StrucDiv

The StrucDiv package provides methods to quantify spatial structural diversity, hereafter structural diversity, in raster data. Raster data means data in gridded field format. The methods consider the spatial arrangement of pixels as pairs, based on Haralick, Shanmugam, and Dinstein (1973). Pixels are considered as pairs at user-specified distances and angles. Distance refers to the distance between the two pixels that are considered a pair. Angle refers to the angle at which two pixels are considered a pair. Angles can be horizontal or vertical direction, or the diagonals at 45° or 135°. the direction-invariant version considers all angles at the same time. The frequencies of pixel pair occurrences are normalized by the total number of pixel pairs, which returns the gray level co-occurrence matrix (GLCM). The total number of pixel pairs depends on the extent of the area under consideration, i.e. on the spatial scale. the spatial scale is defined by the window side length (WSL) of a moving window. In each GLCM, pixel values can be replaced with ranks. Diversity metrics are calculated on every element of the GLCM, and their sum is assigned to the center pixel of the moving window and represents spatial structural diversity of the window. The output map is called a ‘structural diversity map.’ Diversity metrics include common second-order texture metrics: contrast, dissimilarity, homogeneity and entropy (i.e. Shannon entropy). Additionally, structural diversity entropy includes a difference weight δ ∈ {0, 1, 2}, which weighs the difference between pixel values vi and vj, either by absolute, or by squared differences. When δ = 0, structural diversity entropy corresponds to Shannon entropy. Additionally, normalized entropy is available. Normalized entropy is Shannon entropy normalized over maximum entropy, which depends on the spatial scale. These methods can be applied to any continuous data in raster format, and also to categorical data if categories are numbered in a meaningful way, or if entropy or normalized entropy are used. For entropy, normalized entropy, and homogeneity, high numbers of gray levels lead to structureless diversity maps. With these methods, structural diversity features can be detected. Structural diversity features have also been called latent landscape features.

Installation

You can install the released version of StrucDiv from CRAN with:

install.packages("StrucDiv")

Example

Calculate normalized entropy on Normalized Difference Vegetation Index (NDVI) data, which was binned to 15 gray levels (see data documentation). We define the size of the moving window with WSL three, and we consider distance one between pixels (direct neighbors), and all four possible directions in which pixels can be considered as pairs.

entNorm <- StrucDiv(ndvi.15gl, wsl = 5, dist = 1, angle = "all", fun = entropyNorm, na.handling = na.pass)

Haralick, R. M., K. Shanmugam, and I. Dinstein. 1973. “Textural Features for Image Classification.” IEEE Transactions on Systems, Man, and Cybernetics SMC-3 (6): 610–21. https://doi.org/10.1109/TSMC.1973.4309314.