Statistical Approach for Texture Segmentation (An Implementation on a Parallel Inspection System)

Abstract

In this study, Detection of structural defects in textiles is addressed as a specific problem of visual texture analysis and segmentation. An overview of the importance of textile industry in Egypt is demonstrated with specific consideration for the inspection process, problems with current manual inspection systems and the importance of introducing an automatic inspection system arc addressed. In this thesis a parallel automatic textile inspection system was designed and implemented which includes the hardware part (parallel machine, cameras, lighting technique ...etc) in addition to the algorithms and software part.

Interest in digital applications areas: interpretation, and perception.

unagc processing methods stems from two principle improvement of pictorial information for human processing for scene data for autonomous machine

The first category deals with problems like image transnusston, coding, enhancement, blurring...etc, while the second category solves problems dealing with machine perception. In this case, interest focuses on procedures for extracting information from image in a form suitable for computer processing. Often, this information bears little resemblance to visual features that human­ beings usc in inteqHeting the content of an image. Examples of the type of infotmation used in machine perception arc statistical moments, Fourier transform coefficients, and multidimensional distance measures.