http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Image analysis for predicting phenolics in Arabidopsis
( Jayapal Praveen Kumar ),( Rahul Joshi ),( Byoung-kwan Cho ) 한국농업기계학회 2021 한국농업기계학회 학술발표논문집 Vol.26 No.2
Phenolics play a vital role in the development of plants, in specific to the biosynthesis of pigment and lignin. These phenolic compounds are also important to human health due to their antioxidants. They help in managing the blood pressure levels and good blood circulation. Hence, it is important to identify the phenolics compounds in plants in a non-destructive way. In our research, we have taken Arabidopsis thaliana which is the model plant and performed the non-destructive plant phenolic measurement using near-infrared (NIR) imaging and machine learning. Additionally, we have tested the presence of phenolic compound in plants under stress conditions. The different stresses used in our research are four different types of light (White, Red, Blue and Red+Blue). The plant images were captured using the NIR-HSI system with a wavelength of 895-2504 nm. The spectral data are extracted from the images and they are synchronized with HPLC reference values. The partial least square regression (PLSR) model is applied and obtained the model performance with R2 over 0.90 for calibration datasets. The best prediction with R2 value of 0.89 and standard error of prediction (SEP) value of 0.17 is obtained by applying the Savitzky-Golay second derivative preprocessing method.