China
Africa
Explore chapters and articles related to this topic
Identity recognition approach based on the improved phase congruency of the gait energy image
Published in Jimmy C.M. Kao, Wen-Pei Sung, Civil, Architecture and Environmental Engineering, 2017
Ling-yao Jia, Chao Liang, Dong-cheng Shi
ABSTRACT: The use of phase congruency for marking features has significant advantages over gradient-based methods. It is a dimensionless quantity that is invariant to changes in image brightness or contrast, hence it provides an absolute measure of the significance of feature points. In this paper, identity recognition based on improved Log-Gabor phase congruency is proposed. The improved phase congruency algorithm is based on the improved local energy calculation method, frequency spread, and noise compensation tactics. The feature of improved phase congruency is of good location and recognition, and then the LDA method is used to project features into a low-dimensional space. The nearest neighbor classifier based on Euclidean distance is tested in the CASIA gait database. The experimental results show that our approach outperforms other state of the art automatic algorithms in terms of recognition accuracy.
Multimodal Medical Image Fusion in NSCT Domain
Published in Ayman El-Baz, Jasjit S. Suri, Big Data in Multimodal Medical Imaging, 2019
Gaurav Bhatnagar, Zheng Liu, Q.M. Jonathan Wu
Phase congruency is a measure of feature perception in the images which provides an illumination and contrast invariant feature extraction method [53,54]. This approach is based on the Local Energy Model, which postulates that significant features can be found at points in an image where the Fourier components are maximally in phase. Furthermore, the angle at which phase congruency occurs signifies the feature type. The phase congruency approach to feature perception has been used for feature detection. First, logarithmic Gabor filter banks at different discrete orientations are applied to the image, and the local amplitude and phase at a point (x, y) are obtained. The phase congruency, Px,yo, is then calculated for each orientation o as follows. Px,yo=∑nWx,yoAx,yo,ncosϕx,yo,n−ϕ˜x,yo−sinϕx,yo,n−ϕ˜x,yo−T∑nAx,yo,n+ε
Dental Image Segmentation and Classification Using Inception Resnetv2
Published in IETE Journal of Research, 2021
The proposed segmentation technique is used to highlight the minimized threshold level. The detection of such curved structures is affected greatly by the contrast intensity variations that occur within the images. The difference between the curvilinear structures with the background is most common in such medical imaging-based applications. The method holds high accuracy for these computer-aided segmentation techniques that are ideal for processing medical images. First, the background is removed from the input image, and the next segmentation is done for the foreground regions. The boundaries with respect to the minimal contrast structural parts are also detected based on the image gradient. The proposed method is invariant to changes based on image intensity and also with the contrast. A Deep CNN algorithm was also employed in the step segmentation of dental X-Ray imaging. Deep CNN provides clustered functions. These algorithms have the function of grouping related objects and separate dissimilar ones. It helps to resolve the downside of using all data in one go requiring each iteration. It helps to extract the image features based on the phase congruency model based on the image feature points which provide maximal phase structures.
A Modified Framework for Multislice Image Fusion for High Contrast Liver Cancer Detection
Published in IETE Journal of Research, 2020
B. Lakshmi Priya, K. Jayanthi, Biju Pottakkat, G. Ramkumar
Kovesi in [23] formulated phase congruency as a low-level invariant feature of an image. Human visual system is able to identify the significant image features under a wide range of conditions. The interpretation of a scene is not greatly affected by the changes in illumination of a scene and spatial magnification. The significant feature in low-level feature detection is the contrast invariance and this is quantified by means of phase congruency. Phase congruency can be measured by framing a local energy model, which suggests that notable features on image can be from points whose Fourier series components have same phase. Congruency of phase at any angle produces a clearly perceived feature. Phase congruency [24], ‘PC(x)’ of a one-dimensional signal in terms of Fourier series expansion is given by where represents the amplitude of the nth component of Fourier series and represents the local phase of the Fourier component at the point x. The two-dimensional version of phase congruency has been derived in Kovesi [23] and is reframed in Kovesi [25] by first applying log-gabor wavelet-based filter banks at different discrete orientations to the image of interest. Secondly, the local amplitude and phase at pixel corresponding (i, j) are calculated. Where θ denotes the orientation, is weigthing factor based on frequency spread, and are the amplitude and phase of the wavelet scale n. represents the weighted mean phase, T is the noise threshold, and ε is a small constant included to avoid instability due to zero denominator. Since phase congruency provides a low-level invariant feature measure of an image and as it addresses the problem of localization of features, this parameter is employed for fusing low-frequency components in literature [26,27].