Overlapping leaves segmentation method by using hybrid of Chan-Vese model and morphological operators

In the nature, almost all the leaves were overlapping with other leaves. Separating a leaf from another is a step in deeply analyzing each leaf, for example leaf health analysis. Therefore, the image segmentation algorithm on overlapping leaves is needed to separate the target leaf from other leaves automatically. For this reason, this research proposes the overlapping leaves segmentation method by using the Chan-Vese model and the morphological operations. First, Chan-Vese model is applied for image segmentation by minimizing an energy functional for controlling the curve deformation movement and the evolution of the contour curve. Therefore, several morphological operators are used to improve the performance of the Chan-Vase method. This proposed method uses 3 operators which are the opening, dilation and erosion operators. The morphological operators are used for removing the small object and adjusting the result images size to the original image. Four images of natural leaves are used to evaluate the performance of the proposed method. The experimental results show that the proposed method is more accurate than Distance Regularized Level Set Evolution (DRLSE) method especially for the overlapping leaves.