Extraction of leaf vein based on improved Sobel algorithm and hue information

Leaf vein contains important physiological information of the plant, and vein extraction is a key step of plant modeling and identification. Inspired by the previous algorithm of extracting vein in HSI color space, a method that combines the improved Sobel algorithm and hue information is proposed to extract leaf vein in this paper. The former is used to extract the basic contour of leaf vein, and the latter is used to extract the primary vein. Then, the results of the previous two steps are combined. The experimental results show that the algorithm can make up for the deficiencies by using Sobel algorithm and HSI color space to extract leaf vein, and it is capable of accurately extracting the leaf vein. Especially, the leaves which have different colors or nonuniform illumination also can be extracted the leaf vein effectively.