An image processing method for investigating the morphology of cereal endosperm cells

Abstract

ABSTRACTIn cereal seeds, the number, morphology and development of endosperm cells are closely related to grain quality, weight and yield. Endosperm cells differ morphologically in different regions of the seed. Nevertheless, it is important to be able to analyze the morphology of cereal endosperm cells. We established an image processing method to enhance the outlines of endosperm cells. The endosperm cell wall was traced precisely using the “pen tool” in Photoshop software (PS). The tracing was defined as the “work path” and was highlighted using the PS “brush tool.” Images of mature rice, maize and wheat endosperm sections stained with different methods were analyzed using this method. Combined with the whole sections of mature and developing cereal kernels, the processed image exhibited clearly the morphology of endosperm cells in any region of endosperm and at any stage of endosperm development. The processed image was more accurate and efficient for analyzing morphological characteristics than the unprocessed image.