ORB Algorithm Research Articles

Transfer learning for image classification using VGG19: Caltech-101 image data set.

Image classification is getting more attention in the area of computer vision. During the past few years, a lot of research has been done on image classification using classical machine learning and deep learning techniques. Presently, deep learning-based techniques have given stupendous results. The performance of a classification system depends on the quality of features extracted from an image. The better is the quality of extracted features, the more the accuracy will be. Although, numerous deep learning-based methods have shown enormous performance in image classification, still due to various challenges deep learning methods are not able to extract all the important information from the image. This results in a reduction in overall classification accuracy. The goal of the present research is to improve the image classification performance by combining the deep features extracted using popular deep convolutional neural network, VGG19, and various handcrafted feature extraction methods, i.e., SIFT, SURF, ORB, and Shi-Tomasi corner detector algorithm. Further, the extracted features from these methods are classified using various machine learning classification methods, i.e., Gaussian Naïve Bayes, Decision Tree, Random Forest, and eXtreme Gradient Boosting (XGBClassifier) classifier. The experiment is carried out on a benchmark dataset Caltech-101. The experimental results indicate that Random Forest using the combined features give 93.73% accuracy and outperforms other classifiers and methods proposed by other authors. The paper concludes that a single feature extractor whether shallow or deep is not enough to achieve satisfactory results. So, a combined approach using deep learning features and traditional handcrafted features is better for image classification.

A Matching Algorithm T-AKAZE for Image Recognition of Hydroelectric Equipment Failure

Aiming at the shortcomings of the current image-based fault identification algorithm for hydropower station equipment, such as low efficiency and poor accuracy, an improved image matching algorithm based on AKAZE, T-AKAZE was proposed. In this paper, AKAZE algorithm is firstly used to obtain image feature points. Then, the improved algorithm T-AKAZE is used for coarse matching of feature points. On the basis of Tanimoto coarse matching, RANSAC algorithm is used for fine matching to improve the algorithm accuracy. In order to verify the matching effect of the improved algorithm, ORB algorithm, AKAZE algorithm and ORB+Tanimoto algorithm were compared and analyzed. The experimental results show that the proposed algorithm has good robustness for image matching of different scales and similarity, and its timeliness also meets the application requirements, which provides convenience for the staff to identify the fault of hydropower station equipment.

Improved Point-Line Feature Based Visual SLAM Method for Complex Environments.

Traditional visual simultaneous localization and mapping (SLAM) systems rely on point features to estimate camera trajectories. However, feature-based systems are usually not robust in complex environments such as weak textures or obvious brightness changes. To solve this problem, we used more environmental structure information by introducing line segments features and designed a monocular visual SLAM system. This system combines points and line segments to effectively make up for the shortcomings of traditional positioning based only on point features. First, ORB algorithm based on local adaptive threshold was proposed. Subsequently, we not only optimized the extracted line features, but also added a screening step before the traditional descriptor matching to combine the point features matching results with the line features matching. Finally, the weighting idea was introduced. When constructing the optimized cost function, we allocated weights reasonably according to the richness and dispersion of features. Our evaluation on publicly available datasets demonstrated that the improved point-line feature method is competitive with the state-of-the-art methods. In addition, the trajectory graph significantly reduced drift and loss, which proves that our system increases the robustness of SLAM.

A Vision-Inertial Odometer Design Based on ORB and Sliding Window

In order to improve the positioning accuracy and system reliability of the visual odometer in outdoor environment, this paper proposes an unmanned vehicle positioning system using monocular camera and Inertial measurement unit (IMU). The system uses improved ORB algorithm to extract image feature points and establishes a tightly coupled visual-inertial odometer system based on sliding window and nonlinear optimization. Verified by experiment, it can adapt to the environment of changing light, and can complete the positioning task of unmanned vehicles in the outdoor environment.

Improved ORB matching algorithm based on adaptive threshold

Aiming at the problem of ORB feature matching algorithm extracting background pixels as feature points and matching wrong feature points in a complex background environment, an improved ORB algorithm based on adaptive threshold is proposed, and GMS algorithm is used to screen out mismatches in the feature matching stage. First, the algorithm calculates the mean and standard deviation of the image to be matched and the reference image. Then it inputs the obtained data into the adaptive threshold calculation stage to obtain the adaptive threshold. Finally it inputs the adaptive threshold into the feature extraction and matching stage. The experimental results show that the improved ORB algorithm reduces the number of features extracted from the background of the exhibits in the complex environment of the museum, and the matching algorithm combined with the ORB algorithm and GMS increases the correct matching on the basis of slightly shorter time than the original algorithm. The algorithm has strong robustness and real-time performance.

Autonomous recognition technology of carrier robot on various terrain environment

Field carrier robot can help people carry a certain amount of necessary materials in mountaineering, exploration and other fields, and even help field soldiers carry a certain amount of armaments. However, due to the complexity of the field environment, it is difficult for a robot with a single locomotion mechanism to complete such work. A carrier robot with leg/caterpillar/wheel is proposed. The robot can recognize different terrain environment, so as to select the appropriate locomotion mechanism, and can avoid obstacles and plan the path at the same time. We optimize the traditional ORB algorithm to extract the features of the environment image. The random forest classifier is used to classify the terrain, so as to start the corresponding locomotion mechanism. In addition, the extracted ORB feature points are also used for stereo matching to achieve the depth recovery for environmental information. Based on the analysis of two path planning algorithms of DWA and artificial potential field, an artificial potential field algorithm based on DWA is proposed for robot local path planning. In order to verify the effectiveness of these methods, experiments are carried out on the switching for three locomotion mechanisms and path planning in different practical environments. The results show that when the terrain changes, the robot can effectively switch the locomotion mechanism, and can run stably in different environments.

Применение алгоритма Джарвиса для сопровождения объекта на изображениях дальнего инфракрасного диапазона

While solving the tracking of objects for infrared range images the obtaining of the necessary volume relevant basis for neuron network training is problematic enough. The main obstacles are the absence of the far infrared range images base and the complexity of calibration of cameras of various manufacturers for composing the base. Due to these causes and the complexity of marking the borders of the given range objects using the neuron networks, at the moment, is not expedient. In the paper the principles of object tracking on the infrared range images, based on obtaining the key points using the SIFT, SURF, ORB algorithms, have been considered. The specific features of the given methods have been distinguished, and also, the specifics of far infrared range images and of the objects observed on it, has been emphasized. The analysis of the approaches to preserve the features of tracking an object in transition to a new frame of a video image has been performed, their main advantages and disadvantages have been revealed. Based on the obtained results the modernization of the tracking algorithm using the Jarvis algorithm modification has been proposed. The estimation of the influence on speed of additional processing in the tracking algorithm has been experimentally performed. The proposed approach to tracking an object on the infrared image, composed on the ORB algorithm key points and the application of the modified Jervis algorithm, is invariant to turns and scaling of the sought for object and allows to watch it in conditions of high nonhomogeneity of the background and presence of similar objects. The application of the developed algorithm permits to avoid a considerable loss of speed with a significant increase of the object tracking efficiency.

A Method for Identifying Flaws on Product Surface Based on Spatial Connectivity Domain

In this paper, we exploit a method for identifying flaws on product surface based on spatial connectivity domain. A number of algorithms for detecting local features exist that were established to enhance the efficiency and accuracy of identifying interest features, such as AKAZE, BFSIFT, BRIEF, BRISK, ORB, SURF, SIFT and PCA-SIFT algorithm. But the data of flaws on product surface which is similar and consistent with the background intensity became a dilemma to detect the feature of image. In terms of identifying flaws on product surface, the above algorithms are not effective and accurate. Our aim is to enhance the accuracy of detecting the feature of flaws on product surface, so that the product with flaws could be accurately identified in industrial production. Therefore, we propose a method to identify flaws on product surface based on spatial connectivity domain. Compared with some other algorithms, the proposed method is more effective and accurate in detecting the local feature flaws on product surface of auto parts in automotive manufacturing factory.

Deployable Hook Retrieval System for UAV Rescue and Delivery

The rapid development of unmanned aerial vehicles (UAVs) has helped expand their practical use to many industrial applications. However, UAVs sometimes suffer from a flight time limitation and/or a loss in communication. Such undesired malfunctions can endanger public safety and incur economic losses. This paper presents a new class of UAV that can retrieve a disabled or malfunctioned UAV from the ground. We developed a deployable hook retrieval system (DHRS) which integrates three principal mechanisms ( i.e., deployment, slider-linkage-release, and hook release). Each mechanism plays a role in deploying and retrieving multiple hooks while using a simple control strategy. Through a Finite Element Method simulation, the hook was topologically optimized in order to achieve a high strength while reducing weight. The deployed multiple hooks allow the device to capture the target regardless of its orientation. Due to these design strategies, object recognition using a computer vision was simply demonstrated by exploiting ORB and FLANN algorithms. Through an experimental study, we discussed the target range, success rate, and the practical uses that the DHRS could achieve. The results show that the proposed designs were versatile and consistently successful in capturing the targets while addressing constraints such as power consumption, computational load, and lack of prior knowledge or information about the target.

Research on Image Feature Point Matching Based on ORB and RANSAC Algorithm

Image matching is one of the basic problems in the field of machine vision research. In order to improve the accuracy of image feature point matching and enhance the anti-interference ability of the algorithm, for the shortcomings of traditional ORB algorithm, an image feature point matching algorithm based on ORB and RANSAC algorithm is proposed. The algorithm can filter out false matches on the basis of coarse matching results, and filter out exact matching points through the homography matrix. Experiments show that this method can effectively improve the accuracy of image matching and shorten the execution time, and it is very robust to image matching with different scales, ambiguities, brightness and rotation.

A smart login system using face detection and recognition by ORB algorithm

<p>We can identify human faces using a web Camera which is known as Face Detection. This is a very effective technique in computer technology. There are used different types of attendance systems such as log in with the password, punch card, fingerprint, etc. In this research, we have introduced a facial recognition type of biometric system that can identify a specific face by analyzing and comparing patterns of a digital image. This system is the latest login system based on face detection. Primarily, the device captures the face images and stores the captured images into the specific path of the computer relating the information into a database. When any body tries to enter into any room or premises through this login system, the system captures the image of that particular person and matches the image with the stored image. If this image matches with the stored image then the system allows the person to enter the room or premises, otherwise the system denies entry. This face recognition login system is very effective, reliable and secured. This research has used the Viola and Jones algorithm for face detection and ORB for image matching in face recognition and Java, MySql, OpenCV, and iReport are used for implementation.</p>

An Improved ORB Feature Extraction Algorithm

In this paper, an improved ORB feature extraction algorithm is proposed to solve the insufficient detection effect of the traditional ORB algorithm on the image edges as well as the low actual extraction of feature points. First, the image pyramid layering was established, and then on basis of the improved oFAST12-20 corner point algorithm, the method with a setting dynamic threshold, according to the idea of adaptive threshold, was proposed to extract feature points, and the candidate set of feature points was obtained. Then, Harris algorithm was used to calculate the response value of each candidate point, and the response value was sorted from high to low. Finally, the top 80% of the feature points with high response value were extracted as the most important feature points. The experimental results show that the improved oFAST algorithm improved the extraction rate of feature points and reduces the algorithm time compared with the traditional ORB corner detection.

Oil Phase Velocity Measurement of Oil-Water Two-Phase Flow with Low Velocity and High Water Cut Using the Improved ORB and RANSAC Algorithm

Abstract Velocity is an important parameter for fluid flow characteristics in profile logging. Particle tracking velocimetry (PTV) technology is often used to study the flow characteristics of oil wells with low flow velocity and high water cut, and the key to PTV technology is particle matching. The existing particle matching algorithms of PTV technology do not meet the matching demands of oil drops in the oil phase velocity measurement of oil-water two-phase flow with low velocity and high water cut. To raise the particle matching precision, we improved the particle matching algorithm from the oriented FAST and the rotated BRIEF (ORB) feature description and the random sample consensus (RANSAC) algorithm. The simulation and experiment were carried out. Simulation results show that the improved algorithm not only increases the number of matching points but also reduces the computation. The experiment shows that the improved algorithm in this paper not only reduces the computation of the feature description process, reaching half of the computation amount of the original algorithm, but also increases the number of matching results, thus improving the measurement accuracy of oil phase velocity. Compared with the SIFT algorithm and the ORB algorithm, the improved algorithm has the largest number of matching point pairs. And the variation coefficient of this algorithm is 0.039, which indicates that the algorithm is stable. The mean error of oil phase velocity measurement of the improved algorithm is 1.20 %, and the maximum error is 6.16 %, which is much lower than the maximum error of PTV, which is 25.89 %. The improved algorithm overcomes the high computation cost of the SIFT algorithm and achieves the precision of the SIFT algorithm. Therefore, this study contributes to the improvement of the measurement accuracy of oil phase velocity and provides reliable production logging data for oilfield.

Improved ORB Algorithm Using Three-Patch Method and Local Gray Difference.

This paper presents an improved Oriented Features from Accelerated Segment Test (FAST) and Rotated BRIEF (ORB) algorithm named ORB using three-patch and local gray difference (ORB-TPLGD). ORB takes a breakthrough in real-time aspect. However, subtle changes of the image may greatly affect its final binary description. In this paper, the feature description generation is focused. On one hand, instead of pixel patch pairs comparison method used in present ORB algorithm, a three-pixel patch group comparison method is adopted to generate the binary string. In each group, the gray value of the main patch is compared with that of the other two companion patches to determine the corresponding bit of the binary description. On the other hand, the present ORB algorithm simply uses the gray size comparison between pixel patch pairs, while ignoring the information of the gray difference value. In this paper, another binary string based on the gray difference information mentioned above is generated. Finally, the feature fusion method is adopted to combine the binary strings generated in the above two steps to generate a new feature description. Experiment results indicate that our improved ORB algorithm can achieve greater performance than ORB and some other related algorithms.

Image Registration Algorithm Based on Convolutional Neural Network and Local Homography Transformation

In order to overcome the poor robustness of traditional image registration algorithms in illuminating and solving the problem of low accuracy of a learning-based image homography matrix estimation algorithm, an image registration algorithm based on convolutional neural network (CNN) and local homography transformation is proposed. Firstly, to ensure the diversity of samples, a sample and label generation method based on moving direct linear transformation (MDLT) is designed. The generated samples and labels can effectively reflect the local characteristics of images and are suitable for training the CNN model with which multiple pairs of local matching points between two images to be registered can be calculated. Then, the local homography matrices between the two images are estimated by using the MDLT and finally the image registration can be realized. The experimental results show that the proposed image registration algorithm achieves higher accuracy than other commonly used algorithms such as the SIFT, ORB, ECC, and APAP algorithms, as well as another two learning-based algorithms, and it has good robustness for different types of illumination imaging.

Research on Augmented Reality Method Based on Improved ORB Algorithm

Aiming at the problems of time-consuming three-dimensional registration algorithm in augmented reality system, low accuracy of feature point matching and poor registration effect, an improved ORB (S-ORB) algorithm is proposed and applied to the Majiayao painted pottery augmented reality system. Firstly, the scale space is established by Gaussian filtering and downsampling, and the feature points are extracted by FAST corner detection. Secondly, the descriptors of feature points are calculated by rBFEF algorithm. Finally, the feature matching set is obtained by hamming distance and the mismatching point is eliminated by RANSAC algorithm to complete the three-dimensional registration. Make the virtual information stack correctly added to the real scene. The experimental results show that the S-ORB algorithm is better than the ORB, SURF and BRISK algorithms in time and accuracy when the target position is rotated and scaled. It has better real-time and robustness in the AR system of the painted pottery.

Vein Biometric Recognition on a Smartphone

Human recognition on smartphone devices for unlocking, online payment, and bank account verification is one of the significant uses of biometrics. The exponential development and integration of this technology have been established since the introduction in 2013 of the fingerprint mounted sensor in the Apple iPhone 5s by Apple Inc.© (Motorola© Atrix was previously launched in 2011). Nowadays, in the commercial world, the main biometric variants integrated into mobile devices are fingerprint, facial, iris, and voice. In 2019, LG© Electronics announced the first mobile exhibiting vascular biometric recognition, integrated using the palm vein modality: LG© G8 ThinQ (hand ID). In this work, in an attempt to become the become the first research-embedded approach to smartphone vein identification, a novel wrist vascular biometric recognition is designed, implemented, and tested on the Xiaomi© Pocophone F1 and the Xiaomi© Mi 8 devices. The near-infrared camera mounted for facial recognition on these devices accounts for the hardware employed. Two software algorithms, TGS-CVBR® and PIS-CVBR®, are designed and applied to a database generation and the identification task, respectively. The database, named UC3M-Contactless Version 2 (UC3M-CV2), consists of 2400 contactless infrared images from both wrists of 50 different subjects (25 females and 25 males, 100 individual wrists in total), collected in two separate sessions with different environmental light environmental light conditions. The vein biometric recognition, using PIS-CVBR®, is based on the SIFT®, SURF®, and ORB algorithms. The results, discussed according to the ISO/IEC 19795-1:2019 standard, are promising and pave the way for contactless real-time-processing wrist recognition on smartphone devices.

A fast robust template matching method based on feature points

A method for template matching based on the feature matches between a target image and the template is proposed. Firstly, two sets of feature points from two images were extracted by ORB algorithm, and then the key points were matched to get a number of matching point pairs. Secondly, the wrong matches were removed to leverage feature numbers to improve quality. Finally, we explored a grid framework to locate the target object. Experiments demonstrated the great performance of our method.

Improved feature point extraction and mismatch eliminating algorithm

ABSTRACT For the existing image matching algorithms, some inherent shortcomings, such as high mismatch rate and low computational efficiency, give rise to a bad influence on the performance of Visual Simultaneous Localization and Mapping (VSLAM). In this paper, a Grid-Based Motion Statistics for Fast and Random Sample Consensus (GMS-RANSAC) method combining with Multi-Probe Location Sensitive Hash (LSH)-based Adaptive and Generic Corner Detection Based on the Accelerated Segment Test and Oriented FAST and Rotated BRIEF (AGAST-ORB) algorithm is proposed to improve the real time and accuracy of image matching. To this end, the AGAST algorithm and the multi-probe LSH algorithm are firstly integrated into the traditional ORB algorithm to obtain the initial matching set. Specifically, the image feature points are extracted by the AGAST algorithm and then the main direction of feature points is given according to the intensity centroid method to guarantee the rotary invariant of feature points. Based on the extracted feature points, the multi-probe LSH algorithm, benefiting from its high time efficiency, is used to generate the initial matching pairs. In what follows, a GMS-RANSAC algorithm, which is improved by adding a directional similarity constraint model and the traditional RANSAC algorithm, is performed to improve the accuracy of eliminating result further. Finally, the performance test is implemented via a Mikolajczyk standard data set and it is verified that the proposed algorithm has higher matching precision and matching efficiency than traditional image matching algorithms.

Vision-based vehicle detection and counting system using deep learning in highway scenes

Intelligent vehicle detection and counting are becoming increasingly important in the field of highway management. However, due to the different sizes of vehicles, their detection remains a challenge that directly affects the accuracy of vehicle counts. To address this issue, this paper proposes a vision-based vehicle detection and counting system. A new high definition highway vehicle dataset with a total of 57,290 annotated instances in 11,129 images is published in this study. Compared with the existing public datasets, the proposed dataset contains annotated tiny objects in the image, which provides the complete data foundation for vehicle detection based on deep learning. In the proposed vehicle detection and counting system, the highway road surface in the image is first extracted and divided into a remote area and a proximal area by a newly proposed segmentation method; the method is crucial for improving vehicle detection. Then, the above two areas are placed into the YOLOv3 network to detect the type and location of the vehicle. Finally, the vehicle trajectories are obtained by the ORB algorithm, which can be used to judge the driving direction of the vehicle and obtain the number of different vehicles. Several highway surveillance videos based on different scenes are used to verify the proposed methods. The experimental results verify that using the proposed segmentation method can provide higher detection accuracy, especially for the detection of small vehicle objects. Moreover, the novel strategy described in this article performs notably well in judging driving direction and counting vehicles. This paper has general practical significance for the management and control of highway scenes.