蝴蝶兰盆花分级系统及花朵面积提取试验

Potted Phalaenopsis Grading System and Flower AreaExtraction Experiments

为实现蝴蝶兰盆花快速无损分级,以“大辣椒(Big Chili)”蝴蝶兰盆花为对象,研究利用颜色模板匹配算法实现蝴蝶兰花朵图像分割及斑点提取算法获得花朵面积的方法,并通过不同视角下的花朵面积提取试验,分析视角不同所引起的花朵面积差异。结果表明:以蝴蝶兰花朵颜色为模板的颜色模板匹配算法能快速实现蝴蝶兰花朵图像的有效分割。多视角下的蝴蝶兰盆花图像处理结果表明,不同视角下的花朵面积存在0.7%~64.8%的差异。综上所述,为实现蝴蝶兰盆花分级,应旋转并拍摄多视角图像获得比较全面的花朵状态信息,继而根据花朵颜色模板匹配算法提取蝴蝶兰花朵面积,作为评判其花朵开放程度和分级的准确依据。

Using the"Big Chili"potted Phalaenopsis as material,Color Matching Template Algorithm was applied to make the effective segmentation of Phalaenopsis flower,then the next step,Blob Extraction Algorithm was employed to extract the area of Phalaenopsis flower.In the last,it discussed the area extraction results of the same potted Phalaenopsis sample of the different points of view.These Algorithms will provide a fast and non-destructive grading method for potted Phalaenopsis.The results show that the Color Matching Template Algorithm of Phalaenopsis flower can realize the effective segmentation of the Phalaenopsis flower image.The area of flowers from different perspectives of the same potted Phalaenopsis varies from 0.7%to 64.8%.To sum up,in order to accurately evaluate the blooming stage of Phalaenopsis,the appropriate shooting angle should be acquired so that the image can truthfully reflect the real flowering state of the flowers.Then,the area of Phalaenopsis flowers should be extracted according to the flower color matching template algorithm.The results should be served as accurate data for estimating the quality of the flowers.