| In order to engage in photosynthesis, plant leaves absorb CO2) via the opening of pores in their surfaces called stomata. Open stomata, however, result in the evaporation of H2O, which is a detriment to plant function. Thus a particular leaf will seek stomatal apertures through which its need for CO2 is balanced by its aversion to H2O loss. In order to visualize a particular leaf's stomatal aperture, an experimentalist injects the leaf with dye so that it fluoresces when closing its stomata. The regions with a higher relative intensity correspond to areas in which the stomata are closed and the darker regions where the stomata are open. A camera is used to collect the emitted light, and a fluorescence pattern is measured. Images are recorded as these patterns change with time, resulting in a, video sequence. The primary task of this work is to segment these video sequences into fluorescing and non-fluorescing regions. To do this, we propose a 3D segmentation method inspired by the active contours without edges approach of Chan and Vese [15]. The associated partial differential equations are solved within a level-set framework using a three-dimensional semi-implicit numerical scheme. Due to noise in the data, preprocessing is required prior to the segmentation step, and for this we use the PDE based denoising algorithm of [55] with L1-fidelity as proposed by Chan and Esedoglu in [12]. |
