Generic Texture Analysis Applied to Newspaper SegmentationPaul

This paper deals with the segmentation of grey-scale newspaper images into the distinct regions of text, picture and background. Feature vectors are obtained from the image by analysing localised textual characteristics and textual variation. Analysis is performed in a generic way making few contextual assumptions. The contrast, size, orientation and resolution of the image are accounted for by a combination of this feature extraction process and subsequent parameterisations. Given such implicit invariances we are able to perform an initial point based classiication through the use of a modiied quadratic neural network. The use of a simple ood ll based algorithm allows the successful segmentation of newspaper images into distinct rectangular regions. Results of newspaper segmentation show the eeectiveness of these methods. Future work will be based on the local processing performed here and its suggested similarity to the behaviour of neurons in the human visual system.