Automatic cell identification and counting of leaf epidermis for plant phenotyping

Abstract

The counting of leaf epidermal cells is useful to study the plant developmental changes produced by environmental or genetic factors. The scanning electron microscopy can be used, but it is expensive and time-consuming. Methods using optical microscopy are also available, but they still require leaves pre-treatment and manual cell identification. We propose a quick and simple method for counting leaf epidermal cells without leaf treatments and based on automated cell identification and marking. It allows to highly improve the representativeness of leaf epidermis screening, aiming at a high-throughput plant phenotyping approach. • The leaves are pressed between two glass slides without any pre-treatment and digital micrographs are acquired under incident light. • Epidermal cells are automatically identified and counted by means of a “macro" of ImageJ • The cell count obtained applying the procedure of image processing is very close to that obtainable by manual cell identification.