China
Africa
Explore chapters and articles related to this topic
Expression Classification
Published in Yu-Jin Zhang, A Selection of Image Analysis Techniques, 2023
Based on the detection and tracking of facial organs, features for facial expression classification can be further extracted. Gabor transform is often used in expression feature extraction, which is a special short-time Fourier transform, and its window function is Gaussian function. When extracting features with the help of Gabor transform, a group of Gabor filters of different scales and directions are often used (a variety of sampling methods can be selected for this group of filters, and only part of the results are used to reduce the time and space demands of feature extraction and classification (Xu and Zhang 2011)). Using a set of such filters can analyze image grayscale changes in various scales and directions, and can further detect the corner points of objects and the end points of line segments.
Frames and Wavelets: A New Perspective on Sampling Theorems
Published in Ahmed I. Zayed, Advances in Shannon’s Sampling Theory, 2018
One of the earliest mathematical techniques used to remedy this deficiency in the Fourier transform was introduced in 1946 by a Hungarian-British physicist and engineer, Dennis Gabor [11], who won the Nobel Prize in physics in 1971 for his invention and development of the holographic method. Gabor introduced another transform that is akin to the Fourier transform and which is now known as the Gabor transform. The Gabor transform G(f)(ω,t) of a signal f is defined by G(f)(ω,t)=∫−∞∞f(x)g(x−t)e2πiωxdx,
Multimodal Ambulatory Fall Risk Assessment in the Era of Big Data
Published in Ervin Sejdić, Tiago H. Falk, Signal Processing and Machine Learning for Biomedical Big Data, 2018
Due to the specific characteristics of our images, texture seems to be an appropriate feature for describing their contents (e.g., brick, tiles, rocks, carpet) [79]. Texture analysis has been an active research area, and numerous algorithms have been proposed based on different models, for example, gray-level co-occurrence (GLC) matrices and Markov random field (MRF) model [79,88]. In recent works, wavelets have become very popular due to their capacity to provide multiresolution analysis. In particular, the Gabor transform has mathematical and biological properties resembling the characteristics of human visual cortical cells, such as extracting texture features from images for segmentation, object detection, and biometric identification applications.
Performance Analysis of Glioma Brain Tumor Segmentation Using CNN Deep Learning Approach
Published in IETE Journal of Research, 2023
The pixels in enhanced brain MRI image are in spatial domain, where the pixels cannot be modified in this mode. Hence, there is a need for transforming these spatial domain pixels into multiresolution pixels using transformation process. This article utilizes Gabor transform or filter which is used to transform the spatial pixels into multiresolution pixels. Gabor transform works on the principle of Gaussian functions, which can be operated based on its frequency, scale and different angle of orientations. In this paper, multiscale and multiorientation-based Gabor filters are used to transform the spatial pixels in enhanced brain MRI image into various scales and orientation pixels. The Gabor kernel of the Gabor filter is given in the following equation. where and are the standard deviation in x and y direction, respectively, x and y are the pixel coordinates of the image, f is the frequency scale and the standard deviations ( and ) are computed using the following equations.
Fabric defect detection using ACS-based thresholding and GA-based optimal Gabor filter
Published in The Journal of The Textile Institute, 2023
Runhu Zhu, Binjie Xin, Na Deng, Mingzhu Fan
Directly thresholding the fabric defect image is disturbed by the background texture. The Gabor transform is a special case of the short-time Fourier transform. It is found that Gabor-based methods are particularly appropriate for texture analysis and segmentation. Tong et al. (2016) used the composite differential evolution (CoDE) algorithm to optimize the parameters of the Gabor filter. Defects were segmented from the original image background by combining threshold segmentation and fusion operation; Jia et al. (2017) studied the response distribution of the convolution lattice of the Gabor filter to detect defects, and the overall detection rate was 0.975; Jing et al. (2017) proposed two methods based on Gabor filtering and gold image subtraction. Gabor filter combined with specific parameters and the genetic algorithm was used for filtering preprocessing; Li et al. (2019) adopted the optimal elliptic Gabor filter (EGF) and the random drift particle swarm optimization (RDPSO) algorithm to determine the optimal EGF parameters.
Fabric defect detection algorithm using RDPSO-based optimal Gabor filter
Published in The Journal of The Textile Institute, 2019
Yueyang Li, Haichi Luo, Miaomiao Yu, Gaoming Jiang, Honglian Cong
In recent years, a great number of studies have been reported on the automatic fabric defect detection. The detection methods can be classified into five categories, namely, structural, statistical, model-based, learning, and spectral approaches (Hanbay et al., 2016). Among them, the use of spectral approaches on fabric defect detection might be a better choice because the patterns of a great number of fabrics are periodic structures. A large number of studies have been done on fabric defect detection using spectral approaches such as Fourier transform, wavelet transform, and Gabor filters. Fourier-based methods (Chi-Ho & Pang, 2000; Malek, 2012) analyze the spectral distribution of fabric images. However, Fourier analysis might be not suitable for detecting local defects, which usually appear on fabric surface because it is hard to quantify the contribution of each spectral component of the infinite Fourier basis. Through decomposing an image into a hierarchy of localized sub images at different scales, the wavelet transform can provide more localized information. Therefore, wavelet-based methods have been successfully applied for fabric defect detection (Hu, Wang, & Zhang, 2015; Serdaroglu, Ertuzun, & Ercil, 2006). The Gabor transform is a special case of the short-time Fourier transform. Gabor-based methods have been found to be particularly appropriate for fabric defect detection owning to its merits of optimal joint localization in spatial and frequency domains (Hu, 2015; Kumar & Pang, 2002; Tong, Wong, & Kwong, 2016).