Resonon | Application of Resonon Pika XC2

Application of Resonon Pika XC2 on Differentiating Cultivars, Growth Stages, 

Flowers, and Leaves of Industrial Hemp (Cannabis sativa L.)

As an emerging cash crop, industrial hemp (Cannabis sativa L.) grown for cannabidiol (CBD) has spurred a surge of interest in the United States. Cultivar selection and harvest timing are important to produce CBD hemp profitably and avoid economic loss resulting from the tetrahydrocannabinol (THC) concentration in the crop exceeding regulatory limits. Hence there is a need for differentiating CBD hemp cultivars and growth stages to aid in cultivar and genotype selection and optimization of harvest timing. Current methods that rely on visual assessment of plant phenotypes and chemical procedures are limited because of its subjective and destructive nature. Optical sensing technology, which interrogates biological materials non-destructively, is considered an attractive means for addressing the shortcomings of human inspection and analytical testing. Numerous studies have been conducted on using spectroscopic techniques for cultivar/variety differentiation of plants and agricultural products. Spectroscopic sensing, however, only measures small portions of plant tissues and often requires sample treatments (e.g., drying and grinding) and direct contact between samples and the detector. Hyperspectral imaging is a power modality for measuring spectral and spatial information of samples simultaneously. Compared to spectroscopic techniques that are used for point measurements, hyperspectral imaging is advantageous in delivering reliable and comprehensive analysis of characteristics or properties of plant materials with minimal sample preparation, requiring no sample contact, and thus is potentially more suitable for high-throughput, on-site testing. However, so far, no research has been carried out on using hyperspectral imaging for classifying for cultivars and growth stages of CBD hemp.

Based on this, in the attached article “Hyperspectral Imaging With Machine Learning to Differentiate Cultivars, Growth Stages, Flowers, and Leaves of Industrial Hemp (Cannabis sativa L.)”, a group of scientists from Mississippi State University and North Carolina State University acquired hyperspectral images from freshly harvested leaf and flower materials of five cultivars of CBD hemp at five growth stages using a benchtop hyperspectral reflectance imaging system (Pika XC2, Resonon Inc., Bozeman, MT, United States), developed an image processing pipeline for segmenting the plant parts from background and extracting spectra from sample segments, performed exploratory analysis of spectral features of hemp samples, and built classification models to differentiate the cultivars, growth stages, and plant parts. Aiming to present a proof-of-concept validation of a novel hyperspectral imaging-based approach for non-destructive, fast, and objective differentiation of cultivars, growth stages and plant organs (i.e., leaves and flowers) of CBD hemp.

Hemp plants grown in a greenhouse (left), and flower and leaf for sampling (right).

Schematic (left) and photograph (right) of a hyperspectral imaging system for acquiring images from hemp samples.

【Results】:

(Top) Mean spectra of five cultivars (i.e., BX, TJ, CW, FL58, and FL70) of hemp flowers and leaves harvested 4 weeks after flower initiation and (Bottom) mean spectra of the cultivar Cherry Wine harvested at all the five growth stages.

Classification accuracies in differentiating hemp cultivars based on the samples at each growth stage.

Classification accuracies in differentiating plant growth stages (left) for each hemp cultivar (i.e., BX, TJ, CW, FL58, and FL70) and plant organs (right) 

at each growth stage. 100% accuracy is obtained in all the modeling scenarios with zero standard deviation for 30 replications.

 【conclusion】

The authors concluded that the spectral profiles and PC score scatter plots of hemp samples, to a varying degree, revealed the separation among the hemp cultivars, growth stages and plant organs. The rLDA models, using leaf or flower samples at individual growth stages, achieved the classification accuracies of 96.8%-99.6% in the differentiation of hemp cultivars. Pooling leaf and flower samples at all growth stages resulted in deteriorated accuracies compared to modeling samples at individual growth stages. Both growth stages and plant organs need to be factored in model development for hemp cultivar classification. In contrast, in the differentiation of growth stages and plant organs, the rLDA models achieved 100% accuracies consistently. This study shows that hyperspectral imaging can be used for non-destructive and accurate differentiation between hemp cultivars, growth stages and plant organs, and it is a potentially valuable tool for phenotyping, cultivar selection and optimization of harvest timing in CBD hemp production. Extensive research is still needed to develop and deploy hyperspectral imaging technology for field-scale, in-situ applications.