84 research outputs found
Impact of Economic, Social and Meteorological Factors on Load Forecasting in Different Timeframes-A Survey
No full textIntegrating Edge-Based Stereo and Structured Light for Robust Surface Reconstruction
No full textWe present a trinocular stereo vision technique that can be used to build a fast, robust 3-D description of the environment. The proposed technique merges a range image obtained by edge-based stereo with another range image obtained using structured light. In edgebased stereo technique, we utilize the geometrical constraints of the trinocular vision system and consider the uncertainty of image measurements. The edgebased stereo builds a faithful reconstruction of the fixed edges of the surface while the structured light enables us to obtain a 3-D description for smooth surface regions. Integrating these two types of range images improves the accuracy of the reconstructed surfaces and makes it applicable to different types of surfaces. The technique has been successfully applied to several indoor scenes. The reliability and speed of the proposed system make it particularly suitable for real-time applications such as robots and automatic vehicle navigation. 1 Introduction During the l..
Surface Area Distribution Descriptor for object matching
No full textAbstractMatching 3D objects by their similarity is a fundamental problem in computer vision, computer graphics and many other fields. The main challenge in object matching is to find a suitable shape representation that can be used to accurately and quickly discriminate between similar and dissimilar shapes. In this paper we present a new volumetric descriptor to represent 3D objects. The proposed descriptor is used to match objects under rigid transformations including uniform scaling. The descriptor represents the object by dividing it into shells, acquiring the area distribution of the object through those shells. The computed areas are normalised to make the descriptor scale-invariant in addition to rotation and translation invariant. The effectiveness and stability of the proposed descriptor to noise and variant sampling density as well as the effectiveness of the similarity measures are analysed and demonstrated through experimental results
WikiAutoCat: Information Retrieval System for Automatic Categorization of Wikipedia Articles
No full text
Human action recognition using trajectory-based representation
No full textAbstractRecognizing human actions in video sequences has been a challenging problem in the last few years due to its real-world applications. A lot of action representation approaches have been proposed to improve the action recognition performance. Despite the popularity of local features-based approaches together with “Bag-of-Words” model for action representation, it fails to capture adequate spatial or temporal relationships. In an attempt to overcome this problem, a trajectory-based local representation approaches have been proposed to capture the temporal information. This paper introduces an improvement of trajectory-based human action recognition approaches to capture discriminative temporal relationships. In our approach, we extract trajectories by tracking the detected spatio-temporal interest points named “cuboid features” with matching its SIFT descriptors over the consecutive frames. We, also, propose a linking and exploring method to obtain efficient trajectories for motion representation in realistic conditions. Then the volumes around the trajectories’ points are described to represent human actions based on the Bag-of-Words (BOW) model. Finally, a support vector machine is used to classify human actions. The effectiveness of the proposed approach was evaluated on three popular datasets (KTH, Weizmann and UCF sports). Experimental results showed that the proposed approach yields considerable performance improvement over the state-of-the-art approaches
Speeding up Cloth Simulation by Linearizing the Bending Function of the Physical Mass-Spring Model
No full textA Lightweight Approach to In-Place Authoring for Mobile World Browsers
No full textAchieving the ubiquity of in-place content authoring is essential to the next generation of mobile augmented reality browsers. In-place authoring is expected to be intuitive for end users and lightweight for common smart phones. We propose a solution for in-place world labeling that makes use of the smart phones' camera parameters, its built-in sensors and a little help from the user(s). The extracted data is used to calculate the 3D location of a point of interest in the user's scene and validate its presence within the cameras viewing pyramid. The proposed approach allows for single and multiple authors for the same POI which gives much promise for supporting social and collaborative authoring. We describe multiple techniques for calculating the POI location to accommodate for the different interaction scenarios with the users
- …
