. Introduction
Flowers and the ability to recognize them have been attracting humans for over a longer period of time. The
taxonomy of flowers originally contained approximately 8000 plants, but since then it has been extended to
encompass more than 250000 flower species around the world [1]. The flower images captured in a real time will
create a several challenges like variations in viewpoint, scale, illumination, partial occlusions, and multiple
instances. All these challenges need a very refined algorithm to classify flowers. Also, the cluttered background
makes the problem more difficult, as we need to classify the flower image from the background. Moreover, the greatest challenge lies in preserving the intra-class and inter-class variabilities. The classification and grading of
flowers will be the major part in floriculture. For formers, the floriculture has become one of the important
commercial trades in agriculture owing to steady increase in demand of flowers.