Histogram Of Oriented Gradients Features Research Articles

A Multi-Class Objects Detection Coprocessor With Dual Feature Space and Weighted Softmax

A critical mission for mobile robot vision is to detect and classify different objects with low power consumption. In this brief, a multi-class object detection coprocessor is proposed by using the Histogram of Oriented Gradient (HOG) and Local Binary Pattern (LBP) together with a weighted Softmax classifier. Its architecture is compact and hardware-friendly suitable for energy-constrained applications. Firstly, the cell-based feature extraction unit and block-level normalization reuse the SRAMs for storing one-row cell and one-row block. Meanwhile, the working frequency of the feature extraction and block-normalization unit is synchronized to the image sensor for low dynamic power. Then, a block-level one-time sliding-detection-window (OTSDW) mechanism is developed for partial classification with scalable object size. Finally, the Softmax classifier, which is implemented by the look-up table, linear fitting methods, and the fixed-point number, is tested in the Fashion MNIST dataset to evaluate its performance in multi-class classification problems and it reached an accuracy of over 86.2% with 10,180 parameters. The experimental result shows that the hardware-resource usage of the FPGA implementation is capable of 60 fps VGA video with 80.98 mW power consumption. This method uses similar or even fewer hardware resources than that of previous work using only the HOG feature and single-class classifier.

A Robust Feature Descriptor for Biomedical Image Retrieval

Biomedical image retrieval is a crucial side of computer-aided diagnosis. It helps the radiologist and medical specialist to spot and perceive the specific disease. This paper proposed an efficient approach for retrieving similar biomedical images based on the Zernike moment features, curvelet features and histogram of oriented gradients (HOG) feature. The Zernike polynomials based moments set defines the Zernike moment which is a global descriptor and it is capable of extracting both texture and shape information with minimum redundant data. The curvelet transformation is used to compute the edge-based shape information in form of curvelet histogram for the curves with discontinuity and the HOG features calculate the happenings of gradient orientation in the local areas of an image. The experiments were conducted on four benchmark biomedical image databases: HRCT dataset, Emphysema CT database, OASIS MRI database and NEMA MRI database respectively. The performance of the proposed approach was compared with many existing methods and achieved a better retrieval rate on all the four databases.

A Smartphone-Based of Wood Identification Using Image Feature Extraction

Each wood species has their special characteristics which can be differentiated based on their anatomical structures through wood identification. One of the methods is by detecting macroscopic wood image using computer vision. This method is more rapid and accurate to identify wood species compared to the conventional method. In previous work, we have developed a computer vision technique for wood identification by combining Histogram of Oriented Gradient (HOG) and Support Vector Machine (SVM). As smartphone usage increasing worldwide, capturing wood structures using this smart device is very easy to do and can replace the use of digital microscopes. This paper propose a technique for extraction the wood species on smartphone using HOG method as well as the classification method using SVM on android smartphone. SVM was used to classify the extracted wood textures from the HOG features. In our experiments, wood images of 7 wood species were used i.e Mimusops_-elengi, Melanorrhoea wallichii, Acer niveum, Cratoxylon formosum, Agathis endertii, Dyera costulata and Knema glauca. Each species has a total of 100 training images and 100 testing images. The highest accuracy is obtained by Melanorrhoea wallichii and Agathis endertii species with 84.00% score. The Agathis endertii species has the highest sensitivity and the value reaches 86%. Moreover, the Melanorrhoea wallichii species has a highest score for specificity and precision

Hybrid adversarial-discriminative network for leukocyte classification in leukemia.

Leukemia is a lethal disease that is harmful to bone marrow and overall blood health. The classification of white blood cell images is crucial for leukemia diagnosis. The purpose of this study is to classify white blood cells by extracting discriminative information from cell segmentation and combining it with the fine-grained features. We propose a hybrid adversarial residual network with support vector machine (SVM), which utilizes the extracted features to improve the classification accuracy for human peripheral white cells. Firstly, we segment the cell and nucleus by utilizing an adversarial residual network, which contains a segmentation network and a discriminator network. To extract features that can handle the inter-class consistency problem effectively, we introduce the adversarial residual network. Then, we utilize convolutional neural network (CNN) features and histogram of oriented gradient (HOG) features, which can extract discriminative features from images of segmented cell nuclei. To utilize the representative features fully, a discriminative network is introduced to deal with neighboring information at different scales. Finally, we combine the vectors of HOG features with those of CNN features and feed them into a linear SVM to classify white blood cells into six types. We used three methods to evaluate the effect of leukocyte classification based on 5000 leukocyte images acquired from a local hospital. The first approach is to use the CNN features as the input of SVM to classify leukocytes, which achieved 94.23% specificity, 95.10% sensitivity, and 94.41% accuracy. The use of the HOG features for SVM achieved 83.50% specificity, 87.50% sensitivity, and 85.00% accuracy. The use of combined CNN and HOG features achieved 94.57% specificity, 96.11% sensitivity, and 95.93% accuracy. We propose a novel hybrid adversarial-discriminative network for the classification of microscopic leukocyte images. It improves the accuracy of cell classification, reduces the difficulty and time pressure of doctors' work, and economizes the valuable time of doctors in daily clinical diagnosis.

Gravitational search algorithm based optimized deep learning model with diverse set of features for facial expression recognition

Facial expression recognition (FER) is an essential part of effective human–computer interaction and serves as a helpful medium for children and patients who have problems with communication. However, most of the previous studies focus on building a FER model based on supervised and unsupervised approaches. This paper is focused on a semi-supervised deep belief network (DBN) approach to predict the facial expressions from the CK+, Oulu CASIA, MMI, and JAFFE datasets. To achieve accurate classification of the facial expressions, a gravitational search algorithm (GSA) is applied to optimize some parameters in the DBN network. The Histogram oriented gradients (HOG) and 2D-Discrete Wavelet Transform (2D-DWT) are used for feature extraction from the lip, cheek, brow, eye, and furrow patches. The unwanted information present in the image is eliminated using a feature selection approach. The feature extraction is done by the Kernel-principal component analysis to obtain higher-order correlations between input variables and detect non-linear components. The HOG features extracted from the lip patch provides the best performance for accurate facial expression classification. Finally, a comparative analysis to compare the proposed model with different machine learning techniques based on the evaluation criteria. The results demonstrate that DBN-GSA based classifier is more accurate than the rest of the classifiers.

Adaptive Myoelectric Pattern Recognition Based on Hybrid Spatial Features of HD-sEMG Signals

Myoelectric pattern recognition is a useful tool for identifying the user’s intended motion. However, the inherent nonstationary properties of Electromyography (EMG) signals usually limited the use of real time commercial prostheses. These variations cause the degradation of myoelectric control performance and make it unstable over time, across subjects and sessions. In this study, this challenge is overcome by combining the use of robust spatial features and the supervised adaptive learning method to improve the myoelectric performance. Three types of spatial features are proposed based on histogram oriented gradient (HOG) algorithm and intensity features namely H, HI, and AIH features. H features correspond to extracting HOG features from the HD-sEMG map. HI feature is obtained by concatenating the H features with scalar intensity feature that calculated from HD-sEMG map. Finally, the hybrid AIH features are produced by combining the H features with the intensity features matrix (AI) that obtained from the segmented maps. Three sub-databases are used for evaluation. The proposal feature sets are compared with time-domain (TD) and a combination of intensity and center of gravity features (ICG) to show the powerful of these features. The offline results report the superiority of the classifier’s performance in term of precision and sensitivity based on AIH features than other feature sets (i.e. H, HI, TD, ICG) with improvement 4.1%, 3.5%, 2.24%, 5.3% and 6%, 5%, 2.2%, 6.9% respectively. The adaptive classifier based on AIH features outperforms adaptive myoelectric control based on other feature sets and the original version. The adaptive classifier utilized testing data that update the original dataset which in turn has a significant effect on improving the myoelectric performance in the presence of the variation of EMG signal properties.

Classifying white blood cells from a peripheral blood smear image using a histogram of oriented gradient feature of nuclei shapes

Researchers developed various methods and algorithms to classify white blood cells (WBCs) from blood smear images to assist hematologists and to develop an automatic system. Furthermore, the pathological and hematological conditions of WBCs are related to diseases that can be analyzed accurately in a short time. In this work, we proposed a simple technique for WBC classification from a peripheral blood smear image based on the types of cell nuclei. The developed algorithms utilized a histogram of oriented gradient (HOG) feature typically known for application in human disease detection. The segmentation of WBC nuclei utilizes a YCbCr color space and K-means clustering techniques. The HOG feature contains information about the cell nuclei shapes, which then is classified using a support vector machine (SVM) and backpropagation artificial neural network (ANN). The results show that the proposed HOG feature is useful for WBC classification based on the shapes of nuclei. We are able to categorize the type of a WBC based on its nucleus shape with more than 95% accuracy.

Maturity status classification of papaya fruits based on machine learning and transfer learning approach

Papaya (Carica papaya) is a tropical fruit having commercial importance because of its high nutritive and medicinal value. The packaging of papaya fruit as per its maturity status is an essential task in the fruit industry. The manual grading of papaya fruit based on human visual perception is time-consuming and destructive. The objective of this paper is to suggest a novel non-destructive maturity status classification of papaya fruits. The paper suggested two approaches based on machine learning and transfer learning for classification of papaya maturity status. Also, a comparative analysis is carried out with different methods of machine learning and transfer learning. The experimentation is carried out with 300 papaya fruit sample images which includes 100 of each three maturity stages. The machine learning approach includes three sets of features and three classifiers with their different kernel functions. The features and classifiers used in machine learning approaches are local binary pattern (LBP), histogram of oriented gradients (HOG), Gray Level Co-occurrence Matrix (GLCM) and k-nearest neighbour (KNN), support vector machine (SVM), Naïve Bayes respectively. The transfer learning approach includes seven pre-trained models such as ResNet101, ResNet50, ResNet18, VGG19, VGG16, GoogleNet and AlexNet. The weighted KNN with HOG feature outperforms other machine learning-based classification model with 100% of accuracy and 0.099 5 s training time. Again, among the transfer learning approach based classification model VGG19 performs better with 100% accuracy and 1 min 52 s training time with consideration of early stop training. The proposed classification method for maturity classification of papaya fruits, i.e. VGG19 based on transfer learning approach achieved 100% accuracy which is 6% more than the existing method.

The Development of a Practical Artificial Intelligence Tool for Diagnosing and Evaluating Autism Spectrum Disorder: Multicenter Study.

BackgroundAutism spectrum disorder (ASD) is a complex neurodevelopmental disorder with an unknown etiology. Early diagnosis and intervention are key to improving outcomes for patients with ASD. Structural magnetic resonance imaging (sMRI) has been widely used in clinics to facilitate the diagnosis of brain diseases such as brain tumors. However, sMRI is less frequently used to investigate neurological and psychiatric disorders, such as ASD, owing to the subtle, if any, anatomical changes of the brain.ObjectiveThis study aimed to investigate the possibility of identifying structural patterns in the brain of patients with ASD as potential biomarkers in the diagnosis and evaluation of ASD in clinics.MethodsWe developed a novel 2-level histogram-based morphometry (HBM) classification framework in which an algorithm based on a 3D version of the histogram of oriented gradients (HOG) was used to extract features from sMRI data. We applied this framework to distinguish patients with ASD from healthy controls using 4 datasets from the second edition of the Autism Brain Imaging Data Exchange, including the ETH Zürich (ETH), NYU Langone Medical Center: Sample 1, Oregon Health and Science University, and Stanford University (SU) sites. We used a stratified 10-fold cross-validation method to evaluate the model performance, and we applied the Naive Bayes approach to identify the predictive ASD-related brain regions based on classification contributions of each HOG feature.ResultsOn the basis of the 3D HOG feature extraction method, our proposed HBM framework achieved an area under the curve (AUC) of >0.75 in each dataset, with the highest AUC of 0.849 in the ETH site. We compared the 3D HOG algorithm with the original 2D HOG algorithm, which showed an accuracy improvement of >4% in each dataset, with the highest improvement of 14% (6/42) in the SU site. A comparison of the 3D HOG algorithm with the scale-invariant feature transform algorithm showed an AUC improvement of >18% in each dataset. Furthermore, we identified ASD-related brain regions based on the sMRI images. Some of these regions (eg, frontal gyrus, temporal gyrus, cingulate gyrus, postcentral gyrus, precuneus, caudate, and hippocampus) are known to be implicated in ASD in prior neuroimaging literature. We also identified less well-known regions that may play unrecognized roles in ASD and be worth further investigation.ConclusionsOur research suggested that it is possible to identify neuroimaging biomarkers that can distinguish patients with ASD from healthy controls based on the more cost-effective sMRI images of the brain. We also demonstrated the potential of applying data-driven artificial intelligence technology in the clinical setting of neurological and psychiatric disorders, which usually harbor subtle anatomical changes in the brain that are often invisible to the human eye.

Histogram of Oriented Gradients Feature Extraction From Raw Bayer Pattern Images

This brief studies the redundancy in the image processing pipeline for histogram of oriented gradients (HOG) feature extraction. The impact of demosaicing on the extracted HOG features is analyzed and experimented. It is shown that by taking advantage of the inter-channel correlation of natural images, the HOG features can be directly extracted from the Bayer pattern images with proper gamma compression. Due to the elimination of the image processing pipeline, the power consumption and computational complexity of the detection system can be significantly reduced. Experimental results show that the Bayer pattern image-based HOG features can be used in pedestrian detection systems with little performance degradation.

Research on the Detection Method of HOG-SVM for Doping Modified Hardware Trojan

With the globalization of the IC industry, hardware security has become an important topic, especially hardware Trojans (HT) hidden in ICs. In this paper, we presented the method based on histogram of oriented gradient (HOG) and support vector machine (SVM) to detect doping modified HTs, which is generally inserted by modifying the layout of some standard cells and is very difficult to detect. In order to address this problem, this paper firstly takes the standard layout without changing the doping polarity as the Trojan free (TF) layout, and regards the modified layout with changing the doping polarity as the HT layout, and then classifies the experimental data set into HT or TF by extracting the HOG features and using SVM training classification method. Experimental results demonstrate that the experimental accuracy of this method can reach 92.8%.

A Study on Pattern Recognition with the Histograms of Oriented Gradients in Distorted and Noisy Images

Histograms of oriented gradients (HOG) are still one of the most frequently used low-level features for pattern recognition in images. Despite their great popularity and simple implementation performance of the HOG features almost always has been measured on relatively high quality data which are far from real conditions. To fill this gap we experimentally evaluate their performance in the more realistic conditions, based on images affected by different types of noise, such as Gaussian, quantization, and salt-and-pepper, as well on images distorted by occlusions. Different noise scenarios were tested such anti-distortions during training as well as application of a proper denoising method in the recognition stage. As underpinned with experimental results, the negative impact of distortions and noise on object recognition with HOG features can be significantly reduced by employment of a proper denoising strategy.

Pedestrian Detection Based on Two-Stream UDN

Pedestrian detection is the core of the driver assistance system, which collects the road conditions through the radars or cameras on the vehicle, judges whether there is a pedestrian in front of the vehicle, supports decisions such as raising the alarm, automatically slowing down, or emergency stopping to keep pedestrians safe, and improves the security when the vehicle is moving. Suffering from weather, lighting, clothing, large pose variations, and occlusion, the current pedestrian detection still has a certain distance from the practical applications. In recent years, deep networks have shown excellent performance for image detection, recognition, and classification. Some researchers employed deep network for pedestrian detection and achieve great progress, but deep networks need huge computational resources, which make it difficult to put into practical applications. In real scenarios of autonomous vehicles, the computation ability is limited. Thus, the shallow networks such as UDN (Unified Deep Networks) is a better choice, since it performs well while consuming less computation resources. Based on UDN, this paper proposes a new deep network model named two-stream UDN, which augments another branch for solving traditional UDN’s indistinction of the difference between trees/telegraph poles and pedestrians. The new branch accepts the upper third part of the pedestrian image as input, and the partial image has less deformation, stable features, and more distinguished characters from other objects. For the proposed two-stream UDN, multi-input features including the HOG (Histogram of Oriented Gradients) feature, Sobel feature, color feature, and foreground regions extracted by GrabCut segmentation algorithms are fed. Compared with the original input of UDN, the multi-input features are more conducive for pedestrian detection, since the fused HOG features and significant objects are more significant for pedestrian detection. Two-stream UDN is trained through two steps. First, the two sub-networks are trained until converge; then, we fuse results of the two subnets as the final result and feed it back to the two subnets to fine tune network parameters synchronously. To improve the performance, Swish is adopted as the activation function to obtain a faster training speed, and positive samples are mirrored and rotated with small angles to make the positive and negative samples more balanced.

Classification of facial expressions using SVM and HOG

The face is one of the human biometric which is often utilized as an important information of a person. One of the unique information of the face is facial expressions, expressions are information that is given indirectly about an expression of one's feelings. Because facial expressions have a unique pattern for each expression so that the pattern of facial expression will be tested with the computer by utilizing the Histogram of oriented gradient (HOG) descriptor as the extraction of existing features in each expression Face and information acquisition from HOG will be classified by utilizing the Support vector Mechine (SVM) method. The results of facial expression classification by utilizing the Extracaski HOG features reached 76.57% at a value of K = 500 with an average accuracy of 72.57%.

Hardware implementation of real-time pedestrian detection system

Histogram of oriented gradients (HOG) descriptor and support vector machine (SVM) classifier have been considered as a very efficient and robust tools used in the development of pedestrian detection systems. However, the enormous calculations make it difficult to meet real-time requirement specifically for advanced driver assistance system applications. The latter requires higher frame rate and resolution. In the present work, efficient pipelined architecture founded on cell-based scanning and simultaneous linear SVM classification is highlighted. The proposed real-time pedestrian detection architecture performs without external memories, and only 3 × 3 pixels are processed using two line buffers. Indeed, intermediate results will contribute directly to build the cell histograms. Consequently, it allows to reduce the required memory and to speed up the HOG feature extraction procedure. The implementation results on a heterogeneous platform show that the proposed architecture achieves an efficient pedestrian detection for full-HD video recordings (1080 × 1920 pixels) at 60 fps with reduced hardware resource consumption. Using the same hardware resources at 444 MHz, the proposed system turns out to handle a HD video at 180 fps.

A Runway Detection Method Based on Classification Using Optimized Polarimetric Features and HOG Features for PolSAR Images

A novel runway detection algorithm for PolSAR (Polarimetric Synthetic Aperture Radar) images based on optimized polarimetric features and local spatial information is proposed. Existing methods for runway detection for PolSAR images always utilize the parallel line as the primary feature. However, many other ground objects such as rivers and roads also have parallel structures thus affect the performance of these detection methods. The proposed method is based on two stages of classification with polarimetric features and the HOG (Histogram of Oriented Gradient) feature, while avoiding the interference due to the similar morphological features among different ground objects. An FCBF (Fast Correlation Based Filter) is firstly used for optimizing and selecting of the ground objects' polarimetric features of ground targets. Then RF (Random Forest) classifier is employed for extracting ROIs (Region of Interest) which may contain runways. Then HOG features are extracted from these ROIs for further classification with SVM (Support Vector Machines) to detect the runway area. Experimental results with the measured PolSAR data provided by NASA UAVSAR project show that the proposed method can detect runway regions effectively without using the parallel line. Comparative analysis is also conducted on parallel line pattern based algorithms. And the results suggest the effectiveness and performance enhancement of this method.

Palmprint Recognition Using Fusion of Local Binary Pattern and Histogram of Oriented Gradients

In this paper, unique features of the segmented image samples are extracted by using two major feature extraction techniques: Local Binary Pattern (LBP) and Histogram of Oriented Gradients (HOG). After this, these features are fused to get more precise and productive outcomes. The average accuracy of the three distinct datasets that were generated using the LBP and HOG features are determined. To calculate the accuracy of the three distinct models, classification techniques like KNN and SVM, are adopted.

Low-level features based 2D face recognition using machine learning

In modern times, when deep learning-based face recognition is highly in demand, this paper presents machine learning techniques using the low-level feature extraction. Deep learning has a drawback of getting things done in a black-box, however extraction of low-level features viz. histogram of oriented gradients (HOG), speeded up robust features (SURF), and local binary patterns (LBPs) with a machine learning-based classification model presents higher simplicity. This paper presents the experimental demonstrations using 22 variations of machine learning models. Two face datasets, namely, Bosphorus and UMBDB are used for evaluating different classification models. Four experimentations are shown in the implementation section to demonstrate the effect of feature extraction, discretisation, feature variation, and noise in the image under probe. The subspace discriminant ensemble model yields the highest efficiency in classifying faces using HOG features. The effect of various noise attacks on the probe image is shown in the last experimentation.

A Novel Approach for Pose Invariant Face Recognition in Surveillance Videos

The face detection and face recognition methods are introduced to confirm the abnormal human activity in the video surveillance system. Face detection is carried out by Viola-Jones face detector. It is composed of three concepts namely integral image, AdaBoost and the cascade structure. After face detection, Histogram of Oriented Gradient(HOG) and Weighted Local Binary Patterns(WLBP) features are extracted and those are used in Orthogonal Locality Preserving Projection(OLPP) for face recognition. The detected faces may contain pose variation which dramatically degrades the OLPP based face recognition. So, pose-invariant OLLP-based face recognition is proposed where Histogram of Face Orientation(HFO) and Histogram of Face Direction(HFD), HOG and WLBP features are used in OLPP for efficient face recognition.

Real time fruits quality detection with the help of artificial intelligence

One of the major quality of grading fruits is its appearance. Appearance is effects the marketing and choice of consumer. Colour, texture, size, shape are used to find quality of fruit. But the sellers controlling the external quality of fruits to get high profit. In earlier observations, implemented products, computer vision systems for external controlling quality so grading and classification of fruits is based on observations. The proposed system depend on image processing to classify and grade quality of fruits by using mean of image, colour and HOG (Histogram of gradient) feature extractions are used to classify the fruit quality. All machine learning algorithms are used to find the better accuracy of data how it is predicting. In proposed method first data set is collected, then pre-processing is applied for better results. Machine learning algorithms (K-nearest neighbour (KNN), Support Vector Machine (SVM), and PCA is used for dimension reduction and to get good accuracy to implement the system. For big data pre-processing and to get better results Deep learning (CNN) is used to test the fruit in real time world with result and audio sounds. Audio is used to detect object by hearing also.