Title :
Convolution operators as a basis for objective correlates of texture perception
Author :
Benke, K.K. ; Skinner, D.R. ; Woodruff, C.J.
Author_Institution :
Mater. Res. Labs., Australian Dept. of Defence, Ascot Vale, Vic., Australia
Abstract :
A method is described for deriving, from digitized images, objective measures that correlate strongly with simple perceptual judgements on the same images. Each measure is the normalized variance of an image obtained by convolving the original image with a specific local operator. This operator is designed to optimize the correlation between the particular percept and the objective measure, subject to certain specified constraints
Keywords :
pattern recognition; convolution operators; digitized images; objective correlates; objective measures; pattern recognition; perceptual judgements; texture perception; Convolution; Digital images; Graphics; Humans; Image edge detection; Image processing; Layout; Noise shaping; Pattern recognition; Psychology;
Journal_Title :
Systems, Man and Cybernetics, IEEE Transactions on
