Combination of Color, Shape and Texture Features for Orchid Classification

Nurbaity Sabri, M. F. Kamarudin, Raseeda Hamzah, Nur Nabilah Abu Manghsor, Khyrina Airin Fariza Abu Samah, Noraini Hasan

2019 · 11 citations · 9 references

Concepts

Abstract

Orchids is flowering plants that widely evolved worldwide, some of them have the same appearance, colour, and shape. Sometimes florists have difficulty to identify types of orchids due to the same appearance. Botanist also had spent a lot of time to examine each of the orchid types and classify them into a group or orchid family. An image processing technique has been implemented for identifying the type of orchids. The region of interest (ROI) orchid image is extracted using a combination of shape, color and texture feature. From the selected features, Support Vector Machine (SVM) used to classify those features into three (3) types of orchid species named Dendrobium, Phalaenopsis and Vanda. Result show shows an accuracy rate of 82.22%. More experiment using other features to increase the accuracy of classification rate will be done. Besides, more dataset will be used for future work.

References

9