We describe a software signal and image processing laboratory, DataLab-J, which has proved effective in practical operational use. We require: a research environment, a fully operational data analysis system, and a pedagogical tool. The system must be easily extendable and provide a realistic platform upon which novel algorithms can be implemented. On a further dimension, we require the system to handle seamlessly and efficiently three broad data types: digital signals (sequences), images (possibly multiband) and multivariate data sets. The system is implemented in the programming language Java™. DataLab-J has been operational for four years as a platform for many research projects within a university signal and image processing research group.

Image processing, signal processing and computer vision are increasing in importance, and are indeed slowly being considered as core competences in computer science.The specification of computing curricula by the ACM and the IEEE Computer Society states that technical advances over the past decade have increased the importance of topics such as graphics and multimedia, and considers as a core topic Graphics and Visual Computing. With the advent of courses in digital media in electrical engineering and computer science, the increasing importance of computer imaging and vision is manifest, and not only in electrical engineering and computer science departments.The need for development and training of research students is more often than not badly catered for at the present time in Irish universities (Republic, Northern Ireland). Research students are often from diverse discipline and educational system backgrounds. A characteristic of a dynamic and effective graduate and research student environment is not just the raw numbers of PhDs produced, but also relative consensus on the part of the "community" on the most effective major directions of innovation and of focus. On any such qualitative characteristic, Irish universities, institutes of technology and research institutes are not performing well.Successful models for such community-strengthening include doctoral networks at European level, and summer schools at national or regional level.This paper addresses specific issues: (i) the means by which the needs for graduate research level community-strengthening in the areas of signal processing, and image processing and computer vision, can be satisfied; (ii) what the most crucial elements of these fields are, i.e. proposed key areas and curricula.

We describe the choice and assessment of neural network and statistical methods for data modelling, feature selection and forecasting. We deal in particular with how empirical environmental and Earth observation data can be used in conjunction with physical simulation models.

Addressing the problem of automatic fault detection in woven and dyed fabric, we discuss a number of new statistical model-based methods and relate them to a first stage of point/local detection and a second stage of extended pattern detection. One model-based method defines a maximum likelihood binarization of the image. In another model-based method, we describe a discrete Fourier transform-based texture analysis technique that is highly effective for woven textiles in discriminating subtle flaw patterns from the pronounced background of repetitive weaving pattern and random clutter. Finally, we describe a model-based clustering method that can be employed to aggregate perceptual groupings of point and local detections. © 1999 John Wiley & Sons, Inc. Int J Imaging Syst Technol,

Automatic inspection of woven textile fabric is discussed. A two-stage detection process is adopted, with the second stage involving set of novel contextual decision fusion techniques. Three significant problems are addressed: (1) texture feature extraction: Fourier transform features are found to be well matched to the spatially periodic nature of the woven pattern; (2) detection of localized flaw patterns: since prior probabilities are impossible to estimate, and we cannot hope to enumerate all defect classes, a Neyman-Pearson approach is adopted, i.e., flaw detection is via measured deviation from nominal; and (3) detection of extended flaw patterns: the most common flaws are characterized by linear or other cluster shaped patterns; although these are weakly detectable by local detectors, they may be ignored when local detector sensitivity is set to achieve tolerably low false-alarm rates; a local-extended contextual decision fusion technique using morphological filtering enables us to achieve very low composite false-alarm rate. The performance of the system is evaluated on samples of denim fabric containing real defects. The predicted composite false-alarm rate is of the order 1 in 10(13), Or equivalent to 1 per 100 km of fabric roll. Experimental results demonstrate the compatibility of this favorable false-alarm rate with the reliable detection of flaws, which have been chosen for their subtlety and detection difficulty. (C) 1998 Society of Photo-Optical instrumentation Engineers.

We combine image-processing techniques with a powerful new statistical technique to detect linear pattern production faults in woven textiles. Our approach detects a linear pattern in preprocessed images via model-based clustering. It employs an approximate Bayes factor which provides a criterion for assessing the evidence for the presence of a defect. The model used in experimentation is a (possibly highly elliptical) Gaussian cloud superimposed on Poisson clutter. Results are shown for some representative examples, and contrasted with a Hough transform. Software for the statistical modeling is available.

Automatic inspection of woven textile fabric is discussed. Two signiicant problems are addressed: (a) texture feature extraction; Fourier transform features are found to be well matched to the spatially periodic nature of the woven pattern; moreover , they are shift invariant; (b) detector design; maximum likelihood type detectors are inappropriate since we cannot hope to enumerate all defect classes, hence a Neyman-Pearson approach is adopted, i.e. aw detection is via measured deviation from nominal. The performance of the system is evaluated on samples of denim fabric containing real defects.