Key Takeaways:
- A collaborative project between WUR Vision + Robotics researchers and Viscon aims to develop technology for automated selection of healthy young plants for greenhouse transplantation.
- The technology utilizes machine learning and plant phenotyping to identify key health characteristics in plants, potentially saving growers time and resources by focusing on viable plants.
- The project is innovative in its approach to grading plants and predicting their survivability in greenhouse conditions through AI, differing from current manual or semi-automated processes.
- An RGB camera with a pericentric lens, capable of capturing 360-degree images, has been developed to facilitate comprehensive plant health assessment.
- This technology could be applicable not only to tissue-culture-grown plants but also to various important agricultural crops, representing a significant advancement in horticultural automation.
Enhancing Efficiency in Greenhouse Operations
The agricultural industry stands on the brink of a significant technological revolution with the introduction of the Dynamic Machine Learning project, an ambitious initiative aimed at modeling and predicting complex quality traits in fresh produce. At the heart of this project is a pioneering effort to automate the selection of only healthy young plants for transplantation into greenhouses, a process that has traditionally been labor-intensive and subjective.
A Collaborative Endeavor for Horticultural Innovation
This groundbreaking work is the result of a collaboration between researchers from WUR Vision + Robotics and Viscon, who are dedicated to developing a new technology that could transform the horticultural industry. The technology aims to automate the sorting of non-viable plants before they are transferred to the greenhouse, thereby optimizing resource allocation toward the cultivation of healthy plants.
The Power of Machine Learning and Plant Phenotyping
Suzane Pols, team lead of Plant Science at Viscon, highlights the significance of machine learning and plant phenotyping in this endeavor. These cutting-edge fields offer promising tools for enhancing the cultivation process of plants grown from tissue culture, one of the key commercial agricultural crops. By converting visual observations and data into actionable insights, the technology seeks to streamline the selection and sorting process, making it an automated, efficient, and less laborious task for growers.
Defining Plant Health through AI
The question of what constitutes a healthy plant is a complex one within plant science. The project aims to leverage machine learning to identify predictors of plant health, facilitating the separation of healthy plants for improved sorting efficiency. This involves the development of an automated, reliable, and scalable method for plant health assessment, as explained by Lydia Meesters, project leader and scientist in Computer Vision at WUR’s Vision + Robotics programme.
Technological Breakthroughs and Future Directions
A notable innovation of the project is the development of an RGB camera equipped with a pericentric lens, capable of capturing a comprehensive 360-degree view of a plant’s roots in a single image. This technological breakthrough allows for a detailed examination of plant health indicators.
As the project progresses through its feasibility study, the team remains focused on exploring optimal techniques for visualizing plant health traits. The ultimate goal is to refine and scale this technology, paving the way for its application across various agricultural crops and significantly advancing the automation of horticultural processes.
Image provided by Viscon
1 Comment