Multispectral Imaging for Distinguishing Hybrid Forest Seeds

Conventional seed testing methods can be extremely labor-intensive for laboratory technicians, and the destructive nature of many of these tests does not allow for seed conservation after testing.

Seed companies have been looking for new technologies, which allow for rapid, non-destructive, and more ergonomic working standards while testing seeds for purity, health, vigor and more. In this regard, spectral imaging represents an effective technology, not requiring labor-intensive tasks and allowing for non-destructive analysis.

PhD Candidate, Thomas Michelon, researched the effectiveness of the VideometerLab Spectral Imaging technology for the analysis of hybrid forest seed varieties. In particular, he utilized the instrument to analyze Eucalyptus seeds, Corymbia Seeds and hybrids of the two. These seeds are particularly complex to distinguish because of their similar traits and their very small size.

 In July 2024, his research was published on the Journal of Environmental Management.

Dr. Michelon utilized the VideometerLab with 19 bands, ranging from 365 nm to 970 nm i.e. from UV to Visible Spectrum to NIR. He acquired images of the different phenotypic varieties, in order to distinguish one from the other.

His research concluded that the Videometer Spectral Imaging Technology is extremely efficient in predicting and distinguishing hybrid seeds and their progenitors, even with seeds of great phenotypic variability. The technology is indeed consistent in its prediction, removing subjectivity caused by the intense labor of conventional seed testing methods.

To read the full article, you can download Dr. Michelon’s paper here.