Development and evaluation of a new algorithm to extract morphological features from FESEM images: a case study on rice kernels

The automatic extraction of morphological features from images is of fundamental importance for efficiently processing large datasets.In this research, a novel algorithm was developed to extract diverse morphological features from FESEM images, starting from a specific application on rice kernels.In this context, a dataset of 220 images from 54 distinct rice varieties was used to develop and test the algorithm [1].The aim of this approach is to identify the round-shaped starch granules naturally present in rice kernels from FESEM greyscale images and extract a set of morphological features from them.

Hali tarjima qilinmagan