perClass Mira use-cases

Phenotyping characterization for seeds

Plant phenotyping optimizes crop quality or robustness by selecting best genotypes. Spectral imaging offers a unique non-invasive method to capture qualitative parameters of plants, seeds and fruit. perClass Mira enables plant breeders and biologists to quickly build and deploy custom phenotyping work-flows. In this case-study we illustrate an application on seed growth monitoring. This is important to assess the speed of growth and extract relevant parameters.

In this case study we analyse seeds behaviour due to several factors. We look at water stress level and its influence on seeds growth. The data was acquired with different spectral cameras in the VISNIR range. Our objective is three-fold:

1) Classify root and bean for growth characterization.

2) Compute relevant custom spectral indices and export their features.

3) Compare the spectral indices computed with hyperspectral and multispectral sensors

In the white-paper, we illustrates that hyperspectral imaging allows the user to extract different phenotyping parameters to compare quality of seeds. This method can also be used to estimate qualitative parameters such as nitrogen or chlorophyll content in plants. perClass Mira is an easy tool allowing users to extract such parameters of interest from a large number of scans typically needed in the phenotyping studies. The process can also be automated using the perClass Mira batch processing tool. From a grid defined by regions and a template, we can export results in Excel or XML and get results from hundreds scans in an automated way.

Request the whitepaper at