Histogram Of Oriented Gradients Features Research Articles

Human Pose and Path Estimation from Aerial Video Using Dynamic Classifier Selection

We consider the problem of estimating human pose and trajectory by an aerial robot with a monocular camera in near real time. We present a preliminary solution whose distinguishing feature is a dynamic classifier selection architecture. In our solution, each video frame is corrected for perspective using projective transformation. Then, two alternative feature sets are used: (i) Histogram of Oriented Gradients (HOG) of the silhouette, (ii) Convolutional Neural Network (CNN) features of the RGB image. The features (HOG or CNN) are classified using a dynamic classifier. A class is defined as a pose-viewpoint pair, and a total of 64 classes are defined to represent a forward walking and turning gait sequence. Our solution provides three main advantages: (i) Classification is efficient due to dynamic selection (4-class vs. 64-class classification). (ii) Classification errors are confined to neighbors of the true view-points. (iii) The robust temporal relationship between poses is used to resolve the left-right ambiguities of human silhouettes. Experiments conducted on both fronto-parallel videos and aerial videos confirm our solution can achieve accurate pose and trajectory estimation for both scenarios. We found using HOG features provides higher accuracy than using CNN features. For example, applying the HOG-based variant of our scheme to the 'walking on a figure 8-shaped path' dataset (1652 frames) achieved estimation accuracies of 99.6% for viewpoints and 96.2% for number of poses.

Human activity recognition in egocentric video using HOG, GiST and color features

With the rapid increase in digital technology, most research areas are involved in human activity recognition, which can help to analyze the activities of patients. A novel approach for human activity recognition in egocentric video has been invoked in this research article. Generally, only the objects are identified, but the actions are not recognized. With this motivation and new trends, this paper presents an efficient technique to recognize the activities. In our approach, first the various activity dataset is trained, and the feature vector values are stored for various activities, which are applied to the testing inputs. Here, we use a filtering technique, i.e., a median filter followed by a segmentation method using watershed and feature extraction, such as a Histogram of Oriented Gradient (HOG), Color and GiST and a combination of all Features. Features are reduced using a genetic algorithm, and classification is done using Support Vector Machine (SVM) and a Random Forest classifier. The experimental results demonstrate that the Random Forest classifier outperformed the SVM classifier.

Efficient facial expression recognition using histogram of oriented gradients in wavelet domain

Facial expression recognition plays a significant role in human behavior detection. In this study, we present an efficient and fast facial expression recognition system. We introduce a new feature called W_HOG where W indicates discrete wavelet transform and HOG indicates histogram of oriented gradients feature. The proposed framework comprises of four stages: (i) Face processing, (ii) Domain transformation, (iii) Feature extraction and (iv) Expression recognition. Face processing is composed of face detection, cropping and normalization steps. In domain transformation, spatial domain features are transformed into the frequency domain by applying discrete wavelet transform (DWT). Feature extraction is performed by retrieving Histogram of Oriented Gradients (HOG) feature in DWT domain which is termed as W_HOG feature. For expression recognition, W_HOG feature is supplied to a well-designed tree based multiclass support vector machine (SVM) classifier with one-versus-all architecture. The proposed system is trained and tested with benchmark CK+, JAFFE and Yale facial expression datasets. Experimental results of the proposed method are effective towards facial expression recognition and outperforms existing methods.

Toward Always-On Mobile Object Detection: Energy Versus Performance Tradeoffs for Embedded HOG Feature Extraction

This paper studies the effects of front-end imager parameters on object detection performance and energy consumption. A custom version of histograms of oriented gradient (HOG) features based on 2-b pixel ratios is presented and shown to achieve superior object detection performance for the same estimated energy compared with conventional HOG features. A front-end hardware implementation capable of extracting these features at multiple scales is proposed, and a system-level energy analysis is performed. This energy analysis suggests a potential $19\times $ reduction in I/O energy and a $3.3\times $ reduction in back-end detection energy compared with conventional object detection pipelines.

Different Approaches for Face Authentication as Part of a Multimodal Biometrics System

This paper describes different approaches for the face authentication from the features and classification abilities point of view. Authors compare two types of features - Histogram of Oriented Gradients (HOG) and Local Binary Patterns (LBP) including their combination. These parameters are classified using Multilayer Neural Network (MLNN) and Support Vector Machines (SVM). Face authentication consists of several steps. The first step contains Viola-Jones algorithm for face detection. Authors resize the detected face for a fixed vector and afterwards, it is converted into grayscale. Next, feature extraction with a simple Min-Max normalization is applied. Obtained parameters are evaluated by classifiers and for each detected face, authors get posterior probability as the output of the classifier. Different approaches for face authentication are compared with each other using False Acceptance Rate (FAR), False Rejection Rate (FRR), Equal Error Rate (EER), Receiver Operating Characteristic (ROC) and Detection Error Tradeoff (DET) curves. The results are verified with AR Face Database and elaborated in a feature extraction and classifier design point of view. Best results were achieved by HOG feature for SVM classifier. Detailed results are listed in the text below.

Sparsity-based inverse halftoning via semi-coupled multi-dictionary learning and structural clustering

Inverse halftoning is the restoration of a continuous-tone image from its halftone version, which is a critical process for halftone transform, digital archive management and high precision identification of halftone. In this paper, a novel inverse halftoning method based on semi-coupled multi-dictionary learning is proposed to address the cross-style image restoration from halftone images to continuous-tone images. By using semi-coupled multi-dictionary learning, multiple dictionary pairs and their corresponding mapping functions between continuous-tone image and its halftone version could be simultaneously learned. The learned multiple dictionary pairs can well represent the structure characteristics of halftone images and continuous-tone images, respectively. In addition, the mapping functions learned by semi-coupled manner can bridge the gap between the two different style images of halftone image and continuous-tone image. Unlike the existed methods, the proposed method could effectively relax the assumption of the same sparse coding coefficients in coupled dictionary learning. To obtain more accurate mapping functions, a structural clustering method for cross-style image patches is proposed by using SUSAN (smallest univalue segment assimilating nucleus) filtering and HOG (histogram of oriented gradient) features, which can capture the similar structure features from halftone images and continuous-tone images, and thus improve the classification accurate rate of halftone image patches. The experimental results demonstrate that the proposed method can restore higher quality continuous-tone images than that produced by the state-of-the-art methods, which not only reduce the screen noise in smooth regions, but also provide well fine details and clear edges.

Multi-feature fusion tree trunk detection and orchard mobile robot localization using camera/ultrasonic sensors

In a semi-structured orange orchard, multi-tree trunk detection is an effective method for mobile robot localization. However, because of the complex background of the natural orchard environment, dwarf orange tree trunks are easy to misrecognize. In this paper, we present a novel tree trunk detection method based onmultiplecameras and ultrasonicsensors integration technology. These devices are integrated intoa single organicmechanical structure that can rotate to detect the surrounding orchard environment such that they can reduce the non-detectionzone. Multi-feature fusion will be used in this study. First, histograms of oriented gradient (HOG) and support vector machine (SVM) are used to train an initial tree trunk classifier. Next, the gray scale histogram features of the tree trunk and non-trunk images are extracted to optimize the classifier. Finally, the Roberts cross edge detector is used to extract the trunk’s gradient histogram features, which will improve the recognition accuracy of the classifier. The orange tree trunk recognition experiments exhibited a recall rate and accuracy of 92.14% and 95.49%, respectively. On this basis, ultrasonic sensors are used to get the location data of the trunks and a moving average filter is used to reduce the error of mobile robot localization. The experiment showed that the average localization error was approximately 62 mm (2.5%), and the robot moved stably and precisely along the road of the semi-structured orange orchard.

Robust face pseudo-sketch synthesis and recognition using morphological-arithmetic operations and HOG-PCA

In this paper, we propose a simple but yet effective method for synthesizing a pseudo face sketch (pseudo-sketch) from a photo, to be used for face recognition based on sketches drawn by a forensic artist. In contrast to current methods, the proposed method does not require training samples while fairly maintains the salient facial features as the artist do. We also propose a matching method on the basis of the Histograms of Oriented Gradients (HOG) descriptor and Principal Component Analysis (PCA), called HOG-PCA, to handle the similarities between a forensic sketch and a synthesized pseudo-sketch. In this method, we first extract the HOG features for the sketch and pseudo-sketch at regular grid and overlapped patches. The PCA is then applied to address the redundancy in feature representation due to several overlapped patches. Finally, the Nearest Neighbors Classifier (NNC) with the cosine distance is used to classify the sketch and pseudo-sketch pairs as matched or mismatched. Experimental results on CUHK and AR face sketch databases demonstrate that our proposed methods outperform state-of-the-art methods.

Multi-vehicle detection algorithm through combining Harr and HOG features

In order to achieve a better performance of detection and tracking of multi-vehicle targets in complex urban environment, we propose a two-step detection algorithm based on combining the features of Harr and Histogram of Oriented Gradients (HOG). This algorithm makes full use of HOG characteristic advantages for target vehicles, i.e., the good descriptive ability of HOG feature, and the prospect region of interest (ROI) can be extracted using Harr features. Moreover, the extracted HOG features from the ROI target area can be selected through applying the cascade structured AdaBoost classifier features and target area classification. Precise target can be further extracted by using support vector machine (SVM). Experimental results using video collected from real world scenarios are provided, showing that the proposed method possesses higher detecting accuracy and time efficiency than the conventional ones, and it can detect and track the multi-vehicle targets successfully in complex urban environment.

A hybrid technique for gender classification with SLBP and HOG features

Gender classification from facial images plays a significant role in biometric technology viz. gender medicine, surveillance, electronic banking system and human computer interaction. However, it has many challenges due to variations of pose, expression, aging, race, make-up, occlusion and illumination. In the proposed system, spatially enhanced local binary pattern (SLBP) and histogram of oriented gradients (HOG) are extracted to classify the human gender with SVM classifier. This hybrid feature selection has increased the power of the proposed system due to its representation of texture micro-patterns and local shape by capturing the edge or gradient structure form the image. The gender classification accuracy is studied by using the local feature representation of the face images separately and also these features are concatenated to provide a better recognition rate. The combination of two different local descriptors provides good representation of face image and this is given to SVM classifier which classifies as male or female. Also, the proposed work is compared with other two traditional classifiers such as k-nearest neighbor and sparse representation classifier. The performance was evaluated on FERET and LFW database. The highest classification accuracy 99.1% is achieved on FERET database and 95.7% is achieved on LFW database by applying cubic SVM with fusion of SLBP and HOG features.

RETRACTED ARTICLE: An effiecient human tracking system using Haar-like and hog feature extraction

The traditional human tracking systems are often prone to problems which can be deduced either from the subject, camera or background movements which includes major changes in posture, appearance, clothing and lighting of the background. The work introduced here proposes a system for extracting and tracking objects from a video sequence by initializing the process of feature extraction of the selected area of the human. The need for cascade extraction by using Haar-like features, is to basically decrease the use of crude or raw pixel values and then make classification easier. The major issue here is the problem in extracting the required finite set of input for identifying the characteristics necessary for domain encoding rule In the proposed work the major aim is to address the tracking procedure by facilitating the exclusive histogram of oriented gradient (HOG) methodology, using which only the human’s unit element is embraced in the usual tracking system, making it problematic for robust monitoring. The features of Haar-like and HOG methods are combined to propose a tracking system which uses the Haar characteristics for the object’s structure and the HOG features for the edge. A set of mixed features is developed with these two features. Boosting Online’s selection feature is used to select important features and update these online features to understand the optimal choice using cascading SVM classifier.

Automatic Samples Selection Using Histogram of Oriented Gradients (HOG) Feature Distance

Finding victims at a disaster site is the primary goal of Search-and-Rescue (SAR) operations. Many technologies created from research for searching disaster victims through aerial imaging. but, most of them are difficult to detect victims at tsunami disaster sites with victims and backgrounds which are look similar. This research collects post-tsunami aerial imaging data from the internet to builds dataset and model for detecting tsunami disaster victims. Datasets are built based on distance differences from features every sample using Histogram-of-Oriented-Gradient (HOG) method. We use the longest distance to collect samples from photo to generate victim and non-victim samples. We claim steps to collect samples by measuring HOG feature distance from all samples. the longest distance between samples will take as a candidate to build the dataset, then classify victim (positives) and non-victim (negatives) samples manually. The dataset of tsunami disaster victims was re-analyzed using cross-validation Leave-One-Out (LOO) with Support-Vector-Machine (SVM) method. The experimental results show the performance of two test photos with 61.70% precision, 77.60% accuracy, 74.36% recall and f-measure 67.44% to distinguish victim (positives) and non-victim (negatives).

Content-Based Image Retrieval Based on Visual Words Fusion Versus Features Fusion of Local and Global Features

Due to the advancements in digital technologies and social networking, image collections are growing exponentially. The important aim in content-based image retrieval (CBIR) is to reduce the semantic gap issue that improves the performance of image retrieval. In this paper, the objective is achieved by introducing effective visual words fusion technique based on speeded-up robust features (SURF) and histograms of oriented gradients (HOG) feature descriptors. HOG is used to extract global features, whereas SURF is used for the extraction of local features. Global features are preferred for large-scale image retrieval, whereas local features perform better on those systems that support semantic queries with close visual appearance. Moreover, SURF is scale and rotation-invariant as compared to HOG descriptor and it works better for low illumination. On the contrary, HOG performs better for scene-recognition- or activity-recognition-based applications. In the proposed technique, visual words fusion of SURF and HOG feature descriptors is carried which performed better than features fusion of SURF and HOG feature descriptors as well as state-of-the-art CBIR techniques. The proposed technique based on visual words fusion gives classification accuracy of 98.40% using support vector machine while image retrieval accuracy of 80.61%. Qualitative and quantitative analyses performed on four standard image collections namely, Corel-1000, Corel-1500, Corel-5000, and Caltech-256 demonstrate the effectiveness of the proposed technique based on visual words fusion of SURF and HOG feature descriptors.

Hand gesture recognition method based on HOG-LBP features for mobile devices

Hand Gesture recognition becomes a challenging task in computer vision field especially after the appearance of mobile devices. Although gesture recognition algorithms have been widely applied to the human-mobile interaction systems, it is difficult to meet the real time requirements and the robustness in different lightening conditions and backgrounds. This paper provides a description of a practical investigation into ways of improving hand gesture recognition algorithms for computationally limited handheld Android devices. We thoroughly researched possible approaches and find out how they affect the mobile device in terms of execution time and energy consumption. We have identified a combined HOG-LBP methodology whose performance improved upon the detection rate of other systems. In this paper, we present a static hand gesture recognition system for mobile devices by combining the histogram of oriented gradients (HOG) and local binary pattern (LBP) features, which can accurately detect hand poses. With the help of combining HOG and LBP features, we achieved a detection rate of approximately 92% on the enhanced NUS dataset I. This combination performs a better result than the results obtained when using only LBP or HOG features in term of recognition rate. However, in term of execution and detection time, LBP and HOG perform better than the combined HOG-LBP.

Effective and Fast Face Recognition System Using Complementary OC-LBP and HOG Feature Descriptors With SVM Classifier

Selection and implementation of a face descriptor that is both discriminative and computationally efficient is crucial. Local Binary Patterns (LBP) and Histogram of Oriented Gradients (HOG) have been proven effective for face recognition. LBPs are fast to compute and are easy to extract the texture features. OC-LBP descriptors have been proposed to reduce the dimensionality of LBP while increasing the discrimination power. HOG features capture the edge features that are invariant to rotation and light. Owing to the fact that both texture and edge information is important for face representation, this article proposes a framework to combine OC-LBP and HOG. First, OC-LBP and HOG features are extracted, normalized and fused together. Next, classification is achieved using a histogram-based chi-square, square-chord and extended-canberra metrics and SVM with a normalized chi-square kernel. Experiments on three benchmark databases: ORL, Yale and FERET show that the proposed method is fast to compute and outperforms other similar state-of-the-art methods.

An Accurate Multi-Biometric Personal Identification Model using Histogram of Oriented Gradients (HOG)

Biometrics is the detection and description of individuals’ physiological and behavioral ‎features. ‎Many different systems require reliable personal identification schemes to either prove or ‎find out ‎the identity of an individual demanding their services. Multi-biometrics are required inside the current ‎context of large worldwide biometric databases and to provide new developing security ‎demands. There are some distinctive and measurable features used to distinguish individuals known ‎as Biometric Identifiers. Multi-biometric systems tend to integrate multiple identifiers to increase ‎recognition accuracy. Face ‎and digital signature identifiers are still a challenge in many applications, especially in security systems. The fundamental objective of this paper is to integrate both identifiers in an accurate personal identification model. In this paper, a reliable multi-biometric model based on Histogram of Oriented Gradients (HOG) features of a face and digital signature and is able to identify individuals accurately is proposed. The methodology is to adopt many parameters such as weights of HOG features in merging process, the HOG parameters itself, and the distance method in matching process to gain higher accuracy. The proposed model achieves perfect results in personal identification using HOG features of digital signature and face together. The results show that the HOG ‎feature descriptor significantly performs target matching at an average of 100% accuracy ratio for ‎face recognition together with the digital signature. It outperforms existing feature sets with an accuracy of ‎‎84.25% for face only and 97.42% for digital signature only.‎

The Excellent Properties of a Dense Grid-Based HOG Feature on Face Recognition Compared to Gabor and LBP

To effectively represent facial features in complex environments, a face recognition method based on dense grid histograms of oriented gradients (HOG) is proposed. First, the face image is divided by numerous dense grids from which the HOG features are extracted. Then, all the grid HOG feature vectors are composed to realize the feature expression of the whole face, and the nearest neighbor classifier is used for recognition. In the FERET face database with complex changes in illumination, time, and environment, we test the gamma illumination correction, the spatial gradient direction, the size of the block, the standardization, and the face image resolution to find and analyze the optimal HOG parameters for face recognition. Finally, we compare our dense grid HOG with the two famous local facial feature extraction methods: the Gabor wavelet and the local binary pattern (LBP) on face recognition. The experimental results show that the dense grid HOG method is more suitable for the variations in time and environment. The feature extraction times of the dense grid HOG and LBP are similar. However, the dense grid HOG method uses fewer dimensions to obtain a better recognition rate than the LBP. Moreover, the dense grid HOG feature extraction time greatly outperforms the Gabor wavelet feature.

A Joint Multi-Feature and Scale-Adaptive Correlation Filter Tracker

Traditional tracking-by-detection trackers may fail to track targets owing to interferences, such as deformation, occlusion, fast or irregular motion, and background clutter. To improve the process of feature extraction, sample training, and the performance of the traditional kernelized correlation filter (KCF), this paper proposes a new joint multi-feature and scale-adaptive correlation filter tracking algorithm based on histogram of oriented gradients (HOG) and color-naming features. The new algorithm is composed of a position correlation filter tracker and a scale correlation filter tracker. It first uses a mask matrix to cut the training samples, obtains a higher proportion of real samples, and then trains the two correlation filter trackers. One is a position tracker that uses HOG and color-naming features to train two correlation filters, and then combines the results of the two correlation filters to calculate the target position. The second tracker is used to identify the scale with the maximum response size to select the target scale. The proposed algorithm achieves a competitive performance compared with algorithms, such as KCF, compressive tracking, and circulant structure kernel. The results show that this algorithm achieves a better performance when faced with occlusion, deformation, background clutter, and fast or irregular motion, and is able to track changes in target scale.

Occlusion-Aware Correlation Particle Filter Target Tracking Based on RGBD Data

In recent decades, there have been considerable improvements in target-tracking algorithms. However, aspects such as target occlusion, scale variation, and illumination changes still present significant challenges to existing algorithms. In this paper, we describe an occlusion-aware correlation particle filter target-tracking method based on RGBD data. First, we derive a target occlusion judgment mechanism based on a depth image and the histogram of oriented gradients (HOG) feature. We then formulate the tracking mechanism for the target prediction–tracking–optimization–redetection process using a correlation maximum likelihood estimation particle filter algorithm. We propose an adaptive update strategy whereby the system saves a well-tracked model when no occlusion occurs, and then uses this saved model to replace poorly tracked models in the event of occlusion. Furthermore, we consider the scale variation and adjust the target size according to the depth image, but we leave the HOG feature vector dimension of the target area unchanged. Thus, the problems such as model offset, scale variation, and loss of features are corrected over time. The experimental results demonstrate that the proposed target-tracking algorithm can detect target occlusion and track targets well, requires fewer calculations to perform target prediction–tracking–optimization–redetection, reduces the impact of illumination changes, and achieves better real-time performance and accuracy than many existing algorithms.

Real-time vehicle detection and tracking using improved histogram of gradient features and Kalman filters

Intelligent transportation systems and safety driver-assistance systems are important research topics in the field of transportation and traffic management. This study investigates the key problems in front vehicle detection and tracking based on computer vision. A video of a driven vehicle on an urban structured road is used to predict the subsequent motion of the front vehicle. This study provides the following contributions. (1) A new adaptive threshold segmentation algorithm is presented in the image preprocessing phase. This algorithm is resistant to interference from complex environments. (2) Symmetric computation based on a traditional histogram of gradient (HOG) feature vector is added in the vehicle detection phase. Symmetric HOG feature with AdaBoost classification improves the detection rate of the target vehicle. (3) A motion model based on adaptive Kalman filter is established. Experiments show that the prediction of Kalman filter model provides a reliable region for eliminating the interference of shadows and sharply decreasing the missed rate.