Microsoft Research Asia has come up with the new computer vision system which is more beneficial to botanists to identify thousands of species of flowers with nothing but a picture/snap.Image analysis work done by Rui was a perfect match for the botanists who were trying to find out how to sort millions of publicly submitted images of local flowers.
The new project by Microsoft i.e.,smart flower recognition system came into existence from an idea about pairing Microsoft Research's image recognition technology with a huge database of flower images. Because of the existence of millions of flower species with a huge database would be an advantage to the botanists who otherwise may struggle to identify the exact type of flower they encounter.
The image database was contained with 2.6 million floral photographs uploaded by public. The researchers started by teaching the smart system to recognize the images and group the ‘bad’ ones. The final result was about 800,000 flower photos were paired with the Caffe deep-learning network.
The smart recognition system enables domain experts to collect plant distribution in China in an efficient way and this system can help ordinary people to gain more knowledge.
The flower-recognition engine can recognize the type of flower with an accuracy rate of 90+ percent. As reported by Microsoft , developers may soon come up with an application based on this smart flower recognition system, some of which, it seems, will be targeted at the everyday public as well as field experts.