Decoloration of Multi-Color Fabric Images for Fabric Appearance Smoothness Evaluation by Supervised Image-to-Image Translation

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

doi:10.1109/access.2019.2959705

Cited by 8 publications

(7 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, the study still has the following limitations. (1) Although we have explored the method for multi-color fabric image decoloration in our previous study [25], the performance of the smoothness appearance assessment methods should be further discussed on the decolored multicolor fabric images. (2) In order to further improve the training stability of the proposed method, the fabric image sample size should be further extended.…”

Section: Discussionmentioning

confidence: 99%

“…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. Additionally, to improve the multi-color fabric adaptability of the 2D method, we have proposed a decoloration method for multi-color fabric images in the previous study [25]. In this study, we further explore the improvement of the 2D method, and the fabric image data set is the same as our previous studies captured by the proposed system.…”

Section: A Image Acquisitionmentioning

confidence: 99%

“…The possibility and effectiveness of the objective evaluation for the smoothness appearance of fabric without color pattern on the basis of 2D fabric images was demonstrated. In addition, to extend the variety adaptability of 2D methods, another study [25] has proposed a decoloration method for images of multi-color fabrics, which can eliminate the color pattern in the fabric images while preserve the wrinkles information in images. Therefore, the task of fabric smoothness appearance evaluation after image acquisition can be divided into two main steps, i.e.…”

Section: Introductionmentioning

confidence: 99%

See 2 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

Section: Discussionmentioning

confidence: 99%

Section: A Image Acquisitionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

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

show abstract

“…For this kind of task, supervised methods are superior to unsupervised methods with available paired images to train the model. 16 Even though the present methods can draw into different constraints, the color and style information are difficult to be controlled simultaneously, being unsuitable for the appearance generation task of the colored spun yarn fabric. To involve the constraints from different types, the existing methods need to be extended.…”

Section: Introductionmentioning

confidence: 99%

Appearance generation for colored spun yarn fabric based on conditional image‐to‐image translation

Zhang

Xiang

Wang

et al. 2022

Color Research & Application

Self Cite

View full text Add to dashboard Cite

The repeated trial‐weaving and proofing process are traditionally conducted to exhibit the ever‐changing appearance of the colored spun yarn fabric knitted by different colored spun yarns. In this paper, a novel method of appearance generation for colored spun yarn fabric was proposed based on conditional image‐to‐image translation. The generated image is required to have the same color and the fabric style simultaneously as the color card and the assigned style. The common pix2pix model was modified by adding the category label to the image channel to involve the style constraint. To improve the generation performance, the U‐Net architecture was replaced by the residual block architecture. Four commonly used style fabrics and different color cards were adopted to build the image pairs for experiments. The visual inspection and category similarity were used as the evaluation metrics. Experimental results reveal that the proposed method can transfer different color cards to the designated style fabrics based on the constraints, being effective and superior for image generation of colored spun yarns. The proposed scheme can provide references for the designer by presenting the image perception, saving labor, and material resources in the proofing process.

show abstract

“…8–10 The features extracted by different technologies are selected to analyze the two-dimensional (2D) images or three-dimensional (3D) depth maps to describe the surface structure of the fabrics. 11–15 Automatic detection of fabric wrinkle recovery performance is still in the exploratory stage. There are a few studies on improving the automation of the test method using laser or image processing methods.…”

mentioning

confidence: 99%

Instrumental evaluation of fabric shape retention by image analysis

Wang

Tang

Zhang

et al. 2020

Textile Research Journal

Self Cite

View full text Add to dashboard Cite

Fabric shape retention is an essential property for assessing fabric usability and easy-care properties, which needs to be evaluated frequently for quality improvement. At present, certain aspects of shape retention can be characterized by particular devices, such as the crease recovery tester, the fabric drape tester, etc. To effectively and accurately reflect fabric shape retention performance, we developed an automatic crease forming device to simulate fabric crease creation and the shape recovery process in daily life, and objectively assess the shape retention by an image processing method. A specified size specimen was laid flat on the device to create a sharp crease. Then, a video image of the fabric shape recovery is acquired for measuring the evaluation indexes, such as the vertex angle ( VA), height ( H) and shape retention area ( SA). Finally, the results of this proposed method are compared with existing methods. When compared with the existing crease recovery tester, there is good consistency between the VA of the developed measurement system and the recovery angle of the fabric crease recovery tester, which indicates that the proposed method can be used to evaluate the crease recovery of fabrics. Compared with the drapability, there is linear function relationship between the H and SA of the developed measurement system and the draping coefficient of the fabric drape tester, which demonstrates that the proposed method can be used to evaluate the drapability. Therefore, experimental results indicate that the data calculated by our proposed method can be used to determine fabric shape retention.

show abstract

Decoloration of Multi-Color Fabric Images for Fabric Appearance Smoothness Evaluation by Supervised Image-to-Image Translation

Cited by 8 publications

References 27 publications

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

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

Appearance generation for colored spun yarn fabric based on conditional image‐to‐image translation

Instrumental evaluation of fabric shape retention by image analysis

Contact Info

Product

Resources

About