Using Neural Network to Model the Relationship between Plant Surface Color and its Pigment

摘要

Combining the intelligent algorithm such as BP neural network and support vector maching (SVM) with traditional chemical method, this paper models the relationship between plant surface color and its pigment. Using the neural network model constructed above, people can figure out the content of plant pigments by getting the corresponding plant surface color information. Compared with the traditional modeling methods, this method can significantly save time and experimental supplies. Furthermore, it is easy to implement because it needn't touch samples and doesn't cause any damage to samples. So this method provides a practical tool for non-destructive measurements of plant pigments and solutions to explore the mystery of plant color.