Special areas detection on agricultural fields images using evaluations of local brightness variability

Abstract

Problems of practical using of aerial photograph processing methods based on textural and fractal characteristics of images are considered in the paper (on an example of agricultural field images). ASM, Contrast and Entropy are used as textural characteristics. Calculation of fractal signature is based on fractal dimension. Results of joint segmentation based on proposed methods are tested at processing potato field aerial photographs. Also computational speedup problem using MPI are considered.