A Fractal Dimension Estimator For Multispectral Images

We propose a fractal dimension estimator for multispectral images. The estimator is based on the probabilistic box-counting classical approach extended to the multivariate domain of multiband images. In order to validate the estimator, we propose a fractal model for multispectral images based on the fractional Brownian motion. The generated synthetic multispectral fractal images have 7 statistically-independent spectral bands at specific wavelengths in the visible domain. We apply the proposed approach for the complexity characterization of a widely-known remotely-sensed data set and conclude the article.