Using SVM to Build Grading Phalaneopsis Model by their Geometric Characteristics of Leaf and Flower
14:30:00 - 15:00:00
308 , Mathematics Research Center Building (ori. New Math. Bldg.)
Flowering Phalaenopsis orchids and related genera are one of the most valuable potted floriculture crops produced throughout the world. It is one of the most important export flowers in Taiwan. (Exporting mature seeding without spike to fit the need of target market is a higher policy for growers.) According to previous study (Lee and Wang, 1997), spiking rate was positively correlated with leaf area in Phalaneopsis. Instead of correlating leaf area with spiking rate, Tsai (2013) used canonical correlation to show significant correlations between leaf traits and flower traits. Furthermore, Tsai (2013) showed that the second and third leaf traits of tested varieties had the highest correlation with flowering traits. The leaf traits of tested varieties will be measured using algorithm developed in Huang (2000). But it lacks reliable algorithm to predict the quality of flowers. Since support vector machine (SVM) is one of the most often used pattern recognition algorithm, we propose to use SVM to build a classifier based on the second and third leaf traits of tested varieties for grading the phalaneopsis leafs. Based on this grading result, it is expected that it can be used to predict the quality of flowers for every phalaneopsis during big plant stage. Preliminary study results will be presented in this talk.