When we wander around fields (for instance, those around the campus), we can find a number of interesting flowers, but we rarely know their names. We may consult a manual guide on flowers but even if such a guide is on hand, it is time consuming and not an easy task to find a proper section or an exact page showing the flower. It would be useful if we direct a camera, such as a smartphone camera, and then an algorithm will recognize the flower and tell its name. Once we know the name we can find more information about a flower on the web.
This motivates our project. We have created a visual recognition system for automatic flower classification from images. Given an image and only the knowledge that it contains a flower, the system determines the species of that flower. This is a challenging problem due to several factors, including changes in imaging conditions, variations within a species, large similarity across different species, and natural flower deformations.
Project Description
