Structural analysis of textures based on LAW´s filters

Abstract

The field of medical image analysis has grown and these advances have facilitated the development of processes for high-throughput extraction of quantitative features that result in the conversion of images into mineable data and the subsequent analysis of these data for decision support. This paper presents a simple yet efficient image processing approach by proposing a new image feature detector and descriptor based in LAW´s filter. This Algorithm gets texture, shape and layout information; the structures are defined based on a kernels and texture-energy approach that measures the amount of variation within a fixed-size window. The algorithm showed efficiency as a biomarker about the state of human health, vegetables, or any other object animate o inanimate.