Across Regions, Breeding, Most viewed stories, News March 2024, North America, Research, Studies/Reports, Varieties

The digital eye in potato breeding: How machine vision is shaping the future of spuds

A groundbreaking scientific development has emerged from the fields of agriculture and technology, promising a brighter future for potato breeders and farmers worldwide. Published in the Plant Phenome Journal on March 8, 2024, a team of pioneering scientists in the U.S. has introduced a novel, cost-effective phenotyping strategy poised to transform the potato breeding process.

The paper, entitled “A scalable, low-cost phenotyping strategy to assess tuber size, shape, and the colorimetric features of tuber skin and flesh in potato breeding populations,” outlines an innovative approach to evaluating crucial tuber characteristics through advanced machine vision.

In the quest to cultivate the perfect potato, size, shape, color, and resistance to defects play pivotal roles in the acceptance of new cultivars. However, the traditional methods of measuring these traits have been labor-intensive and imprecise, posing significant challenges to potato breeders and farmers alike.

Enter the game-changing solution from a team of visionary scientists: a scalable, semiautomated workflow that employs machine vision to evaluate these essential tuber characteristics with unprecedented precision. This technology, akin to giving computers the power to see and analyze potatoes as a human expert would, has been successfully applied to the 189 F1 progeny of the A08241 breeding population progeny by the research team, showcasing its potential to revolutionize potato breeding practices.

Key Highlights:

  • Superhero Vision for Spuds: Machine vision technology can now accurately quantify potato tuber size, shape, and color features, offering a superhero-esque capability to see the finest details of potato characteristics.
  • Weighing with a Glance: In a remarkable feat, the system can infer tuber weight from simple two-dimensional images, eliminating the need for physical weighing and streamlining the assessment process.
  • Beyond Round or Long: Moving past traditional aspect ratio measurements, this technology utilizes biomass profiles to provide a nuanced understanding of tuber shape, offering a more comprehensive view of a potato’s physical attributes.
  • AI to the Rescue: Leveraging deep learning models, the system can identify tubers with the hollow-heart defect, ensuring that only the best quality potatoes make it through the breeding process.
  • Genetics at Play: The research also highlights that the traits measured by machine vision are predominantly under genetic control, emphasizing the role of inheritance in potato breeding and offering new insights into how we select and propagate desirable traits.


This breakthrough represents a significant leap forward in potato breeding, offering a scalable and cost-effective tool that could lead to the development of superior potato varieties. By harnessing the power of machine vision and artificial intelligence, breeders can now make more informed decisions, accelerate the breeding process, and ultimately meet the growing global demand for this staple food in a more efficient manner.

Article authored by Potato News Today
Source: Plant Phenome Journal open access paper
Scientific paper:
Cover image: Credit WikimediaImages from Pixabay

Editor & Publisher: Lukie Pieterse

Feel free to get in touch with Lukie!
He’ll be happy to share your company’s news stories on Potato News Today:
Connect on LinkedIn
Follow on Twitter
About us



FPS Food Process Solutions