As a popular topic in automation, fabric defect detection is a necessary and essential step of quality control in the textile manufacturing industry. The main challenge for automatically detecting fabric damage, in most cases, is the complex structure of the textile. This article presents a two-stage approach, combining novel and traditional algorithms to enhance image enhancement and defect detection. The first stage is a new combined local and global transform domain-based image enhancement algorithm using block-based alpha-rooting. In the second stage, we construct a neural network based on the modern architecture to detect fabric damage accurately. This solution allows localizing defects with higher accuracy than traditional methods of machine learning and modern methods of deep learning. All experiments were carried out using a public database with examples of damage to the TILDA fabric dataset.
Analytical equations of a new spline approximation method for filtering impulse noise in images are obtained. The proposed method differs from the known ones: when filtering images, one-dimensional sequential spline functions are used for direct and inverse transformations, and the processing is performed in rows and columns. In this work, experimental studies based on computer simulation using special test images on the background of impulse noise were conducted. Experimental studies have shown the operability and high efficiency of the developed method, which allow to improve the quality of image filtering by up to 10 dB. In this case, the properties of spline functions make it possible to abandon the use of various masks, that is, to abandon inefficient linear methods of image filtering. The method can be used to create digital image processing systems in the industry, to create autonomous robots, under observation conditions that complicate the registration process, and in the absence of a priori information about the form of background noise.
Pomortsev UDK 62t.3t7Based on mathematical techniques of Markov processes, the usefulness of constructing and utilization of a group measure to increase storage reliability of a physical quantity unit under the conditions of metrological autonomy and decrease in economic expenditures for verification (calibration) of htdividual measuring devices is demonstrated. Utilization of the methodology proposed allows us to provide a scientifically justified determination of an increased intercalibration interval of measuring devices.At the contemporary stage of development of the standard base of the country, attention of specialists, managers and organizers of investigations of the metrological service is focussed on optimization of temporal and material expenditures in the course of utilization of measuring devices.One of the problems to be solved in this instance is to assure metrological autonomy. A route for a solution of this problem may serve the construction -from the initial measuring devices of the given physical quantity -of a group measure which may be utilized for carrying out verification of the period when the functioning of the permanent scheme of transmitting the size of a unit of this physical quantity is disrupted or discontinued [ 1-3]. In turn, in the course of normal operation construction of a group measure can reduce to a large extent economic expenditures for calibration of individual measuring devices contained in the group measure. This is due to the fact that the intercalibration interval of a single measuring device is smaller than the intercalibration interval of a group measure and the calibration can be carried out over a long time period. The assertion above is based on the feature that systematic errors of measuring devices contained in a group measure are viewed in their totality as random variables [4] and with increase of the totality, the confidence interval which characterizes the degree of reproducibility of the measurement results decreases. With a decrease of the confidence interval, the confidence probability Pd increases; hence for an a priori given value Pd this allows us to retain with greater reliability the unit of a physical quantity at a given time interval, and as a result we can achieve an increase in the duration of the storage period of the given unit at Pd = const.The dependence of the time interval between the consecutive calibrations of individually taken measure T l maintaining the same unit of a physical quantity for the equal boundaries of a confidence interval (-A; +A) and the same time interval T~ for n measures (Pd = const) is presented in graphical form in Fig. 1. Thus, when utilizing a group measure of n measuring devices, an intercalibration interval of their totality can be larger than the intercalibration interval of individual measuring devices. Hence the problem of a rigorous estimation of the maximally admissible scientifically substantiated intercalibration interval is at present quite timely.The aim of this paper is to develop -based on the mathemati...
Abstract. In this paper, we consider a method for defects detection in a video sequence, which consists of three main steps; frame compensation, preprocessing by a detector, which is base on the ranking of pixel values, and the classification of all pixels having anomalous values using convolutional neural networks. The effectiveness of the proposed method shown in comparison with the known techniques on several frames of the video sequence with damaged in natural conditions. The analysis of the obtained results indicates the high efficiency of the proposed method. The additional use of machine learning as postprocessing significantly reduce the likelihood of false alarm.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.