Local Feature Extraction Method Research Articles

Efficient Crowd Anomaly Detection Using Sparse Feature Tracking and Neural Network

Crowd anomaly detection is crucial in enhancing surveillance and crowd management. This paper proposes an efficient approach that combines spatial and temporal visual descriptors, sparse feature tracking, and neural networks for efficient crowd anomaly detection. The proposed approach utilises diverse local feature extraction methods, including SIFT, FAST, and AKAZE, with a sparse feature tracking technique to ensure accurate and consistent tracking. Delaunay triangulation is employed to represent the spatial distribution of features in an efficient way. Visual descriptors are categorised into individual behaviour descriptors and interactive descriptors to capture the temporal and spatial characteristics of crowd dynamics and behaviour, respectively. Neural networks are then utilised to classify these descriptors and pinpoint anomalies, making use of their strong learning capabilities. A significant component of our study is the assessment of how dimensionality reduction methods, particularly autoencoders and PCA, affect the feature set’s performance. This assessment aims to balance computational efficiency and detection accuracy. Tests conducted on benchmark crowd datasets highlight the effectiveness of our method in identifying anomalies. Our approach offers a nuanced understanding of crowd movement and patterns by emphasising both individual and collective characteristics. The visual and local descriptors facilitate high-level analysis by closely relating to semantic information and crowd behaviour. The analysis observed shows that this approach offers an efficient framework for crowd anomaly detection, contributing to improved crowd management and public safety. The proposed model achieves accuracy of 99.5 %, 96.1%, 99.0% and 88.5% in the UMN scenes 1, 2, and 3 and violence in crowds datasets, respectively.

A novel gearbox local fault feature extraction method based on quality coefficient and dictionary learning

The performance of sparse decomposition is directly determined by the similarity between impact atoms and the actual fault impact waveform. The shift-invariant K-singular value decomposition (K-SVD) dictionary learning algorithm is capable of training impact patterns from vibration signals collected by sensors to construct impact atoms, thereby extracting fault impact components from the vibration signals. However, the impact pattern training using the shift-invariant K-SVD algorithm is influenced by the presence of harmonics and white noise in the gear transmission system vibration signals. To solve the above problems, a novel gearbox local fault feature extraction method based on the quality coefficient and dictionary learning is proposed in this paper. Firstly, the original signal is decomposed into a series of intrinsic mode functions (IMFs) by empirical mode decomposition. Then, a new quality coefficient is proposed by comprehensively considering the intensity of the impact, harmonics and noise components in each IMF, as well as the degree of correlation with the original signal. The IMF with the largest quality coefficient is used to train the impact pattern and solve the sparse coefficients. Finally, the orthogonal matching pursuit algorithm is adopted to solve the sparse coefficients, which are used to reconstruct the fault impact response signal from the dictionary. Simulation and experimental analysis demonstrate the superior performance of the proposed method compared to other state-of-the-art methods.

MSFA: Multi‐stage feature aggregation network for multi‐label image recognition

AbstractMulti‐label image recognition (MLR) is a significant branch of image classification that aims to assign multiple categorical labels to each input. Previous research has focused on enhancing the learning of category‐related regional features. However, the potential impact of multi‐scale distributions in intra‐ and inter‐category targets on MLR tends to be neglected. Besides, semantic consistency for categories is restricted to be considered on single‐scale features, resulting in suboptimal feature extraction. To address the limitations of above, a Multi‐stage Feature Aggregation (MSFA) network is proposed. In MSFA, a novel local feature extraction method is suggested to progressively extract category‐related high‐resolution local features in both spatial and channel dimensions. Subsequently, local and global features are fused without additional up‐ and down‐sampling to enrich the scale diversity of the features while incorporating refined class‐specific information. Furthermore, a hierarchical prediction scheme for MLR is proposed, which generates classification confidence corresponding to different scales under hierarchical loss supervision. Consequently, the final output of the network comes from the joint prediction by the classifiers on multi‐scale features, ensuring a stronger feature extraction capability. The extensive experiments have been carried on VOC and MS‐COCO datasets, and the superiority of MSFA over existing mainstream methods has been verified.

Strain measurement with adaptive local feature extraction method based on special fiber OFDR system.

The optical fiber distributed strain sensor based on the optical frequency domain reflectometer (OFDR) preserves its dominant position in short-distance measurement fields with high spatial resolution, such as biomedical treatment, soft robot, etc. However, owing to the weak intensity of the Rayleigh backscattered signal (RBS) in the single-mode fiber (SMF) and complex computation, the large strain changes cannot be precisely and rapidly demodulated by the traditional cross-correlation method. In this work, the OFDR with backscattering enhanced optical fiber (BEOF) is proposed and demonstrated for fast and large strain measurement. By enhancing the RBS amplitude, the signal-to-noise ratio (SNR) is improved, resulting in a higher similarity between the reference signal and test signal, which is beneficial for the expansion of the strain measurement range. Moreover, the adaptive local feature extraction and matching (ALFEM) algorithm is presented and demonstrated, which replaces the traditional cross-correlation method for strain demodulation and fast measurement. On account of the enhancement ratio of BEOF, the dominant characteristic data segment can be extracted from whole wavelength data. In the experiments, the enhancing ratio of BEOF is designed as 10, resulting in the spatial resolution reaches 400µm and the strain measurement range is greatly increased to 4800µɛ. Further, the effectiveness of the ALFEM algorithm has been verified, in which the strain demodulation time is approximately 25% of that of the traditional method. This scheme fully exploits the enhancement characteristic of the BEOF and is also applicable to the systems based on other types of BEOF, different strain changes and sensing distances.

Research on Face Attribute Recognition Technology Based on Fine-Grained Features

Abstract This study explores fine-grained feature-based face attribute recognition techniques to enhance the accuracy of face recognition in low-resolution and complex environments. The article proposes a global feature extraction method and a local texture feature extraction method to extract global and regional features by enhancing feature reuse and information flow through dense connectivity and ShuffleNet V2 framework. Then, a multiscale feature exchange method is used to fuse different scale features to enhance the capture of detail information. Finally, efficient feature integration is achieved by the multiscale feature fusion method. Experimental results on the CK+ and FER2013 datasets show that the accuracy of this method on face expression recognition reaches 97.24% and 95.93%, respectively, and the average recognition accuracy in the face attribute recognition experiments on the CelebA dataset is 97.11%, which is significantly better than the comparison algorithm. In addition, the analysis of the recognition effect on low-resolution faces shows that this paper’s method achieves a recognition accuracy of 54.03% at a resolution of 15 × 15 and a high accuracy of over 99% at resolutions of 70 × 70 and above. These results show that the face attribute recognition technique based on fine-grained features proposed in this paper significantly improves recognition accuracy.

An Exploration of Face Recognition Methods Based on LBP Algorithm and PCA Analysis

Local Binary Pattern is an operator that can be effectively used for facial description, and its application field is relatively wide, which has made great contribution to the research and technical support related to face recognition. The traditional local texture feature extraction method of LBP operator is relatively single and cannot fully synthesize the problem of neighborhood correlation. Therefore, this paper proposes a face recognition method based on the combination of LBP algorithm and PCA principal component analysis method for the above problems. Through the application of this method, it can effectively weaken the influence of noise on the results and show the basic characteristic situation of the sub-block region, and the experiments of this method have confirmed that, compared with the use of PCA, LBP and other methods alone, the method in this study has a significant improvement in the recognition rate.

A deep learning-based local feature extraction method for improved image matching and surface reconstruction from Yutu-2 PCAM images on the Moon

Topographic reconstruction of the lunar surface or other planets is important to engineering applications and scientific research in a planetary exploration mission. The typical methods of terrain reconstruction are usually based on photogrammetry techniques. Structure-from-Motion (SfM) is one of the most effective and commonly used photogrammetric technologies that estimate three-dimensional structures from two-dimensional image sequences. To find correspondences and align photos, SfM approaches require invariant features, such as Scale Invariant Feature Transform (SIFT). The hand-crafted features, however, seriously degrade performance due to weak texture and low light conditions causing image matching failure, which will directly affect the accuracy and robustness of reconstruction. Robust local descriptors based on deep learning outperform hand-crafted descriptors since convolutional neural networks are more robust than hand-engineered representations. Therefore, a novel and robust deep learning-based local feature extraction method is proposed, comprising two branch networks integrated with attention mechanisms for generating reliable keypoints and descriptors, respectively. Furthermore, a 3D terrain surface reconstruction workflow is constructed by combining it with the modern advanced image matching method and SfM system. The effectiveness of the proposed method and the workflow were verified in experiments using Panoramic Camera (PCAM) images acquired from three waypoints explored by the Yutu-2 lunar rover during the Chang’e-4 mission. We also illustrate how our approach supports other applications, such as creating panoramic mosaics of surface imagery. This provides a new and powerful method for planetary terrain reconstruction at a high spatial resolution that can meet the requirements for rover navigation and positioning, as well as geological analysis of the Moon and other planets. The source codes developed in this study are openly available at https://github.com/Atypical-Programmer/Deep_Reconstruction_Workflow.

Unsupervised Feature Extraction for Various Computer-Aided Diagnosis Using Multiple Convolutional Autoencoders and 2.5-Dimensional Local Image Analysis

There are growing expectations for AI computer-aided diagnosis: computer-aided diagnosis (CAD) systems can be used to improve the accuracy of diagnostic imaging. However, it is not easy to collect large amounts of disease medical image data with lesion area annotations for the supervised learning of CAD systems. This study proposes an unsupervised local image feature extraction method running without such disease medical image datasets. Local image features are one of the key determinants of system performance. The proposed method requires only a normal image dataset that does not include lesions and can be collected easier than a disease dataset. The unsupervised features are extracted by applying multiple convolutional autoencoders to analyze various 2.5-dimensional images. The proposed method is evaluated by two kinds of problems: the detection of cerebral aneurysms in head MRA images and the detection of lung nodules in chest CT images. In both cases, the performance is high, showing an AUC of more than 0.96. These results show that the proposed method can automatically learn features that are useful for lesion recognition from lesion-free normal data, regardless of the type of image or lesion.

MIXTURE FEATURE EXTRACTION BASED ON LOCAL BINARY PATTERN AND GREY-LEVEL CO-OCCURRENCE MATRIX TECHNIQUES FOR MOUTH EXPRESSION RECOGNITION

Some academics struggle to recognize facial emotions based on pattern recognition. In general, this recognition utilizes all facial features. However, this study was limited to identifying facial emotions in a single facial region. In this study, lips, one of the facial features that can reveal a person's expression, are utilized. Using a combination of local binary pattern feature extraction (LBP) and grey level co-occurrence matrix (GLCM) methods and a multiclass support vector machine classification approach for feature extraction in facial images. The concept begins with image segmentation to create an image of a mouth. Experiments were also conducted for various tests, and the outcomes of these experiments revealed a recognition performance of up to 95%. This result was obtained through experiments in which 10% to 40% of the data were evaluated. These findings are beneficial and can be applied to expression recognition in online learning media to monitor the audience's condition directly.

Classification of Chest X-ray COVID-19 Images Using the Local Binary Pattern Feature Extraction Method

Background and Purpose: COVID-19, which started in December 2019, caused significant loss of life and economic losses. Early diagnosis of the COVID-19 is important to reduce the risk of death. Therefore, studies have increased to detect COVID-19 with machine learning methods automatically. Materials and Methods: In this study, the dataset consists of 15153 X-ray images for 4961 patient cases in three classes: Viral Pneumonia, Normal and COVID-19. Firstly, the dataset was preprocessed. And then, the dataset was given to the Cubic Support Vector Machine (Cubic SVM), Linear Discriminant (LD), Quadratic Discriminant (QD), Ensemble, Kernel Naive Bayes (KNB), K-Nearest Neighbor Weighted (KNN Weighted) classification methods as input data. Then, the Local Binary Model (LBP) texture operator was applied for feature extraction. Results: These values were increased from 94.1% (without LBP) to 98.05% using the LBP method. The Cubic SVM method's highest accuracy was observed in these two applications. Conclusions: This study demonstrates that the performance of the presented methods with LBP feature extraction is improved.

Nuclear cataract classification in anterior segment OCT based on clinical global–local features

Nuclear cataract (NC) is a priority ocular disease of blindness and vision impairment globally. Early intervention and cataract surgery can improve the vision and life quality of NC patients. Anterior segment coherence tomography (AS-OCT) imaging is a non-invasive way to capture the NC opacity objectively and quantitatively. Recent clinical research has shown that there exists a strong opacity correlation relationship between NC severity levels and the mean density on AS-OCT images. In this paper, we present an effective NC classification framework on AS-OCT images, based on feature extraction and feature importance analysis. Motivated by previous clinical knowledge, our method extracts the clinical global–local features, and then applies Pearson’s correlation coefficient and recursive feature elimination methods to analyze the feature importance. Finally, an ensemble logistic regression is employed to distinguish NC, which considers different optimization methods’ characteristics. A dataset with 11,442 AS-OCT images is collected to evaluate the method. The results show that the proposed method achieves 86.96% accuracy and 88.70% macro-sensitivity, respectively. The performance comparison analysis also demonstrates that the global–local feature extraction method improves about 2% accuracy than the single region-based feature extraction method.

Cognitive IoT Vision System Using Weighted Guided Harris Corner Feature Detector for Visually Impaired People

India has an estimated 12 million visually impaired people and is home to the world’s largest number in any country. Smart walking stick devices use various technologies including machine vision and different sensors for improving the safe movement of visually impaired persons. In machine vision, accurately recognizing an object that is near to them is still a challenging task. This paper provides a system to enable safe navigation and guidance for visually impaired people by implementing an object recognition module in the smart walking stick that uses a local feature extraction method to recognize an object under different image transformations. To provide stability and robustness, the Weighted Guided Harris Corner Feature Detector (WGHCFD) method is proposed to extract feature points from the image. WGHCFD discriminates image features competently and is suitable for different real-world conditions. The WGHCFD method evaluates the most popular Oxford benchmark datasets, and it achieves greater repeatability and matching score than existing feature detectors. In addition, the proposed WGHCFD method is tested with a smart stick and achieves 99.8% recognition rate under different transformation conditions for the safe navigation of visually impaired people.

An improved lightweight deep neural network with knowledge distillation for local feature extraction and visual localization using images and LiDAR point clouds

Visual localization nowadays is a research hotspot in computer vision and photogrammetry. It can provide meter level or higher localization accuracy under the conditions without GPS signals. However, achieving efficient, robust and high-accuracy visual localization under the condition of day-night changes is still challenging. To deal with this problem, we develop an improved lightweight deep neural network with knowledge distillation to efficiently extract deep local features from imagery while maintaining strong robustness for day-night visual localization. Furthermore, to further improve the accuracy of visual localization, we use aligned dense LiDAR point clouds and imagery collected by a new portable camera-LiDAR integrated device to build a prior map, and directly utilize the 2D-3D correspondences between 2D local feature points extracted by our lightweight network and 3D laser points retrieved from the prior map for localization. Moreover, we build our own ground truth point cloud dataset at 5 cm accuracy to evaluate the accuracy of the constructed prior map as well as a day-night dataset including prior map and verification data for the evaluation of the proposed visual localization method. The experimental results prove that our visual localization method achieves a balance between the efficiency and robustness while improving localization accuracy for day-night visual localization. In a comparison with a variety of state-of-the-art local feature extraction methods based on deep neural networks, our lightweight network has the least number of parameters (0.2 million) and reaches the highest feature extraction efficiency (92 frames per second), which is on par with that of the classic real-time ORB feature extraction method. Furthermore, our network remains competitive with other advanced deep local feature extraction networks in feature matching and day-night visual localization. In addition, evaluations performed on our own dataset demonstrate that our visual localization method using images and LiDAR point clouds provides a localization error of 1.2 m under the conditions of day-night changes, which is much smaller than those achieved by a state-of-the-art, purely visual localization method.

Multi-instance Finger Vein-based Authentication with Secured Templates

The illegitimate access to biometric templates is one of the major issues to be handled for authentication systems. In this work, we propose to use two instances of finger vein images which inherits the advantages of a robust multi-modal biometric authentication system without needing different sensors. Two local texture feature extraction methods are experimented on standard finger-vein datasets. Fused discriminating features with reduced dimension lowers down the system computational cost. A cancelable template protection scheme as Gaussian Random Projection based Index-of-Max is then applied for embedding privacy and security to the templates. Foremost template protection properties like revocability, non-invertibility and unlinkability are observed to be significantly obeyed by the proposed system with considerable authentication performance. Recognition performance of the proposed methods are compared with some previously executed finger vein systems and observed to be less complex and overperforming on the combined basis of authentication and template protection. Thus, the proposed system utilizes multiple evidence and provides a balanced performance with respect to authentication, template protection and computational cost.

A Face Recognition System Using ACO-BPNN Model for Optimizing the Teaching Management System.

The basic idea of face recognition technology is to compare the matching degree between the standard face image marked with identity information and the static or dynamic face collected from the actual scene, which includes two main research contents: face feature extraction and face feature recognition. Traditional identification generally proves who we are through certificates, passwords, or certificates plus passwords. With the development of science and technology, face recognition technology will occupy an increasingly important position. Inspired by the human brain, the artificial neural network (ANN) is an information extraction system based on imitating the basic function and structure of the human brain and abstracted by physical and mathematical research methods. Based on the traditional BP neural network model, this paper proposes an ant colony algorithm-enabled BP neural network (ACO-BPNN) model and applies it to face recognition. Experimental results show that, similar to other face recognition techniques, the facial feature location needs to adapt to various changes of faces to the maximum extent, so the recognition and classification effect of the whole face feature extracted from the whole face image on the changes of such partial areas is not good, while the local feature extraction method based on ACO-BPNN can achieve a good recognition and classification effect.

Pengenalan Logo Kendaraan Menggunakan Metode Local Binary Pattern dan Random Forest

The vehicle logo is one of the features that can be used to identify a vehicle. Even so, a lot of Intelligent Transport System which are developed nowadays has yet to use a vehicle logo recognition system as one of its vehicle identification tools. Hence there are still cases of traffic crimes that haven't been able to be examined by the system, such as cases of counterfeiting vehicle license plates. Vehicle logo recognition itself could be done by using various feature extraction and classification methods. This research project uses the Local Binary Pattern feature extraction method which is often used for many kinds of image recognition systems. Then, the classification method used is Random Forest which is known to be effective and accurate for various classification problems. The data used for this study were as many as 2000 vehicle logo images consisting of 5 brand classes, namely Honda, Kia, Mazda, Mitsubishi, and Toyota. The results of the tests carried out obtained the best accuracy value of 88.89% for the front view logo image dataset, 77.03% for the side view logo image dataset, and 83% for the dataset with both types of images.

Robust video content analysis schemes for human action recognition.

Action recognition is a challenging time series classification task that has received much attention in the recent past due to its importance in critical applications, such as surveillance, visual behavior study, topic discovery, security, and content retrieval. The main objective of the research is to develop a robust and high-performance human action recognition techniques. A combination of local and holistic feature extraction methods used through analyzing the most effective features to extract to reach the objective, followed by using simple and high-performance machine learning algorithms. This paper presents three robust action recognition techniques based on a series of image analysis methods to detect activities in different scenes. The general scheme architecture consists of shot boundary detection, shot frame rate re-sampling, and compact feature vector extraction. This process is achieved by emphasizing variations and extracting strong patterns in feature vectors before classification. The proposed schemes are tested on datasets with cluttered backgrounds, low- or high-resolution videos, different viewpoints, and different camera motion conditions, namely, the Hollywood-2, KTH, UCF11 (YouTube actions), and Weizmann datasets. The proposed schemes resulted in highly accurate video analysis results compared to those of other works based on four widely used datasets. The First, Second, and Third Schemes provides recognition accuracies of 57.8%, 73.6%, and 52.0% on Hollywood2, 94.5%, 97.0%, and 59.3% on KTH, 94.5%, 95.6%, and 94.2% on UCF11, and 98.9%, 97.8% and 100% on Weizmann. Each of the proposed schemes provides high recognition accuracy compared to other state-of-art methods. Especially, the Second Scheme as it gives excellent comparable results to other benchmarked approaches.

Learning to locate for fine-grained image recognition

In this paper, we propose an end-to-end weakly supervised method for fine-grained image recognition called bounding box-part location method(BBPL), which can locate the object and part precisely without part annotations. The proposed method includes three modules: object detection, ObjectMask, and classification. Firstly, the object detection module predicts the bounding boxes, and the predicted bounding boxes are employed to generate a mask through ObjectMask module. The generated mask can suppress the background interference during recognition. Secondly, the classification module can be further divided into two branches, which are global feature classification and local feature classification. In global feature classification branch, global feature is extracted to get global classification result. While in local feature classification branch, salient point is first detected through our novel salient point detection module, which can greatly reduce the consuming-time compared with the most existing local feature extraction methods. Further, the local feature is extracted in these detected salient points, and local classification result is obtained by local feature classification branch. Finally, we get the final result by fusing the results of two classification branches together. With experiments on three widely used fine-grained image recognition datasets (CUB-200-2011, Stanford Cars, Stanford Dogs), our method can achieve the state-of-the-art performance.

Unsupervised Doppler Radar Based Activity Recognition for e-Healthcare

Passive radio frequency (RF) sensing and monitoring of human daily activities in elderly care homes is an emerging topic. Micro-Doppler radars are an appealing solution considering their non-intrusiveness, deep penetration, and high-distance range. Unsupervised activity recognition using Doppler radar data has not received attention, in spite of its importance in case of unlabelled or poorly labelled activities in real scenarios. This study proposes two unsupervised feature extraction methods for the purpose of human activity monitoring using Doppler-streams. These include a local Discrete Cosine Transform (DCT)-based feature extraction method and a local entropy-based feature extraction method. In addition, a novel application of Convolutional Variational Autoencoder (CVAE) feature extraction is employed for the first time for Doppler radar data. The three feature extraction architectures are compared with the previously used Convolutional Autoencoder (CAE) and linear feature extraction based on Principal Component Analysis (PCA) and 2DPCA. Unsupervised clustering is performed using K-Means and K-Medoids. The results show the superiority of DCT-based method, entropy-based method, and CVAE features compared to CAE, PCA, and 2DPCA, with more than 5\%-20\% average accuracy. In regards to computation time, the two proposed methods are noticeably much faster than the existing CVAE. Furthermore, for high-dimensional data visualisation, three manifold learning techniques are considered. The methods are compared for the projection of raw data as well as the encoded CVAE features. All three methods show an improved visualisation ability when applied to the encoded CVAE features.

A local feature extraction method for UAV-based image registration based on virtual line descriptors

Image local feature extraction is extensively utilized in the field of photogrammetry where the spatial distribution of features is important in high-quality image matching, particularly in high-resolution unmanned aerial vehicle (UAV) image registration. Presently, the spatial distribution problems are considered in some local feature extraction methods, though these methods are designed for point descriptors. Line descriptors are more robust to repetitive patterns compared to point descriptors and have attracted extensive attention in recent years. Hence, a feature extraction method is designed in this paper for line descriptors based on the K-connected virtual line descriptors matching method. Using the four measures, the quality of local features is quantified, and a regular gridding strategy based on the quality of local features is applied in the feature selection procedure. The proposed feature extraction method was successfully applied to match various simulated and real UAV-based images. Based on the experimental results using real images, it is indicated that two evaluation criteria, namely the spatial distribution quality of features and the number of correct matches, are improved to at least 12% and 15%, respectively, for verifying the capability of the proposed method to enhance matching performance.