences and PGA’s annual meetings and other relevant conferences in Alberta.