Two laboratory-based multispectral imaging systems were developed. The image features from color (red, green,and blue) and multispectral bands (550, 710, and 810 nm) were evaluated in predicting chlorophyll and nitrate contents of potatoleaves grown in the greenhouse. Histogram-based image features were extracted from normalized r and g and multispectralband images. The normalized r feature ( r average energy) was found to be negatively correlated (R2 = 0.90) withchlorophyll contents of leaves. Multispectral band image features were found to be linearly correlated (R2 = 0.95) withchlorophyll contents. Average prediction accuracies varied from 90.5% to 92.3% for normalized color and multispectral bandimage features. Average and maximum prediction accuracies were 75.1% and 99.0%, respectively, for petiole nitrate predictionusing normalized r image feature. The multiple linear regression between multispectral band features and nitrate showed alinear relationship (R2 = 0.84). However, average and maximum prediction accuracies were 28.7% and 93.4%, respectively,for predicting petiole nitrate using multispectral band image features.
No takes yet. Share an insight, caveat, or question.
Borhan et al. (2004) studied this question.