In-situ leaf-surface micro-topography - the case of stomata detection

Understanding the interaction of leaf stomatal state and environmental conditions is one prerequisite to predict leaf photosynthesis and transpiration rates at different spatial and temporal scales, which adds to the implementation of smarter climate control algorithms in greenhouses. Non-invasively capturing data of stomatal density, size and aperture in living plants comes along with several challenging problems. The equipment for image acquisition needs to be applicable under field or greenhouse conditions as one should be able to acquire in-situ microscopic images with a minimum magnification of about ×300. This usually requires extension tubes. In situ acquired microscopic leaf images tend to be out of focus due to leaf movements and a limited depth of field. Therefore, several focus layers are usually acquired, with only partial areas being in focus in each image. An easy to handle and mobile photo-microscope was assembled and a computation method was developed to obtain microscopic images from leaves of living hibiscus plants using the focus stacking technique in the greenhouse. Based on MATLAB, a convenient method for stomatal segmentation and density estimation using an aligned and all in focus image is presented.