HOG Features Research Articles

Automatic Facial Expression Recognition Using Modified LPQ and HOG Features with Stacked Deep Convolutional Autoencoders

Automatic Facial Expression Recognition Using Modified LPQ and HOG Features with Stacked Deep Convolutional Autoencoders

Method for enhanced gesture recognition under low light conditions based on wearable mechanoluminescence sensors

Gestures, as one of the key modes of human–computer interaction, are currently identified and processed mainly through two methods: machine vision imaging and wearable stress/strain sensing.However, the former has high requirements for lighting and background.This paper innovatively proposes a feature fusion algorithm that combines mechanoluminescent sensors for gesture recognition under low-light conditions. Unlike existing methods, this algorithm creates a system that integrates mechanoluminescent sensors for visual gesture recognition. By triggering the sensor’s luminescence through mechanical stress, it provides additional light source information to enhance image quality in low-light environments.For this purpose, this study employed the fluorescent emission properties of ZnS:Cu to fabricate a mechanoluminescence sensor. Its mechanoluminescence response characteristics were measured. Based on these findings, an algorithm for the fusion of gesture information features was developed for low illumination backgrounds. This algorithm decomposes the gesture visual images through YCbCr, extracting the HOG features of the Cr channel (Cr-HOG). Simultaneously, feature fusion is performed using the pooling layer (VGG16-Conv5-pool) following the convolutional layer (Conv5-3) in the VGG16 model. A one-to-one composite SVM classifier is constructed to complete the training and testing of the gesture recognition model. Ultimately, the feasibility of this enhanced gesture recognition method was validated through unmanned vehicle control experiment. Experimental results demonstrated that in environments with an illuminance level below 10 LUX measured by a light meter, the average recognition rate of the wearable mechanoluminescence sensor feature fusion algorithm reached 95.82 %. This is 32.24 % higher than the recognition rate of the single Cr-HOG feature classification and 28.27 % higher than the single VGG16-Conv5-pool feature classification. Compared with feature extraction using ResNet-50 and DenseNet-121 networks, the recognition rate for classifying gestures under low-light conditions is approximately 30 % higher.Compared with other image enhancement recognition methods, the recognition can be increased by up to 30.88 %.

Robust Hand Gesture Recognition Using HOG Features and machine learning

Robust Hand Gesture Recognition Using HOG Features and machine learning

Test-Time Adaptation via Style and Structure Guidance for Histological Image Registration

Image registration plays a crucial role in histological image analysis, encompassing tasks like multi-modality fusion and disease grading. Traditional registration methods optimize objective functions for each image pair, yielding reliable accuracy but demanding heavy inference burdens. Recently, learning-based registration methods utilize networks to learn the optimization process during training and apply a one-step forward process during testing. While these methods offer promising registration performance with reduced inference time, they remain sensitive to appearance variances and local structure changes commonly encountered in histological image registration scenarios. In this paper, for the first time, we propose a novel test-time adaptation method for histological image registration, aiming to improve the generalization ability of learning-based methods. Specifically, we design two operations, style guidance and shape guidance, for the test-time adaptation process. The former leverages style representations encoded by feature statistics to address the issue of appearance variances, while the latter incorporates shape representations encoded by HOG features to improve registration accuracy in regions with structural changes. Furthermore, we consider the continuity of the model during the test-time adaptation process. Different from the previous methods initialized by a given trained model, we introduce a smoothing strategy to leverage historical models for better generalization. We conduct experiments with several representative learning-based backbones on the public histological dataset, demonstrating the superior registration performance of our test-time adaptation method.

Tri-stage offline Telugu character recognition system based on fusion of HOG and ULBP

The domain of pattern recognition has experienced a notable upsurge in the prevalence of handwritten character recognition, which is a research area that has been widely recognized for decades.However, this task presents significant challenges, especially for Indian languages, due to the vast array of character shapes and curved linear structures they possess. Despite some research having been conducted on Indian languages such as Kannada, Hindi, Tamil Devanagari, and Telugu, among others, the field is still in its nascent stages of development. The nascent phase of handwritten Telugu script presents a promising opportunity for further investigation and exploration. We introduce a Tri-stage Telugu handwritten recognition system that utilizes hyper parameters of k- NN, SVM, and Random Forest based on the fusion of HOG and ULBP features. By implementing a Tri-stage classifier in a sequential manner, the system's recognition accuracy is significantly enhanced.The proposed Tri-stage classifier achieves exceptionally high accuracies on Telugu handwritten character HP datasets.

Semantic and Context Understanding for Sentiment Analysis in Hindi Handwritten Character Recognition Using a Multiresolution Technique

The rapid growth of Web 2.0, which enables people to generate, communicate, and share information, has resulted in an increase in the total number of users. In developing countries, online users’ sentiment influences decision-making, social views, individual consumption decisions, and entity quality monitoring. As a result, more accurate sentiment analysis, particularly in their native language such as Hindi, is preferred over crude binary categorization. This is because of the abundance of web-based data in Indian languages such as Hindi, Marathi, Kannada, Tamil, and so on. Analyzing this data and recovering valuable and relevant information from handwritten text has become extremely important. Despite years of research and development, no optical writing recognition (OCR) system has ever been certified as completely reliable. The first step in any pattern recognition system is feature selection. In many fields, feature selection is studied as a combinatorial optimization problem. The primary goal of feature selection is to reduce the number of redundant and ineffective traits in the recognition system. This feature selection is used to maintain or improve the performance of the classifier used by the recognition system: A support vector machine (SVM) technique could be used to solve this character recognition problem. The Hindi character recognition system recognizes Hindi characters by employing morphological operations, edge detection, HOG feature extraction, and an SVM-based classifier. The proposed model outperformed the current state-of-the-art method, achieving an accuracy of 96.77%.

Integrated global and local feature extraction and classification from computerized tomography (CT) images for lung cancer classification.

Despite being the second most often diagnosed form of cancer, lung cancers are rarely found in the general population. It is proposed in this study to employ amethodology of extracting both global and local features from CT scan images for the identification of lung cancer. Data gathering, globalised and localised training as well as testing the model are all part of this structure. This study makes use of 800 CT scan images. Images are pre-processed by warping and cropping in advance of the global testing step. Each image is represented by afeature vector employing eight distinct types of image characteristics, which are taken from the images. After creating feature vectors, three machine learning methods are employed to create detection models. Every medical image has been partitioned over aseries of simple divisions throughout the training and testing process locally. To describe each block, feature vectors are derived from the image features that worked effectively in the general phase of the experiment. Similar extracted features are then used to build detection systems for all picture blocks using the learning strategies that were effective in the global stage. SVM using Haar Wavelet characteristics had an accuracy, sensitivity, and specificity of 89%, 90%, and 89%, respectively. One might get 90%‑accurate results with SVM and 91%‑sensitive and 91%‑specific results using SVM plus HOG features. Finally, the utilisation of SVM with Gabor Filter characteristics achieved the greatest correctness, specificity, and sensitivity values, particularly 87%, 86%, and 87%, respectively (Tab. 3, Fig. 7, Ref. 18). Keywords: feature extraction, support vector machine, lung cancer, classification, machine learning.

Research on Teaching Characteristics of Innovative Civic and Political Education in Colleges and Universities Based on HOG Feature Extraction

Abstract With the rapid development of information technology and scientific education, the work of ideological and political education in colleges and universities has also undergone important changes. In this paper, the gradient value of the learning behavior data of college ideological and political students is extracted by HOG features, and the extracted student behavior data is inputted into the learning dashboard to be analyzed, and the pedagogical feedback of ideological and political learning is obtained. Two classes, category features and numeric features, are categorized by the Kprototype algorithm for the educational dataset. Moreover, through the learning dashboard, learners’ learning-related data is graphically visualized in multiple dimensions to explore the factors that affect the effect of learning Civics and Politics in colleges and universities. Finally the teaching of college ideology and politics in University G is used as the metadata of the study to analyze the utility of the learning dashboard innovative teaching tool in teaching college ideology and politics. The results show that the effect value of the learning dashboard on the learning effect of Civics is 0.804, and the effect value on the emotion of Civics learners is 0.385, respectively, which both have a large degree of positive influence. The use of a learning dashboard can positively impact the learning behavior of Civics learners.

Research on the Application of Intelligent Recognition Technology in the Prediction of Violation Behaviour at Electricity Work Sites

Abstract To realize the safe operation of electric power site, this paper proposes an intelligent recognition technology to automatically identify violations. This study successfully constructs a face detection model for power operation sites by combining deep convolutional neural networks and target detection algorithms. A three-way connected feature pyramid structure containing a neuron self-processing module is adopted, and an accuracy test is completed using a Tri-FPN-based target detection network, significantly improving recognition accuracy. In this paper, we also utilized the on-site images collected by video surveillance equipment, combined with CNN algorithm and HOG feature extraction technology to effectively identify the violations and provide early warning of the breaches of the personnel at the power operation site. MAP curves evaluated the detection performance, and the results showed that the head recognition rate was up to 0.9913, and the accuracy rate of all violations exceeded 0.9350.The high accuracy of CNN-based feature fusion extraction algorithm in the recognition of violations of personnel at the site of electric power operation provides effective technical support to ensure personnel safety.

Research on the Application of Behavioral Image Feature Capture in Basketball Game Video

<p>In order to realize intelligent image recognition of foul behavior in basketball games, this paper designs a feature capture method of video foul behavior based on improved bilateral filtering algorithm. Adaptive bilateral filtering is used to denoise the video image, and the optical flow feature and HOG feature of the behavior in the denoised image are obtained by image combination feature extraction method, which is fused to a combined eigenvector. The combined features were taken as the target recognition samples, and the multi-back propagation neural network was used to identify the foul behavior features. The particle filter was used to capture the video features and identify the location of the video behavior features. The experimental results show that this method can accurately capture the changes of video behavior characteristics, and has applicable performance in the identification of foul behavior in basketball matches.</p> <p> </p>

Two-Stage Traffic Sign Classification System

Traffic Sign Recognition (TSR) has received widespread attention due to increased traffic accidents caused by a failure to recognize road traffic signs. Most Traffic Sign Recognition remains challenging due to the illumination of natural scenes, and the diversity in traffic signs [31]. To solve the problem, we proposed a two-stage classification system to overcome the above-mentioned challenges and classify the traffic sign boards efficiently. During the first phase, colour based histogram equalizations applied to eliminates the illumination factors and then traffic sign boards are classified into many shapes, which includes circular, triangular, hexagonal, and diamond traffic signs boards based on the modified HOG feature descriptors and various machine learning classifiers are compared [12]. In the second phase, circular and triangular traffic signs are further categorized into specific subclasses using circular and triangular signs trained Convolutional Neural Network (CNN) respectively to overcome intra-class variation among the similar sign board. Experimentation was conducted on German Traffic Sign Recognition Benchmark dataset (GTSRB) [30] and the results show 99.96% top-1 accuracy during the first stage of classification. In the second stage, there is accuracy of 99.40% for circular traffic signs and 97.50% for triangular traffic signs.

The positive significance of ideological and political teaching in colleges and universities based on HOG feature extraction to college students’ innovation and entrepreneurship

Abstract This paper designs an intelligent ideological teaching model integrating innovation and entrepreneurship education under the guidance of the “6C+6E” theory. In this model, the features are first extracted by the convolution operation of the student face image through the Gabor algorithm. The features extracted by HOG algorithm based on integral graph are concatenated to get the fusion of student face features. Then SVM algorithm is used to classify and recognize the student’s facial features. In the classroom, applying this ideological and political teaching mode, the overall concentration situation of the class reached 0.9988 points at the best time. In contrast, the concentration of the traditional innovation and entrepreneurship teaching classroom were all lower than 0.4 points, and the highest rate of students’ heads up in the overall classroom reached 98.19%. After the implementation of the teaching, there is a significant change in the mean score of students’ awareness of entrepreneurial value (P=0.011<0.05) and understanding of entrepreneurial business opportunities (P=0.016<0.05), which has increased by 50.45% and 36.41% relative to the pre-implementation of the teaching. The Civics teaching model constructed in this paper lays the foundation for the effective development of innovation and entrepreneurship education in colleges and universities and provides a reference basis for the improvement of students’ innovation and entrepreneurship awareness.

Research on Building a New Mode of Civic and Political Education by Image Recognition Technology for Popularizing Socialist Core Values

Abstract Given the current state of ideological education in colleges and universities, we propose leveraging image recognition technology to foster innovation and digital advancement in this mode of education. The weighted average method is used to greyscale the video images of students’ classrooms, and after binarizing the images using the OTSU algorithm, interference factors are avoided by denoising and angle correction. We combine the motion history image (MHI) and HOG features to create a new feature that depicts Civics classroom behavior. The SVM classifier then performs the classification of classroom behavior, ultimately constructing the Civics classroom behavior recognition model based on HOC+MHII+SVM. This paper’s model and research data are combined to conduct an example analysis of Civics education in colleges and universities. It can be concluded from the analysis that, in the case of simultaneously adopting SVM for recognition and classification, it is found that HOC+MHI feature fusion has a more significant effect on classroom behavior recognition than FPN+Multi-layer CBAM, and the difference of its various indexes is maintained within the range of 0.01~0.08. In addition, the experimental group has a significant difference relationship in the four dimensions of demonstrating the concept of honor and shame, the concept of national development, the value of development, and the value of society, P<0.05. This study provides theoretical references for research on the innovation of the ideological and political education model and is conducive to the promotion of cultivating the formation of students’ socialist core values.

A Ranked-Aware GA with HoG Features for Infant Cry Classification

A Ranked-Aware GA with HoG Features for Infant Cry Classification

Face Recognition Algorithm based on Optimal Fusion of LBP and HOG Features

Face Recognition Algorithm based on Optimal Fusion of LBP and HOG Features

Data Augmentation for Genetic Programming-Driven Late Merging of HOG and Uniform LBP Features for Texture Classification

The categorization of texture images requires the identification and extraction of meaningful keypoints, a crucial step in ensuring the precise representation of textured images. The literature has introduced numerous descriptors in order to detect and capture both local and global texture characteristics. These descriptors vary in their effectiveness depending on the specific application. However, it is generally accepted that they complement each other by compensating for their strengths and weaknesses. Therefore, many approaches focused on combining different types of image descriptors generating robust features that are invariant to image transformations. These solutions mostly focused on one way to deal with information and faced more problems as they need human intervention and a large set of data. The acknowledged benefits of combining multiple types of descriptors are accompanied by challenges specially. These challenges arise due to differences in properties, such as locality and sparsity, as well as the heterogeneity exhibited by generated features. To address this issue comprehensively, we present an approach that benefits from genetic programming techniques to generate and combine two distinct texture classifiers. It incorporates histograms of oriented gradients and local binary patterns descriptors, which capture different textures. To effectively fuse the results of both classifiers, the proposed approach employs a late fusion process along with data augmentation methods while using a limited amount of training data. To evaluate the performance of the proposed approach in texture image classification tasks, we have conducted extensive experiments on six challenging datasets encompassing various variations. We have also investigated its performance in a cross dataset problem where the model has been trained on instances of a dataset before being tested on samples of another dataset. The obtained results clearly demonstrate that the proposed approach surpasses other relevant low-level approaches, as well as existing GP-based and CNN methods specifically designed for describing and classifying textures. Thanks to its ability to simultaneously leverage multiple descriptors, the suggested solution shows a high potential for real-world applications, particularly in handling various image changes with robustness.

Research on Digital Construction and Design of Minority Clothing Based on Multivariate Statistical Analysis

Abstract Digitizing minority costumes is a modern way to protect non-heritage. In this paper, after analyzing the Hu invariant moment features and HOG features, we propose a feature extraction method for ethnic minority costumes that integrates the fusion features of HOG features and Hu invariant moment features. On this basis, the structure of the extracted multivariate feature data is analyzed, and the KPCA method of multivariate statistical analysis is adopted to process the multivariate data of ethnic minority costumes and digitally construct and design the ethnic minority costumes. In addition, experiments were conducted to explore the effect of minority costumes on feature extraction and the effect of digitization construction and design. The results show that the extraction accuracy of six ethnic costumes is stable at 0.75, 0.45, 0.50, 0.75, 0.72, and 0.68, respectively, and its Top-10 accuracy reaches 0.61 at the highest level, while the accuracy of retrieving only some of them is less than 0.40, and the effects of constructed ethnic costume outlines are all around 0.9. The output results of ethnic dresses constructed based on this research are roughly similar to the original ethnic dresses, which is conducive to the inheritance of ethnic dress culture.

Physiological characteristics inspired hidden human object detection model

The current target detection algorithms provide the unsatisfactory performance on the task of detecting hidden human targets. In this study, we put forward the physiological characteristics inspired hidden human object detection model considering the spatio-temporal physiological features and their interdependent relationships. The experimental results of homemade hidden human object dataset demonstrate that the proposed model generates the detection accuracy of 64%, 44%, and 54% for indoor scene, outdoor scene, and overall dataset, respectively, outperforming the YOLO v4 models and the models based on HOG, LBP, and Haar features, with at least 22% promotion in detection accuracy. The ablation experiments indicate the effectiveness of each module of the method. In the future, the proposed model or the corresponding modeling idea has the potential to be applied to military rescue, public security investigation and other fields. Once the paper is accepted, we will make the homemade dataset publicly available.

Evaluation of infrared image detection based on HOG feature extraction in swimmer training effectiveness

Evaluation of infrared image detection based on HOG feature extraction in swimmer training effectiveness

Comparison of machine learning models on breast cancer risk prediction challenge

Breast cancer is a malignant tumor that poses a serious risk to women's life and wellbeing. To make matters worse, this cancer is less symptomatic in its early stages. It is not easily diagnosed through traditional means. The topic of this essay is to investigate machine learning for the determination of breast cancer.. The methods based on machine learning are as followed: automated nuclear section segmentation model, BCRecommender System, DNNs, and computer-aided diagnosis model (CADM). The methods studied are all based on the BreCaHad dataset and use a comparative metric.To measure the performance of each model, the accuracy, F1-score, specificity, and precision are used. The result shows that the approaches based on machine learning work well in diagnosing breast carcinoma, with high accuracy. Most of them have a percentage over 90% in accuracy and some of them are even higher than 95%. However, some of the models work poorly, such as layer 1 of BCRecommender with 61.06% accuracy and EfficientNetB0 with 72.96%. With every aspect taken into consideration, computer-aided diagnosis (CADM: Using combined features of HOG, WPD, ResNet as well as PCA + SVM) has the greatest advantage in diagnosing BC.