This paper presents a novel approach for defect detection using a wavelet-domain Hidden Markov Tree (HMT)1 model and a level set segmentation technique. The background, which is assumed to contain homogeneous texture, is modeled off-line with HMT. Using this model, a region map of the defect image is produced on-line through likelihood calculations, accumulated in a coarse-to-fine manner in the wavelet domain. As expected, the region map is basically separated into two regions: 1) the defects, and 2) the background. A level-set segmentation technique is then applied to this region map to locate the defects. This approach is tested with images of defective fabric, as well as x-ray images of cotton with trash. The proposed method shows promising preliminary results, suggesting that it may be extended to a more general approach of defect detection.
The overall quality of a fabric is dependent on a number of factors. Among these is the fabric’s tendency to wrinkle after home laundering - referred to as smoothness. Wrinkle grading is a subjective process involving human graders who compare fabric samples to replicas, representing various degrees of wrinkling. This process is also operator dependent, expensive, and it lacks the ability to adequately describe the many subtle differences that exist between grades. Therefore, the textile industry needs an automated system that can describe wrinkles on a fabric surface in an objective and repeatable manner. In this paper, we describe a computer vision system developed in a previous work and examine the effectiveness of new features extracted from the wavelet domain independent mixture model and a landform classification technique. Shown to be useful in texture classification, features from the wavelet domain independent mixture model are measured based on the two-population characteristic of the wavelet domain. The second technique uses topographical analysis methods originally developed for geographical landform classification that have been successfully applied to digital elevation models of the Earth’s surface. These new measurements, representing quantitative descriptions of the surface of a fabric in both the frequency and spatial domains, are compared to the existing industry grading standard using a fuzzy classifier. Results show a good correlation with technicians’ grades.
We describe a strategy for the content-based compression of mammograms. In this two-step strategy, the clinically important structures are first identified via a fractal-based segmentation method. Then, a modified version of JPEG2000 is applied in such a way that lossless compression is applied to the extracted structures from the first step, while a lossy compression is applied to the remaining regions. Preliminary results demonstrate that this strategy can achieve high compression ratios (up to 50:1) without compromising the diagnostic quality of the mammograms.
A vision system for the automatic quantification of fabric geometric distortion has been implemented and tested. The intended utility of this system is to replace the manual measurement of fabric shrinkage or growth as governed by the AATCC (American Association of Textile Chemists and Colorists) Test Method 135. In the near future, other capabilities, such as automatic quantification of fabric smoothness, will also be incorporated. The system uses commercial, off-the-shelf hardware components, together with a customized image processing algorithm to capture digital images of pre-marked fabric swatches and to accurately measure the distance between the benchmarks before and after laundering. The primary focus of this paper is a description of the algorithm that detects these benchmarks. This robust algorithm detects the marks without regard to: (1) changes in the texture or the color of the swatches, (2) inter-fabric changes in the benchmark colors, (3) changes in the fabric contrast due to scanning or laundering, (4) presence of noise, or (5) slight rotations of the swatches during scanning. The presented system has been under routine testing at the International Textile Center of Texas Tech University, as well as the laboratories of Cotton Inc., with the computed dimensional changes and the manual measurements possessing a nearly perfect linear correlation.