Log-Gabor Filter Research Articles

Deep convolutional neural networks for bias field correction of brain magnetic resonance images

As a low-frequency and smooth signal, the bias field has a certain destructive effect on magnetic resonance (MR) images and is the main obstacle for doctors' diagnosis and image processing (such as segmentation, texture analysis, and registration). Before analyzing a damaged MR image, a preprocessing step is required to correct the bias field in the image. Unlike traditional bias field removal algorithms based on signal models and a priori assumptions, deep learning methods do not require precisely modeling signals and bias fields and do not need to adjust parameters. An MR image with the bias field is input and the corrected MR image is output after the deep neural network being trained on a large training set. In this paper, we propose taking the original image and the local feature images of the bias field in multiple frequency bands obtained by a Log-Gabor filter bank as input, correcting the bias field of a brain MR image through a deep separable convolutional neural network. Meanwhile, to speed up the training process and improve bias correction performance, we apply residual learning and batch normalization. We conducted the same test on BrainWeb simulation database and Human Connectome Project real data set, the consistency of qualitative and quantitative evaluation shows that our training model demonstrates better performance than the traditional state-of-the-art N4 and non-iterative multi-scale (NIMS) methods. Especially for the images with high-intensity non-uniformity level, the bias field has been well corrected.

A psychophysical performance-based approach to the quality assessment of image processing algorithms.

Image processing algorithms are used to improve digital image representations in either their appearance or storage efficiency. The merit of these algorithms depends, in part, on visual perception by human observers. However, in practice, most are assessed numerically, and the perceptual metrics that do exist are criterion sensitive with several shortcomings. Here we propose an objective performance-based perceptual measure of image quality and demonstrate this by comparing the efficacy of a denoising algorithm for a variety of filters. For baseline, we measured detection thresholds for a white noise signal added to one of a pair of natural images in a two-alternative forced-choice (2AFC) paradigm where each image was selected randomly from a set of n = 308 on each trial. In a series of experimental conditions, the stimulus image pairs were passed through various configurations of a denoising algorithm. The differences in noise detection thresholds with and without denoising are objective perceptual measures of the ability of the algorithm to render noise invisible. This was a factor of two (6dB) in our experiment and consistent across a range of filter bandwidths and types. We also found that thresholds in all conditions converged on a common value of PSNR, offering support for this metric. We discuss how the 2AFC approach might be used for other algorithms including compression, deblurring and edge-detection. Finally, we provide a derivation for our Cartesian-separable log-Gabor filters, with polar parameters. For the biological vision community this has some advantages over the more typical (i) polar-separable variety and (ii) Cartesian-separable variety with Cartesian parameters.

Efficient palmprint biometric identification systems using deep learning and feature selection methods

Over the past two decades, several studies have paid great attention to biometric palmprint recognition. Recently, most methods in literature adopted deep learning due to their high recognition accuracy and the capability to adapt with different acquisition palmprint images. However, high-dimensional data with a large number of uncorrelated and redundant features remain a challenge due to computational complexity issues. Feature selection is a process of selecting a subset of relevant features, which aims to decrease the dimensionality, reduce the running time, and improve the accuracy. In this paper, we propose efficient unimodal and multimodal biometric systems based on deep learning and feature selection. Our approach called simplified PalmNet–Gabor concentrates on the improvement of the PalmNet for fast recognition of multispectral and contactless palmprint images. Therefore, we used Log-Gabor filters in the preprocessing to increase the contrast of palmprint features. Then, we reduced the number of features using feature selection and dimensionality reduction procedures. For the multimodal system, we fused modalities at the matching score level to improve system performance. The proposed method effectively improves the accuracy of the PalmNet and reduces the number of features as well the computational time. We validated the proposed method on four public palmprint databases, two multispectral databases, CASIA and PolyU, and two contactless databases, Tongji and PolyU 2D/3D. Experiments show that our approach achieves a high recognition rate while using a substantially lower number of features.

Robust Multimodal Remote Sensing Image Registration Based on Local Statistical Frequency Information

Multimodal remote sensing image registration is a prerequisite for comprehensive application of remote sensing image data. However, inconsistent imaging environment and conditions often lead to obvious geometric deformations and significant contrast differences between multimodal remote sensing images, which makes the common feature extraction extremely difficult, resulting in their registration still being a challenging task. To address this issue, a robust local statistics-based registration framework is proposed, and the constructed descriptors are invariant to contrast changes and geometric transformations induced by imaging conditions. Firstly, maximum phase congruency of local frequency information is performed by optimizing the control parameters. Then, salient feature points are located according to the phase congruency response map. Subsequently, the geometric and contrast invariant descriptors are constructed based on a joint local frequency information map that combines Log-Gabor filter responses over multiple scales and orientations. Finally, image matching is achieved by finding the corresponding descriptors; image registration is further completed by calculating the transformation between the corresponding feature points. The proposed registration framework was evaluated on four different multimodal image datasets with varying degrees of contrast differences and geometric deformations. Experimental results demonstrated that our method outperformed several state-of-the-art methods in terms of robustness and precision, confirming its effectiveness.

A Convolutional Neural Network for Learning Local Feature Descriptors on Multispectral Images

This work presents a novel convolutional neural network, termed multispectral features network (MF-Net), for learning local feature descriptors in multispectral images. Unlike most existing solutions, which primarily handle images from the visible light spectrum, we propose a learning-based method that deals with image data acquired from different spectrum bands. To design our convolutional neural network, we introduce a new layer that incorporates Log-Gabor filters to enhance the network capability to work with nonlinear intensity variations in images captured from different electromagnetic frequencies spectrum. This layer, entitled mapping layer, can be easily integrated into different network architectures. To demonstrate the efficacy and limitations of our method, we went on experiments with two distinct datasets extensively used in previous works composed of image pairs from the visible spectrum and the infrared spectrum. Experimental results with datasets containing images obtained from visible light and infrared spectrum show that our method can accurately match features, outperforming some state-of-the-art learning-based algorithms.

3MRS: An Effective Coarse-to-Fine Matching Method for Multimodal Remote Sensing Imagery

The fusion of image data from multiple sensors is crucial for many applications. However, there are significant nonlinear intensity deformations between images from different kinds of sensors, leading to matching failure. To address this need, this paper proposes an effective coarse-to-fine matching method for multimodal remote sensing images (3MRS). In the coarse matching stage, feature points are first detected on a maximum moment map calculated with a phase congruency model. Then, feature description is conducted using an index map constructed by finding the index of the maximum value in all orientations of convolved images obtained using a set of log-Gabor filters. At last, several matches are built through image matching and outlier removal, which can be used to estimate a reliable affine transformation model between the images. In the stage of fine matching, we develop a novel template matching method based on the log-Gabor convolution image sequence and match the template features with a 3D phase correlation matching strategy, given that the initial correspondences are achieved with the estimated transformation. Results show that compared with SIFT, and three state-of-the-art methods designed for multimodal image matching, PSO-SIFT, HAPCG, and RIFT, only 3MRS successfully matched all six types of multimodal remote sensing image pairs: optical–optical, optical–infrared, optical–depth, optical–map, optical–SAR, and day–night, with each including ten different image pairs. On average, the number of correct matches (NCM) of 3MRS was 164.47, 123.91, 4.88, and 4.33 times that of SIFT, PSO-SIFT, HAPCG, and RIFT for the successfully matched image pairs of each method. In terms of accuracy, the root-mean-square error of correct matches for 3MRS, SIFT, PSO-SIFT, HAPCG, and RIFT are 1.47, 1.98, 1.79, 2.83, and 2.45 pixels, respectively, revealing that 3MRS got the highest accuracy. Even though the total running time of 3MRS was the longest, the efficiency for obtaining one correct match is the highest considering the most significant number of matches. The source code of 3MRS and the experimental datasets and detailed results are publicly available.

Textural Analysis for Medicinal Plants Identification Using Log Gabor Filters

Texture plays a crucial role in computer vision, providing valuable information about image regions. Log Gabor filters that mimic the human eye’s visual cortex are used as feature extractors to identify medicinal plants based on the leaf textural features. This method was tested on a dataset developed from the Centre of Plant Medicine Research, Ghana, consisting of forty-nine (49) plant species as well as the Flavia and Swedish Leaf datasets, which are benchmark datasets. The Log Gabor filter outperformed the Gabor filters, which have been extensively used in this area when tested on nine supervised classifiers (K-Nearest Neighbour, Support Vector Machine, Naïve Bayes, Logistic Regression, Decision tree, Random Forest, Multilayer Perceptron, Gradient Boosting and Stochastic Gradient Descent) with 10-fold cross-validation. The Support Vector Machine and Multilayer Perceptron were the best performing classifiers for both Log Gabor filter and Gabor filter in terms of accuracy, precision, true positive rate, F1 score and false positive rate. The Log Gabor filter’s highest accuracy was 79% for Mydatastet, 97% for Flavia, and 98% for the Swedish Leaf dataset whiles the Gabor filter’s highest accuracy was 66% for Mydatastet, 92% for Flavia and 96% for the Swedish Leaf dataset.

Feature Matching of Multimodal Images Based on Nonlinear Diffusion and Progressive Filtering

Traditional image feature matching methods cannot obtain satisfactory results for multi-modal images in most cases because different imaging mechanisms bring significant nonlinear radiation distortion differences and geometric distortion. The key to multi-modal image matching is trying to eliminate the nonlinear radiation distortion and extract more robust features. This paper proposes a new robust feature matching method for multi-modal images. Our method starts by detecting feature points on phase congruency maps in nonlinear scale space and then removing mismatches by progressive filtering. Specifically, the phase congruency maps are generated by the Log-Gabor filter. Then, the feature points on phase congruency maps are detected in nonlinear scale space constructed by the nonlinear diffusion filter. Subsequently, the structure descriptor is established by the Log-Gabor filter, and the initial correspondences are constructed by bilateral matching. Finally, an iterative strategy is used to remove mismatches by progressive filtering. We perform comparison experiments on our proposed method with the SIFT, RIFT, VFC, LLT, LPM, and mTopKPR methods using multi-modal images. The algorithms of each method are comprehensively evaluated both qualitatively and quantitatively. Our experimental results indicate the superiority of our method over the other six matching methods.

Comprehensive edge direction descriptor for fingerprint liveness detection

The edge direction is one of the most discriminative image information. The existing edge direction based feature extraction methods only extract the explicit edge direction information but ignore the complementary implicit edge direction information. In this paper, we propose a comprehensive edge direction information based image feature extraction method for fingerprint liveness detection. The main novelties of our method can be summarized as follows. (1) We propose the singular value decomposition (SVD) based image Log-Gabor transform energy extraction method, with the energy image preserving the most dominant convolution response. (2) We extract the explicit and implicit edge direction information from the Log-Gabor transform of original image and SVD energy image respectively. (3) Adopting the Log-Gabor filters as the codewords, we propose the orientation normalization based scale Non-K-Maximum suppression Log-Gabor transform encoding method and the positive–negative separation mean pooling method. Our method is of lower computational complexity compared with existing codeword based image representation methods. Extensive experimental evaluations on three benchmark databases show that the proposed method yields desirable performance compared with the state-of-the-art methods.

A novel hybrid face recognition framework based on a low-resolution camera for biometric applications

In research work, human face recognition is an essential biometric symbol persistently continued so far due to its different levels of applications in society. Since the appearance of the human faces can have many variations due to issues like the effect of illumination, expression and face pose. These differences are correlated with one another, which results in a helpless ability to recognize a particular person's face. The motivation behind our work in this paper is to give a new framework for face recognition based on frequency analysis that contributes to solving the distinguishing proof issues with enormous varieties of boundaries like the effect of illumination, expression, and face pose. Here three algorithms combined for provable results: i) Difference of Gaussian filtered discrete wavelet transform (DDWT) for feature extraction; ii) Log Gabor (LG) filter for feature extraction; and iv) Multiclass support vector machine classifier, where feature coefficients of DDWT and LG filter are fused for classification and parameters evaluation. The evaluation of our experiment is carried out on a large database consisting of 15 persons of each 200-face image which are captured using a 5-megapixel low-resolution web camera and yielding satisfactory results on various parameters compared to existing methods.

Quantification of retinal blood leakage in fundus fluorescein angiography in a retinal angiogenesis model

Blood leakage from the vessels in the eye is the hallmark of many vascular eye diseases. One of the preclinical mouse models of retinal blood leakage, the very-low-density-lipoprotein receptor deficient mouse (Vldlr−/−), is used for drug screening and mechanistic studies. Vessel leakage is usually examined using Fundus fluorescein angiography (FFA). However, interpreting FFA images of the Vldlr−/− model is challenging as no automated and objective techniques exist for this model. A pipeline has been developed for quantifying leakage intensity and area including three tasks: (i) blood leakage identification, (ii) blood vessel segmentation, and (iii) image registration. Morphological operations followed by log-Gabor quadrature filters were used to identify leakage regions. In addition, a novel optic disk detection algorithm based on graph analysis was developed for registering the images at different timepoints. Blood leakage intensity and area measured by the methodology were compared to ground truth quantifications produced by two annotators. The relative difference between the quantifications from the method and those obtained from ground truth images was around 10% ± 6% for leakage intensity and 17% ± 8% for leakage region. The Pearson correlation coefficient between the method results and the ground truth was around 0.98 for leakage intensity and 0.94 for leakage region. Therefore, we presented a computational method for quantifying retinal vascular leakage and vessels using FFA in a preclinical angiogenesis model, the Vldlr−/− model.

Defect Detection on a Wind Turbine Blade Based on Digital Image Processing

Wind power generation is a widely used power generation technology. Among these, the wind turbine blade is an important part of a wind turbine. If the wind turbine blade is damaged, it will cause serious consequences. The traditional methods of defect detection for wind turbine blades are mainly manual detection and acoustic nondestructive detection, which are unsafe and time-consuming, and have low accuracy. In order to detect the defects on wind turbine blades more safely, conveniently, and accurately, this paper studied a defect detection method for wind turbine blades based on digital image processing. Because the log-Gabor filter used needed to extract features through multiple filter templates, the number of output images was large. Firstly, this paper used the Lévy flight strategy to improve the PSO algorithm to create the LPSO algorithm. The improved LPSO algorithm could successfully solve the PSO algorithm’s problem of falling into the local optimal solution. Then, the LPSO algorithm and log-Gabor filter were used to generate an adaptive filter, which could directly output the optimal results in multiple feature extraction images. Finally, a classifier based on HOG + SVM was used to identify and classify the defect types. The method extracted and identified the scratch-type, crack-type, sand-hole-type, and spot-type defects, and the recognition rate was more than 92%.

Quality assessment of screen content images based on multi-stage dictionary learning

In this paper, we propose an effective method for quality assessment of screen content images (SCIs) based on multi-stage dictionary learning. To simulate the brain’s layered processing of signals, we proposed a hierarchical feature extraction strategy, which is called multi-stage dictionary learning, to simulate the hierarchical information processing of brain. First, the standard deviation of normalized map obtained from training image is used to select the training data in a certain proportion, which can ensure the learning efficiency and reduce the training burden. Next, the reconstructed map is weighted as the input of the next-stage dictionary learning. Then using the trained dictionary, the sparse representation is applied to extract features. Meanwhile, considering that some important features may be ignored in the process of multi-stage dictionary learning, we use Log Gabor filter to extract feature maps, and then calculate the correlation between feature maps as another kind of compensation features. Final, for the two feature sets, we choose SVR and feature codebook to learn two objective scores, and then use the adaptive weighting strategy to get the final objective quality score. Experimental results show that the proposed method is superior to several mainstream SCIs metrics on two publicly available databases.

BVMatch: Lidar-Based Place Recognition Using Bird's-Eye View Images

Recognizing places using Lidar in large-scale environments is challenging due to the sparse nature of point cloud data. In this letter we present BVMatch, a Lidar-based frame-to-frame place recognition framework, that is capable of estimating 2D relative poses. Based on the assumption that the ground area can be approximated as a plane, we uniformly discretize the ground area into grids and project 3D Lidar scans to bird's-eye view (BV) images. We further use a bank of Log-Gabor filters to build a maximum index map (MIM) that encodes the orientation information of the structures in the images. We analyze the orientation characteristics of MIM theoretically and introduce a novel descriptor called bird's-eye view feature transform (BVFT). The proposed BVFT is insensitive to rotation and intensity variations of BV images. Leveraging the BVFT descriptors, we unify the Lidar place recognition and pose estimation tasks into the BVMatch framework. The experiments conducted on three large-scale datasets show that BVMatch outperforms the state-of-the-art methods in terms of both recall rate of place recognition and pose estimation accuracy.

Deep Learning for Face Recognition under Complex Illumination Conditions Based on Log-Gabor and LBP

The complex light conditions, and this is one of the most important and difficult problems in practical face recognition, in this paper, we propose a new deep learning-based method to solve the problem of the effect of the light of the changes in the facial recognition process. First, the primary treatment of the lighting can be used to improve the negative effects of intensive changes in the lighting of a photo of a face, and for a second, the Log-Gabor filters in order to get the images used in the Log-Gabor features at different scales and in different directions, and then, the LBP (Local Binary Pattern) features on the image subblock is obtained. Finally, the histogram of the texture of the features of the formation and the visual layer of the deep belief network (DBN) to drop, and then the classification and the recognition is done with a deep-learning-DBN. The experimental results show that superior performance can be obtained in the application of the strategy in comparison to some of the modern technology.

Comparative study of ultrasonic techniques for reconstructing the multilayer structure of composites

The multilayer structure of fiber-reinforced polymers may be extracted by ultrasonic pulse-echo inspection. Depending on the employed ultrasonic frequency and subsequent processing methodology, different depth resolution and dynamic depth range can be achieved. This study compares the performance of different ultrasonic pulse-echo approaches for extracting the ply-by-ply structure of multilayered composites. The following methodologies are studied: Method 1: 50 MHz, 15 MHz, and 5 MHz ultrasound with low-pass filtering using analysis of the instantaneous amplitude, Method 2: 15 MHz ultrasound with Wiener deconvolution (and autoregressive spectral extrapolation) using analysis of the instantaneous amplitude, and Method 3: 5 MHz ultrasound with low-pass or log-Gabor filtering using analysis of the instantaneous phase. In the simulation study, the performance of the various techniques is investigated on synthetic data representative for a 24-ply carbon fiber reinforced polymer. The robustness of the techniques is evaluated for different signal-to-noise ratios. The various techniques are further investigated on experimental data of a 24-ply cross-ply carbon fiber reinforced polymer. The ply-by-ply structure is extracted and presented in the form of both B-scan and C-scan images. The thickness of each ply is estimated for quantitative analysis. The obtained results indicate that the 5 MHz ultrasound coupled to analytic-signal analysis with log-Gabor filter shows the best performance for reconstructing the multilayer structure of the studied composites.

An advanced gradient texture feature descriptor based on phase information for infrared and visible image matching

Infrared and visible image matching has many applications in remote sensing, computer vision, military fields, etc. The differences in the many characteristics of infrared images and visible images make a robust feature description vital but difficult. Texture orientation information retains the general properties in refrared and visible images, and multi-scale, and multi-oriented Gabor filters can accurately reveal the texture orientation information. This paper presents a feature descriptor by capturing the phase information between neighboring pixels with Log-Gabor filters. Firstly, the original matching image is enhanced via histogram equalization to emphasize the regions of interest, and the gradient magnitude for each pixel is computed to extract the image profile, which advances the performance of the algorithm significantly. Secondly, multi-scale and multi-oriented Log–Gabor filters are utilized to obtain the angle information for different scales and phases in the neighboring region of each pixel, and the angle information is indexed by computing the maximum of energy, including the magnitude, real part, and imaginary part to generate the marked image in which the histograms of the subregion of the detected keypoints are employed to generate the feature descriptors. Finally, we advocate five evaluation measures for testing the performance of the algorithm. The proposed approach is evaluated with four data sets composed of images obtained in visible light and infrared spectra, and its performance is compared with the performance of the state-of-the-art algorithms: Scale-invariant feature transform(SIFT), Speeded up robust features(SURF), Oriented fast and rotated BRIEF(ORB), the edge-oriented histogram descriptor (EHD), the phase congruency edge-oriented histogram discriptor (PCEHD), and the Log–Gabor histogram descriptor (LGHD). The experimental results indicate that the performance of the proposed approach is higher than that of other state-of-the-art algorithms.

A Hybrid Swarm and Gravitation-based feature selection algorithm for handwritten Indic script classification problem

In any multi-script environment, handwritten script classification is an unavoidable pre-requisite before the document images are fed to their respective Optical Character Recognition (OCR) engines. Over the years, this complex pattern classification problem has been solved by researchers proposing various feature vectors mostly having large dimensions, thereby increasing the computation complexity of the whole classification model. Feature Selection (FS) can serve as an intermediate step to reduce the size of the feature vectors by restricting them only to the essential and relevant features. In the present work, we have addressed this issue by introducing a new FS algorithm, called Hybrid Swarm and Gravitation-based FS (HSGFS). This algorithm has been applied over three feature vectors introduced in the literature recently—Distance-Hough Transform (DHT), Histogram of Oriented Gradients (HOG), and Modified log-Gabor (MLG) filter Transform. Three state-of-the-art classifiers, namely, Multi-Layer Perceptron (MLP), K-Nearest Neighbour (KNN), and Support Vector Machine (SVM), are used to evaluate the optimal subset of features generated by the proposed FS model. Handwritten datasets at block, text line, and word level, consisting of officially recognized 12 Indic scripts, are prepared for experimentation. An average improvement in the range of 2–5% is achieved in the classification accuracy by utilizing only about 75–80% of the original feature vectors on all three datasets. The proposed method also shows better performance when compared to some popularly used FS models. The codes used for implementing HSGFS can be found in the following Github link: https://github.com/Ritam-Guha/HSGFS.

A Novel GA-Based Optimized Approach for Regional Multimodal Medical Image Fusion With Superpixel Segmentation

For multimodal medical image fusion problems, most of the existing fusion approaches are based on pixel-level. However, the pixel-based fusion method tends to lose local and spatial information as the relationships between pixels are not considered appropriately, which has much influence on the quality of the fusion results. To address this issue, a region-based multimodal medical image fusion framework is proposed based on superpixel segmentation and a post-processing optimization method in this paper. In this framework, the average image of the source medical images is firstly obtained by a weighted averaging method. To effectively obtain homogeneous regions and preserve the complete information of image details, the fast linear spectral clustering(LSC) superpixel algorithm is carried out to segment the average image and get superpixel labels. For each region of the medical images, log-gabor filter(LGF) and sum modified laplacian(SML) are adopted to extract texture feature and contrast feature for the measurement of region importance. The most important regions are selected and the decision map is generated by comparison. Moreover, to get a more accurate decision map, a new post-processing optimized method based on genetic algorithm(GA) is given. A weighted strategy is applied to the extracted features and the weighting factor can be adaptively adjusted by GA. The effectiveness of the proposed fusion method is validated by conducting experiments on eight pairs of medical images from diverse modalities. In addition, seven other mainstream medical image fusion methods are adopted for comparing the performance of fusion. Experimental results in terms of qualitative and quantitative evaluation demonstrate that the proposed method can achieve state-of-the-art performance for multimodal medical image fusion problems.

El Yapımı Tabanlı ve Derin Öğrenme Yöntemlerini Kullanan Yüz Yanıltma Önleme Şeması

Malicious parties which impersonate systems by fake identities affect recognition performance of biometric systems. This study focuses on a strength anti-spoofing scheme based on decision level fusion to monitor individuals in term of real and fake. The proposed fake detection scheme involves consideration of both handcrafted and deep learned techniques on face images to differentiate real and fake individuals. In this context, convolutional neural network (CNN) and Log-Gabor filter methods are used to learn deep representations and extract facial features of images respectively. In order to improve the robustness of proposed anti-spoofing framework, fusion of Log-Gabor and CNN methods is considered by applying decision-level-fusion technique. Finally, the performance of proposed anti- spoofing scheme is examined on public spoof databases such as Print-Attack and Replay-Attack face databases to detect fake facial images.