Flower Petals Measure Based On Image Processing

Flower is an important performance element of the design art field, plant classification has played a very important role in the landscaping of landscape design,and the petals number of the flower plants provides a very important characteristics for this job, so the study of this paper has great academic value and practical value.Along with the development of information technology, image processing technology has become the indispensable strong tool for scientific research. This paper analyzes the limitations of traditional image segmentation algorithm, and the drawback of existing algorithm is discussed,such as they often can’t get a satisfactory result in the border area. Because the demand for completeness and accuracy of target outline is higher, the article introduces the foreground extraction algorithm solving for corolla region’s segment.This algorithm requires an additional mark map to show constraints information.Currently,there are mainly two ways:trimap and strokes. The former is too time-consuming, and the property of stability or real-time is bad.The latter way is simple, but segmentation result is not very effective. On the basis of studying for a lot of natural flowers images, this paper made a study of the automatic gainning signature information. Putting forward a new way of thinking about automatic generating mark figure according to the color of the features of flower images. Experimental result shows that this method not only effectively reduce the computation time, but also can obtain satisfactory segmentation result. The Serviceability is good for images with a low resolution or blur images aswell.In part of feature calculation, judge flowers whether having petal through the minimum circumscribed circle first. Choosing the algorithm of distance of centre of mass by comparing the pros and cons of three calculation methods by comparing three methods.In order to ensure the correctness of the algorithm, propose the quadric area judgement combining with concave angular point.