Automatic grain texture analysis using integral transforms

Abstract

Abstract Tonewood for musical instruments is quarter-sawn and frequently quality-graded based on visual appearance, mechanical and acoustic properties. The assessment uses simple human (subjective) observation, and two “experts” can rate the same sample differently. This paper describes the application of integral transforms (Fourier and Radon) for automatic (objective) assessment of the visual appearance of 10 Sitka spruce (Picea sitchensis) sample images. This work considers surface classification on the basis of grain orientation, count, spacing, and evenness or uniformity.