Academia.edu no longer supports Internet Explorer.
To browse Academia.edu and the wider internet faster and more securely, please take a few seconds to upgrade your browser.
by AJAY KUMAR
2003, Pattern Recognition
Loading Preview
Sorry, preview is currently unavailable. You can download the paper by clicking the button above.
In the textile manufacturing industry, fabric defect detection becomes a necessary and essential step in quality control. The investment in this field is more than economical when reduction in labor cost and associated benefits are considered. Moreover, the development of a wholly automated inspection system requires efficient and robust algorithms. To overcome this problem, in this paper, we present a new fabric defect detection scheme which uses the local homogeneity and neural network. Its first step consists in computing a new homogeneity image denoted as í µí°»-image. The second step is devoted to the application of the discrete cosine transform (DCT) to the í µí°»-image and the extraction of different representative energy features of each DCT block. These energy features are used by the back-propagation neural network to judge the existence of fabric defect. Simulations on different fabric images and different defect aspects show that the proposed method achieves an average accuracy of 97.35%.
2011, Journal of Multimedia
In this paper, it is aimed to compare the performance of spectral based fault detection methods in quality control by testing on the same environment. The most widely used spectral based approaches as Fourier Transform, Wavelet Transform, Gabor Transform were used to extract features of the faulty fabric samples. By using statistical functions feature selection was done so huge dimensionality of features was decreased. The selected features are taken as inputs for feed-forward network (with the back propagation algorithm) to classify faulty fabrics in categories; weft, wrap and oil. All computations were performed in Matlab program so as to satisfy all conditions as the same. The analyses' results show the Wavelet transform in the classification for three type defects was more efficient than the others, on other hand; Fourier transform in terms of processing time is faster than the others.
Textile industry is one of the largest and oldest sectors in the India and has a formidable presence in national economy in terms of output, investment and employment. Due to increasing demand for quality fabrics it is thus important to produce the defect free high quality fabric. Visual inspection system consumes a lot of time and are error prone. The price of the fabric is reduced to 45%-65% due to presence of various defects. The purpose of this paper is to automate the detection and classification of texture defects by computerize software. The proposed method uses a statistical based approach for the inspection and detection of the defect on woven/knitted fabric collected from the textile industry. In this the images are acquired, pre-processed, restored and normalized to extract the statistical feature using computer vision. The extracted features are given an input to the artificial neural network decision tree classifier to compute the weighted factor for detecting and classifying the type of defects. An automatic defect detection system can increase the texture defect detection percentage and will reduce the fabrication and labour cost and improves the quality of the product. Keywords: Defect detection, Statistical approach, Computer vision, Decision tree classifier, neural network. Call for Papers: https://sites.google.com/site/ijcsis/
1997, Machine Vision and Applications
The global market for textile industry is highly competitive nowadays. Quality control in production process in textile industry has been a key factor for retaining existence in such competitive market. Automated textile inspection systems are very useful in this respect, because manual inspection is time consuming and not accurate enough. Hence, automated textile inspection systems have been drawing plenty of attention of the researchers of different countries in order to replace manual inspection. Defect detection and defect classification are the two major problems that are posed by the research of automated textile inspection systems. In this paper, we perform an extensive investigation on the applicability of genetic algorithm (GA) in the context of textile defect classification using neural network (NN). We observe the effect of tuning different network parameters and explain the reasons. We empirically find a suitable NN model in the context of textile defect classification. We compare the performance of this model with that of the classification models implemented by others.
2001, Real-time Imaging
2002, Systems, Man, and Cybernetics, Part B: …
This paper proposes a method for fabric defect detection based on neural network. The global market for textile industry is highly competitive nowadays. Neural network is widely used to extract features from images for texture segmentation. The proposed scheme involves two challenging problem .i.e. defect detection and defect classification. Scene analysis and feature selection play an important role in classification process. The complexity of the subsequent steps increases and the classification task becomes hard by selecting an inappropriate feature set. So possibly an appropriate set of geometric features is taken into account in order to address the problem of neural network-based textile defect classification. Statistical approach is used to extract the features.
1998, International Journal of Computer Mathematics
2000, Pattern Analysis & Applications
1999, Machine Vision Applications in Industrial Inspection VII
1999
A real-time pilot system for defect detection and classification of web textile fabric is presented in this paper. The general hardware and software platform, developed for solving this problem, is presented and a powerful novel method for defect detection is proposed. This method gives good results in the detection of low contrast defects under real industrial conditions, where the presence of many types of noise is an inevitable phenomenon. For the defect classification an artificial neural network, trained by using a back-propagation algorithm, is implemented. Using a reduced number of possible defect classes, the system gives consistent and repeatable results with sufficient speed.
2012, Computers and Mathematics with Applications
Detection of external defects on potatoes is the most important technology in the realization of automatic potato sorting stations. This paper presents a hierarchical grading method applied to the potatoes. In this work a potato defect detection combining with size sorting system using the machine vision will be proposed. This work also will focus on the mathematics methods used in automation with a particular emphasis on the issues associated with designing, implementing and using classification algorithms to solve equations. In the first step, a simple size sorting based on mathematical binarization is described, and the second step is to segment the defects; to do this, color based classifiers are used. All the detection standards for this work are referenced from the United States Agriculture Department, and Canadian Food Industries. Results show that we have a high accuracy in both size sorting and classification. Experimental results show that support vector machines have very high accuracy and speed between classifiers for defect detection.
2013, Journal of Visual Communication and Image Representation
2009, Expert Systems With Applications
The inspection of semi-finished and finished garments is very important for quality control in the clothing industry. Unfortunately, garment inspection still relies on manual operation while studies on garment automatic inspection are limited. In this paper, a novel hybrid model through integration of genetic algorithm (GA) and neural network is proposed to classify the type of garment defects. To process the garment sample images, a morphological filter, a method based on GA to find out an optimal structuring element, was presented. A segmented window technique is developed to segment images into several classes using monochrome single-loop ribwork of knitted garment. Four characteristic variables were collected and input into a back-propagation (BP) neural network to classify the sample images. According to the experimental results, the proposed method achieves very high accuracy rate of recognition and thus provides decision support in defect classification.
2000, Image and Vision Computing
2018, Signal, Image and Video Processing
The modified local binary pattern is a method that can produce high-precision features for detection and diagnosis of texture images; in this paper, a method is proposed to detect the texture defects based on this algorithm. The proposed method includes two main phases. The first phase is based on clustering technique to fabric normal texture modeling, and the second phase is a threshold to decide about the fabric defects selection. The total dataset in this research contains 596 texture images from different databases including Isfahan textile dataset, UHK dataset, products and TILDA dataset. The fabric defects are generated because of pressure cracks and has effects, woof defects, warp defects and spool slacking. Finally, a noticeable detection rate about 91.86% with a higher rate of 92.02% sensitivity is achieved for the total given dataset. All of the reported results from tests are achieved by applying the proposed method on the explained dataset.
2013, Procedia CIRP
2008, Optical and Digital Image Processing
2011, Journal of Electronic Imaging
During the manufacturing of textiles, several types of defects occur in the fabrics. This paper explores the characterization of the fabric textures using the conventional approaches such as Gabor filter, Gabor wavelet and Gauss Markov random field (MRF) and the well-known method for surface roughness measurement in the mechanical engineering called topothesy. The topothesy and fractal dimension known as fractal parameters represent not only the roughness but also the affine self-similarity in fabric textures. The fabric texture features are tested on the database of four types of defective fabric samples, viz., torn fabric, oil stain, miss pick and interlacing of two webs, collected from the cloth mills of Berhampur. A comparison of the results of defect detection in fabrics indicates that the topothesy fractal dimension features outperform those of Gabor filter, Gabor wavelets and Gauss MRF.
—Image mining is concerned with knowledge discovery in image databases. It is the extension of data mining algorithms to image processing domain. Image mining plays a vital role in extracting useful information from images. In computer aided plant identification and classification system the image mining will take a crucial role for the fabric defect classification. Quality inspection is an important aspect of modern industrial manufacturing. In textile industry production, automate fabric inspection is important for maintain the fabric quality. In modern textile industry, Tissue online Automatic Inspection (TAI) is becoming an attractive alternative to Human Vision Inspection (HVI). HVI needs a high level of attention nevertheless leading to low performance in terms of tissue inspection. Automatic fabric inspection is valuable for maintenance of fabric quality. Defect inspection of fabric is a process which accomplished with human visual look-over using semi-automated way but it is labor prone and costly. Many sewing, knitting and dyeing units involve both manual and automated processes. Detecting faults in fabric manually, by human visual inspection is a tedious task. Its accuracy depends upon the skill of human operator and varies from person to person to address this difficulty a method is proposed for textile defect identification and classification based on image mining. The detection of local fabric defects is one of the most intriguing problems in computer vision. Texture analysis plays an important role in the automated visual inspection of texture images to detect their defects. Various approaches for fabric defect detection have been proposed in past and in this paper; we proposed a method based on texture analysis, association rule classifier, threshold segmentation and histogram to identify the textile defects. A feature extractor is designed based on Gray Level Co-occurrence Matrix (GLCM). An association rule miner is used as a classifier to identify the textile defects.
Human inspection is the traditional means to promise the quality of fabric. It helps instant correction of small defects, but errors left due to human eye occurs because of fatigue and fine defects are often undetected. Therefore, automated inspection of fabric defect is then a natural way to improve fabric quality and reduce labor costs. The most difficulty faced by industrial inspection problems deals with the textured materials such as textile web, paper, and wood. In order to deal with the problem of defect detection in fabric defect images we have a proposed an efficient algorithm to easily detect the region containing fabric defects. We have converted the input RGB images into other color spaces i.e. Lab and ycbcr which are more close to human perception and gives better results in feature extraction as well as classification stage. For feature detection we have intensity values from these two color spaces in which rough separation of two classes has been evaluated using CSLBP features and DCT algorithms and neural object has been trained and tested to get segment out the defect region from the rest of the fabric image. Experimental results showed better accuracy on various types of defects i.e. hole, broken needle, dye spot, slub defects etc.
1998, Kybernetes
2007, Vision Systems: Applications
2007, Applied Optics
The automatic segmentation of flaws in woven fabrics is achieved by applying Fourier analysis to the image of the sample under inspection, without considering any reference image. No prior information about the fabric structure or the defect is required. The algorithm is based on the structural feature extraction of the weave repeat from the Fourier transform of the sample image. These features are used to define a set of multiresolution bandpass filters, adapted to the fabric structure, that operate in the Fourier domain. Inverse Fourier transformation, binarization, and merging of the information obtained at different scales lead to the output image that contains flaws segmented from the fabric background. The whole process is fully automatic and can be implemented either optically or electronically. Experimental results are presented and discussed for a variety of fabrics and defects.
2010, Expert Systems With Applications
2017, Albaha University Journal of Basic and Applied Sciences
Technological development accompanied the need to get a high-quality welding. In this research, an automatic technique is introduced to detect, recognize and classify welding defects in radiographic (x-ray) images, using texture features. Image processing techniques, including converting color images to grayscale, filtering and resizing images were applied to help in the image array of weld images and welding defect detection. Therefore, a proposed program was built in-house to automatically classify and recognize the most common types of welding defects met in practice. The introduced technique has been tested on eleven welding defects which are: center line crack, cap undercut, elongated slag lines, lack of interpass fusion, lack of root penetration, lack of side wall fusion, misalignment, root crack, root pass aligned, root undercut, and transverse crack (n = 35 for each). The overall average discrimination rate is about 94.29%. The introduced technique can find promising application of digital image processing technique to the field of welding defect inspection compared with traditional methods.
2007, Pattern recognition
2014, Neurocomputing
1999, … and Systems, 1999. …
In this paper a pilot system for defect detection and classification of web textile fabric in real-time is presented. The general hardware and software platform, developed for solving this problem, is presented while a powerful novel method for defect detection after ...
2010, Image and Vision Computing
2010, 2010 20th International Conference on Pattern Recognition