The field of quantitative morphology has long been important in biological investigations. Various elements of an organism's morphology, such as size and shape, are easily quantified, and standard methods for the analysis of these components exist (i.e., geometric morphometrics). However, methods for reliably quantifying textures and patterns are currently lacking. In this paper we propose a technique for quantifying grayscale images of biological textures and patterns. With our method, the textural properties of an image are represented as a foot pattern of 2-dimensional Cartesian coordinates, obtained via an evolutionary algorithm that minimizes pattern entropy. The pixels of the foot pattern are then assigned labels using one of two techniques: complete enumeration, or by minimizing the differences between sets of landmarks (using a heuristic search for the optimal assignment). The labeled landmark coordinates are then treated as input data for standard quantitative morphometric analysis. With this approach we were able to statistically distinguish between foot patterns generated from two different textural images drawn from the backs of salamanders. Thus, morphological textures and patterns may be quantified, and sets of textures statistically compared.