Yield is a crucial trait in crops and is used to assess the efficiency and productivity of farming practices. It is influenced by various factors, including crop variety, soil quality, irrigation, climate, harvesting and post-harvest handling, among others with implications for farmers, researchers, and policymakers.
To increase the efficiency and productivity, farmers and researchers continually work to improve crop yields through different strategies, including the development of new agricultural technologies and sustainable farming practices. Accurate measurement and analysis of crop yield data are essential for making informed decisions to optimize agricultural production.
As a response to the importance of this factor, the development of PHSPP which offers automatic, real-time estimation of yield based on the count, size, color, shape etc. of seeds combined with its weight and moisture content, was implemented for the Bean Program by the Mechatronics & Automation Lab at the Alliance Bioversity International and CIAT Palmira’s campus in Colombia.
The automatic count of seeds from photographs of different genotypes is possible applying computer vision techniques. Also, to obtain data on phenotypic traits such as color, shape, size, and brilliance are feasible using lighting-controlled conditions. Crossing this information with the bean breeding program's databases could easily identify genotypes to improve according to selection and evaluation criteria.
This system is currently in the validation and training stage with the next parameters:
• Market Class
• Primary seed color
• Secondary seed color
• Seed count
• Seed Weight (100 Seed weight and total seed weight)
• Seed Moisture Content
In what context is this tool useful?
The PHSPP is useful in a context where seed morphological information is valuable in breeding programs, plant identification, seed banking, crop production, seed germination, ecological studies, and other research fields. It enables accurate identification, conservation, propagation, and understanding of plant species, supporting various applications in agriculture, and scientific inquiry.
This tool empowers breeders, rigorously testing thousands of bean genotypes to select the best, enhancing yield potential, disease resistance, and seed quality. It's now deployed in Colombia and Uganda, promising better beans for improved livelihoods and sustainability.
Jennifer Wilker - email@example.com (Principal Scientist)
Fabricio Soto - firstname.lastname@example.org (Technical Support).