Algorithm Design in Leaf Surface Separation by Degree in HSV Color Model and Estimation of Leaf Area by Linear Regression

Abstract

Plant leaves are very important for their respiration and photosynthesis. The two processes are significant factors for their growth. Measuring leave dimension is very important in studying and analyzing the photosynthesis of plants. Leaf dimension assessment with image evaluation is the most widely technique used for presenting. This paper proposed the algorithm of image segmentation to classify image elements and calculate leaf surface with a threshold segmentation technique by using the constant threshold in gray color model and calculating the degree of green color in the HSV models. Segmentation technique is used to separate good surface out of defective surface of leaf image. Moreover, this paper also proposed leaf area estimation with linear regression analysis with the pixel value on the leaf surface. Further to sixty experiments, they showed the accuracy to separate elements of good surface and defective surface are 98.72% and 96.47% respectively.