Assessing fabric stain release with a GPU implementation of statistical snakes

Abstract

Stain release is the degree to which a stained substrate approaches its original unsoiled appearance as a result of care procedure. Stain release has a significant impact on the pricing of the fabric and, hence, needs to be quantified in an objective manner. In this paper, an automatic approach for the objective assessment of fabric stain release that utilizes region-based statistical snakes, is presented. This deformable contour approach employs a pressure energy term in the parametric snake model in conjunction with statistical information (hence, statistical snakes) extracted from the image to segment the stain and subsequently assign a stain release grade. This algorithm has been parallelized on a General Purpose Graphical Processing Unit (GPGPU) for accelerated and simultaneous segmentation of multiple stains on a fabric. The computational power of the GPGPU is attributed to its hardware and software architecture, which enables multiple and identical snake kernels to be processed in parallel on several streaming processors. The detection and segmentation results of this machine vision scheme are illustrated as part of the validation study. These results establish the efficacy of the proposed approach in producing accurate results in a repeatable manner. In addition, this paper presents a comparison between the benchmarking results for the algorithm on the CPU and the GPGPU.

Graphical processing unit–based machine vision system for simultaneous measurement of shrinkage and soil release in fabrics

“…If the stain size was either too small or too large with respect to the fabric image background, then the system failed to report the presence of a stain. The latest work in the arena of soil release is that conducted by Kamalakannan et al, 9 wherein energy minimizing curves, termed as statistical snakes 10 are customized to segment the stains in the fabric images. Using a sizeable data set, it was demonstrated that there was an improvement with regard to the stain segmentation accuracy as compared to that in Refs.…”

“…To that end, it should be mentioned that the recent emergence of fast and specialized computing resources have popularized machine vision schemes that operate on digitized fabric images and produce objective measurements. [3][4][5][6][7][8][9] With regard to soil release, one of the first attempts to detect oily stains was performed by Shin et al, 3 wherein a texture-based mask was used to detect the stain. Chen et al, 4 proposed the use of a backpropagation neural network to inspect fabric defects, such as lack of yarns and oily stains.…”

“…The work in Ref. 9 necessitates the user to provide a reasonable initialization regarding the statistics of the region of interest by clicking within the stain. The second shortcoming relates to the issue of optimally setting the parameters in the snake model.…”

“…The locations of the stain centers, in turn, are used to create seed regions for initializing the statistical snakes presented in Ref. 9. In addition, the algorithm in Ref.…”

“…The problem of parameter setting is tackled by incorporating a statistical procedure in Ref. 9, which elicits the optimal parameter value based on the distributions of the stain and the fabric background. Furthermore, this scheme is incorporated into the snake evolution process, which results in the adaptive evolution of the snake parameters.…”