Videometer at Seed Quality Innovation Day 2026: Advancing Digital Seed Analysis with Spectral Imaging and AI
On February 13th, 2026, Videometer will take part at the Seed Quality Innovation Day, held in Woodland, California and organized by InSight Seed Technologies and Aginnovation, sales partners to Videometer in the U.S.
Our CEO, Jens Michael Carstensen, will give a presentation on how spectral imaging and AI-based models are transforming seed processing and quality control.
From Physical Seeds To Digital Seeds
Seed quality assessment has traditionally relied on manual inspection, destructive testing, and time-consuming laboratory workflows. In his presentation, Spectral imaging for seed processing and QC – AI models based on high-quality physico-chemical data, Dr. Carstensen, will introduce the concept of digital seeds: rich digital representations of seeds that capture physical, chemical, and biological properties in one single, non-destructive analysis.
Using LED-based spectral imaging across UV (ultra-violet), VIS (visible), and NIR (near-infrared) wavelengths, seeds and grains can be analyzed far beyond what is possible with conventional RGB imaging. This enables objective and repeatable evaluation of parameters such as seed purity, damage, health, maturity, germination potential, and coating quality, directly from image data.
Why Spectral Imaging Matters For Seed Quality
A key focus of the presentation is the difference between seed appearance and seed chemistry. While visual inspection is strongly influenced by illumination, gloss, and geometry, spectral imaging isolates the diffuse reflectance and fluorescence signals that are directly linked to seed coat chemistry and internal properties.
By combining high-quality spectral data with machine learning models trained on reference seed collections, it becomes possible to:
- Classify crop seeds, weeds, and inert material with high accuracy
- Detect mechanical damage, mold, and specific pathogens
- Assess maturity, vigor, and early germination potential
- Quantify and validate seed coatings and treatments
These capabilities support faster decision-making, improved consistency, and scalable quality control in both R&D and production environments.
AI Models Built On Reliable Physical and Chemical Data
Rather than treating AI as a black box, the approach presented emphasizes data quality first. High signal-to-noise spectral images, standardized illumination, and calibrated systems form the foundation for robust and transferable AI models . This allows seed companies, laboratories, and genebanks to build reliable digital workflows that can be validated, shared, and maintained over time.
Meet us in Woodland, California
Seed Quality Innovation Day brings seed professionals, researchers, and technology providers together to explore the future of seed testing and quality assurance. We look forward to contributing to the discussion and sharing practical insights from more than two decades of spectral imaging experience in agrifood applications.
If you’re attending Seed Quality Innovation Day 2026, we hope you find Dr. Carstensen’s presentation insightful and invite you to connect with the Videometer, InSight Seed Technologies and Aginnovation teams, to learn more about digital seed analysis and advanced spectral imaging solutions.