An Objective Fabric Smoothness Assessment Method Based on a Multi-Scale Spatial Masking Model

Wang, Jingan; Shi, Kangjun; Wang, Lei; Li, Zhengxin; Pan, Ruru; Gao, Weidong

doi:10.1109/access.2019.2918845

Cited by 10 publications

(22 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In our previous studies [24,29,31], we put forward the following point: although the observation of the 2D method is indirect, they can still provide sufficient information for the smoothness evaluation of the fabric. we have established a 2D fabric image acquisition system and optimized the illumination environment for fabric smoothness information extraction [24], and the view was verified by the experiments in the studies.…”

Section: A Image Acquisitionmentioning

confidence: 99%

“…Length, surface area, volume under the surface, mean principle curvatures, mean max twist of every sub-block of the depth map [13]; arithmetic average roughness, root mean square roughness, 10-point height, bearing surface ratio, wrinkle sharpness, wrinkle density [14,22]; fractal dimension [20]; maximum amplitude, sharpness, density, maximum amplitude of the first derivative of the cross profile of the edges [15,21]; mean, mean deviation, and standard deviation of the height values in every row [26]; wrinkling density, wrinkling hardness, tip-angle, wrinkling roughness [27]; dense-SIFT feature with sparse coding [4]. In our previous study, we extract the image features by a visual masking model [29] for human visual system. The experiment in the study proved that this feature has a strong ability to characterize the smoothness appearance of the 2D fabric images.…”

Section: B Hand-crafted Featuresmentioning

confidence: 99%

“…The smoothness appearance of fabrics is a subjective concept, which is mainly concerned by the perception of the human vision under a specific environment. A feature set named multi-scale spatial masking feature (MS-SMF) [29] is proposed with the consideration of the relationship between the HVS characteristics and the fabric smoothness appearance in images. As it was verified that the MS-SMF has a superior ability to describe the fabric smoothness appearance than other hand-crafted features used in previous studies, the MS-SMF is used as the hand-crafted features used in the proposed image classification framework.…”

Section: A Hand-crafted Featuresmentioning

confidence: 99%

“…A recent study for 3D image feature extraction [4] proposes to use the dense-sift features encoded by linear spatial pyramid matching using sparse coding(ScSPM), which simulates the sparse structure of human visual system (HVS) [28]. For the 2D methods, our previous study [29] established a multi-scale spatial masking model to extract hand-crafted features for fabric smoothness appearance, which models the spatial masking effect of the HVS. Differing from the hand-crafted features constructed by human, convolutional neural networks (CNNs) can extract image features by self-driven learning, which are built with reference to the hierarchical information processing structure of the HVS [30].…”

Section: Introductionmentioning

confidence: 99%

“…Combining the two can describe the fabric smoothness appearance evaluation process more comprehensively. As the research basis, our previous studies have proposed a set of effective hand-crafted feature [29] and demonstrated the superiority of the CNN features [31] for fabric smoothness assessment. The experimental prediction results from these two types of features showed a certain complementarity in the sample distribution, which supports our viewpoints.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Fusing Convolutional Neural Network Features With Hand-Crafted Features for Objective Fabric Smoothness Appearance Assessment

et al. 2020

Self Cite

View full text Add to dashboard Cite

In the textile and apparel industry, it remains a challenging task to evaluate the fabric smoothness appearance objectively. In existing studies, with computer vision technology, researchers use the hand-crafted image features and deep convolutional neural network (CNN) based image features to describe the fabric smoothness appearance. This paper presents an image classification framework to evaluate the fabric smoothness appearance degree. The framework contains a feature fusion module to fuse the handcrafted and CNN features to take both advantages of them. The framework uses the multi-scale spatial masking model and a pre-trained CNN to extract hand-crafted and CNN features of fabric images respectively. In addition, a mislabeled sample filtering module is set in the framework, which helps to avoid the negative impact of mislabeled samples in training. In the experiments, the proposed framework achieves 85.2%, 96.1%, and 100% average evaluation accuracies under errors of 0 degree, 0.5 degree, and 1 degree respectively. The experiments on the feature fusion and mislabeled sample filtering verified their effectiveness in improving the evaluation accuracies and the label noise robustness. The proposed method outperforms the state-of-theart methods for fabric smoothness assessment. Promisingly, this paper can provide novel research ideas for image-based fabric smoothness assessment and other similar tasks.

show abstract