Measurement of Environmentally Influenced Variations in Anthocyanin Accumulations in Brassica rapa subsp. Chinensis (Bok Choy) Using Hyperspectral Imaging

Abstract

Dietary supplements of anthocyanin-rich vegetables have been known to increase potential health benefits for humans. The optimization of environmental conditions to increase the level of anthocyanin accumulations in vegetables during the cultivation periods is particularly important in terms of the improvement of agricultural values in the indoor farm using artificial light and climate controlling systems. This study reports on the measurement of variations in anthocyanin accumulations in leaf tissues of four different cultivars in Brassica rapa var. chinensis (bok choy) grown under the different environmental conditions of the indoor farm using hyperspectral imaging. Anthocyanin accumulations estimated by hyperspectral imaging were compared with the measured anthocyanin accumulation obtained by destructive analysis. Between hyperspectral imaging and destructive analysis values, no significant differences in anthocyanin accumulation were observed across four bok choy cultivars grown under the anthocyanin stimulation environmental condition, whereas the estimated anthocyanin accumulations displayed cultivar-dependent significant differences, suggesting that hyperspectral imaging can be employed to measure variations in anthocyanin accumulations of different bok choy cultivars. Increased accumulation of anthocyanin under the stimulation condition for anthocyanin accumulation was observed in “purple magic” and “red stem” by both hyperspectral imaging and destructive analysis. In the different growth stages, no significant differences in anthocyanin accumulation were found in each cultivar by both hyperspectral imaging and destructive analysis. These results suggest that hyperspectral imaging can provide comparable analytic capability with destructive analysis to measure variations in anthocyanin accumulation that occurred under the different light and temperature conditions of the indoor farm. Leaf image analysis measuring the percentage of purple color area in the total leaf area displayed successful classification of anthocyanin accumulation in four bok choy cultivars in comparison to hyperspectral imaging and destructive analysis, but it also showed limitation to reflect the level of color saturation caused by anthocyanin accumulation under different environmental conditions in “red stem,” “white stem,” and “green stem.” Finally, our hyperspectral imaging system was modified to be applied onto the high-throughput plant phenotyping system, and its test to analyze the variation of anthocyanin accumulation in four cultivars showed comparable results with the result of the destructive analysis.

Image-based phenotyping to estimate anthocyanin concentrations in lettuce

“…There are many successful examples of anthocyanin predictions based on hyperspectral imaging combined with machine learning techniques ( Chen et al., 2015 ; Askey et al., 2019 ; Simko, 2020 ; Cho et al., 2021 ; Kim et al., 2021 ). However, these machine learning models require hyperspectral imaging systems with similar wavelengths as used to develop the machine learning model.…”

Hyperspectral Imaging-Based Multiple Predicting Models for Functional Component Contents in Brassica juncea

“…Notably, among the ten pre-processing combinations tested, the 6th combination registered R V 2 and RMSEP values of 0.8273 and 2.4277, respectively, indicating the highest level of prediction performance. This may be because the distribution of anthocyanins, referred to as 'purple magic' by Kim et al [37], accounted for the largest purple areas over the plant body. When dark-red areas on B. juncea leaves were visibly greater, the anthocyanin content was higher, which may have influenced the spectrum extracted from hyperspectral images.…”

Non-destructive prediction of rosmarinic acid content in basil plants using a portable hyperspectral imaging system and ensemble learning algorithms

“…However, extracting useful plant traits from the complex and vast HSI data requires various approaches, including vegetation indices, multiple linear regression, partial least square regression (PLSR), principal component regression, machine learning algorithms, and radiative transfer modeling [9]. Vegetation indices have been developed for rapid monitoring of phytochemicals and nutrition status, such as chlorophyll content in corn leaves [10], potassium content in rice leaves [11], and anthocyanin in bok choy leaves [12]. PLSR is a commonly used regression method for predicting metabolite content from HSI data.…”