The Infona portal uses cookies, i.e. strings of text saved by a browser on the user's device. The portal can access those files and use them to remember the user's data, such as their chosen settings (screen view, interface language, etc.), or their login data. By using the Infona portal the user accepts automatic saving and using this information for portal operation purposes. More information on the subject can be found in the Privacy Policy and Terms of Service. By closing this window the user confirms that they have read the information on cookie usage, and they accept the privacy policy and the way cookies are used by the portal. You can change the cookie settings in your browser.
The body action recognition is one of the key technologies of the computer vision. As the fact that the features of body action usually reside on low dimensional manifolds embedded in a high dimensional ambient space, a new method of body action recognition based on manifold learning is proposed in this paper. In the proposed method, a Linear Local Embedding of Difference (DLLE) algorithm is applied...
In this paper, a monocular 3D dense reconstruction technique based on optical flow feedback is proposed to achieve fast and accurate 3D stereoscopic modeling in the real scene. Corresponding pixel pairs are robustly matched by inter-frame optical flow fields, and the five-point algorithm is employed to determine relative pose of the moving camera, based on which sparse point cloud is generated and...
The main objective of this paper is to develop 3D point cloud alignment technique by using an improved absolute orientation algorithm based on unit quaternion. In this method, Scale Invariant Features Transform (SIFT) is used to find corresponding feature points, and Random Sample Consensus (RANSAC) is used as robust estimator to remove false matches in the point cloud group. The unit quaternion solution...
Visual object tracking is a fundamental research topic in computer vision. In this paper, we proposed a novel hybrid tracking method based on Pulse Coupled Neural Network (PCNN) and Multiple Instance Learning (MIL). Most modern trackers may be inaccurate when the training samples are imprecise which causes drift. To resolve these problems, MIL method is introduced into the tracking task, which can...
In this paper, we propose a novel view-invariant gait authentication method based on silhouette contours analysis and view estimation. The approach extracts Lucas-Kanade based gait flow image and head and shoulder mean shape (LKGFI-HSMS) of a human by using the Lucas-Kanade0s method and procrustes shape analysis (PSA). LKGFI-HSMS can preserve the dynamic and static features of a gait sequence. The...
In this paper, a vision-based human tracking approach integrating Histogram of Oriented Gradient (HOG) and Hu moment feature is presented under the Particle Filter framework. Our motivation stems from the fact that traditional Particle Filter (PF) based on single feature is not robust when the background color, illuminate and target deformation change, and it may lead to lose target or fail to detect...
This paper presents a hierarchical method of region division for mobile robot system based on spectral clustering algorithm. Firstly, to improve the exploration efficiency, topological map is constructed incrementally using a hybrid map model with grid-topplogy structure. And then, topology-based Voronoi Diagram is introduced to divide the map and express the whole environment uniquely. Meanwhile,...
In this paper, we present a Robot Technology component implementation of the simultaneous path planning and topological mapping algorithm. Firstly, Simultaneous Path Planning and Topological Mapping (SP2ATM) algorithm is applied for robot path planning and environment exploration with map construction incrementally. And at the same time, Rao-Blackwellized Particle Filter (RBPF) is utilized for robot's...
According to the two-dimensional coordinate of users body gravity center, a motion control method was proposed for the omni-directional intelligent wheelchair. Fuzzy Kohonen clustering neural network was used in the clustering process of human gravity center in order to judge the occupants driving direction intention. To reduce the effect caused by abnormal clustering center data points which were...
This paper presents a fuzzy-based intelligent control strategy allowing a mobile robot to safely follow a given person. The robot is embedded with two sensors: a RFID and a stereo camera. The RFID can locate the given person with an ID tag, and the stereo camera can be used to detect the target. Based on the two sensors, a robust control strategy is designed according to the target's speed and his...
Under the structure of robot technology middleware(RTM), this paper presents a distributed method for mobile robot simultaneous localization and mapping(SLAM) to address the problem of 3D modeling in complex indoor environment.We integrate the image feature and depth information to establish the correspondence-based iterative closest point (ICP) algorithm for localizing the robot precisely. With the...
Optical flow estimation is one of the key technologies in computer vision and image processing. However, constancy of the grey value which is used in traditional variational optical flow computation technology is sensitive to the constant changes of illumination and non-translational displacements. To solve this problem, the advanced data terms including the gradient value constancy assumptions and...
This paper presents a person detection and tracking method for a mobile robot by fusing the data from Radio Frequency Identification (RFID) and stereo camera. The RFID system detects a person wearing an ID tag and a course position estimate of the person is obtained. The stereo camera is used for person detection based on the compressive sensing theory. Less Haar-like features are extracted from compressed...
Variational methods are among the most accurate techniques for estimating the optic flow. In this paper, an estimation technology of variational optical flow based on Image and Flow-driven regularization methods are introduced to preserve the discontinuities and yield dense flow fields. Additionally, variational frame work together with the image pyramid and gaussian filter method are applied in the...
In this paper, we present a monocular vision tracking approach based on Hu moment recognition in Particle Filter framework. Our motivation stems from the fact that when the target rotates, translates and scales, target losing or mistracking always happens in image. To solve this problem, we propose an object tracking technique based on Particle Filter and Hu moment by using monocular vision method...
How to tackle the difficulty involved in optical flow calculation caused by illumination inconsistency in adjacent video frames remains an open challenge. This motivates us to investigate how to obtain a valid and accurate optical flow field which is robust to illumination change based on structure-texture decomposition. In this paper, we attempt to reveal the involved technique details in this ROF...
This paper presents an effective Simultaneous Localization and Map Building (SLAM) technique for indoor mobile robot navigation based on OpenMP. Particle Filter (PF) based SLAM provides an effective indoor mobile robot navigation framework, but real-time performance of PF needs improving due to their inherent complex and intensive computation. OpenMP is the product of the multi-core technology development...
To improve the human gait recognition rate, a method based on the modified gait flow image (GFI) considering the walking direction is used in this paper. The walking direction depends on the height and position at the walking beginning point and ending point in a gait cycle. GFI is generated to represent the motion characteristics of a gait without constructing any model. Lacus-Kanade's approach is...
In this paper, we presents a method of human detecting and tracking. To improve the robustness and real-time performance of detecting and tracking, we propose an autonomous robot detecting and tracking approach based on stereo vision and EKF. In view of diverse human motion posture, Hu moment of the head and shoulder contour is adopted to recognize human body. In addition, EKF is used to correct and...
Multi-robotic system is widely used in exploring in large-scale unknown environment and performing the complex tasks. This paper presents a method of local map merging for Multi-robotic system using RTM as communication platform. We integrate Scale-Invariant Feature Transform (SIFT) feature matching information with iterative closest point (ICP) algorithm to realize the local map merging. We use the...
Set the date range to filter the displayed results. You can set a starting date, ending date or both. You can enter the dates manually or choose them from the calendar.