Filters








1,211 Hits in 12.5 sec

Registration of Partial 3D Point Clouds Acquired from a Multi-view Camera for Indoor Scene Reconstruction

S. KIM
2006 IEICE transactions on information and systems  
Secondly, 3D point clouds acquired from two views are projected onto the same image plane, and two-step integer mapping is applied to search for correspondences through the modified KLT.  ...  In this paper, a novel projection-based method is presented to register partial 3D point clouds, acquired from a multi-view camera, for 3D reconstruction of an indoor scene.  ...  Acknowledgments This research was supported by the MIC, Korea, under the ITRC support program supervised by the IITA, and in part by CTRC at GIST.  ... 
doi:10.1093/ietisy/e89-d.1.62 fatcat:ztdtk7cfhbgqpeuoedkuk7q26i

Real-time Monocular Dense Mapping for Augmented Reality

Tangli Xue, Hongcheng Luo, Danpeng Cheng, Zikang Yuan, Xin Yang
2017 Proceedings of the 2017 ACM on Multimedia Conference - MM '17  
In this paper, we present a novel method for real-time monocular dense mapping based on the piecewise planarity assumption for poorly textured regions.  ...  Monocular simultaneous localization and mapping (SLAM) is a key enabling technique for many augmented reality (AR) applications.  ...  The coordinate of a 3D point in the world coordinates is denoted as P w = (x, y, z) T ∈ R 3 and its projection to a 2D frame i is 2D coordinates in image coordinates denoted as u = (u, v) T ∈ Ω which can  ... 
doi:10.1145/3123266.3123348 dblp:conf/mm/XueLCYY17 fatcat:oez32nsqwzcnpibxwwfxz2tiyy

A SLAM System Based on RGBD Image and Point-Line Feature

Dan Li, Shuang Liu, Weilai Xiang, Qiwei Tan, Kaicheng Yuan, Zhen Zhang, Yingsong Hu
2021 IEEE Access  
In the local mapping and loop detection module, the Plucker coordinates are applied to express the spatial line and define the re-projection error of the straight line, so that the back-end optimization  ...  Ulteriorly, the depth information of RGBD image is used to restore the 3D information of point and line features, improving the accuracy of camera track positioning.  ...  ∼ indicates that the last dimension of the homogeneous coordinates is removed.  ... 
doi:10.1109/access.2021.3049467 fatcat:lycck7i4kvdblab3b4vzmb3l5e

HIERARCHICAL STRUCTURE FROM MOTION COMBINING GLOBAL IMAGE ORIENTATION AND STRUCTURELESS BUNDLE ADJUSTMENT

A. Cefalu, N. Haala, D. Fritsch
2017 The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences  
Global image orientation techniques aim at estimating camera rotations and positions for a whole set of images simultaneously.  ...  Furthermore we substitute classical bundle adjustment by a structureless approach based on epipolar geometry and augmented with a scale consistency constraint.  ...  As in section 3.2, let be the projection of a point onto an image plane in homogenous coordinates.  ... 
doi:10.5194/isprs-archives-xlii-1-w1-535-2017 fatcat:hlhowoy3pngp3bh7go7r6cg4za

A Survey of Structure from Motion [article]

Onur Ozyesil, Vladislav Voroninski, Ronen Basri, Amit Singer
2017 arXiv   pre-print
More specifically, after touching upon the early factorization-based techniques for motion and structure estimation, we provide a detailed account of some of the recent camera location estimation methods  ...  The structure from motion (SfM) problem in computer vision is the problem of recovering the three-dimensional (3D) structure of a stationary scene from a set of projective measurements, represented as  ...  coordinates by P j ∈ R 4 and C i denoting the i'th camera matrix, p ij ∈ R 3 is the projection of the j'th point onto the i'th camera plane (in the homogeneous coordinates) and λ ij are the undetermined  ... 
arXiv:1701.08493v2 fatcat:44hrgmeiujaqtggdztx57ew44i

Bundle Adjustment — A Modern Synthesis [chapter]

Bill Triggs, Philip F. McLauchlan, Richard I. Hartley, Andrew W. Fitzgibbon
2000 Lecture Notes in Computer Science  
The theory is developed for general robust cost functions rather than restricting attention to traditional nonlinear least squares.  ...  This paper is a survey of the theory and methods of photogrammetric bundle adjustment, aimed at potential implementors in the computer vision community.  ...  E.g., suppose that image points (u/w, v/w) are generated by a homogeneous projection equation (u, v, w) = P (X, Y, Z, 1) , where P is the 3 × 4 homogeneous image projection matrix.  ... 
doi:10.1007/3-540-44480-7_21 fatcat:jm2ewshh5bejhnv5dixghvp6y4

Super-resolution of a 3-dimensional scene from novel viewpoints

Kyle Nelson, Asim Bhatti, Saeid Nahavandi
2012 2012 12th International Conference on Control Automation Robotics & Vision (ICARCV)  
The optimal view encapsulating the most scene information is located via pattern search optimisation. ii This thesis analyses two separate techniques for transferring input image information to the optimal  ...  Super-resolution is an image enhancement method that increases the spatial resolution of video or images.  ...  A projection matrix may be formulated as = [ | ], where is a 3×3 rotation matrix describing the orientation of the camera coordinate frame relative to world coordinate frame, is a 3×1 translation vector  ... 
doi:10.1109/icarcv.2012.6485347 dblp:conf/icarcv/NelsonBN12 fatcat:wfakbldhzfhl3jlkcwsuu7v34m

Building Change Detection Using High Resolution Remotely Sensed Data and GIS

Natalia Sofina, Manfred Ehlers
2016 IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing  
as the planning and coordination of recovery activities.  ...  This study presents a GIS-based approach for the detection of destroyed buildings.  ...  coordinates on the paper-based maps.  ... 
doi:10.1109/jstars.2016.2542074 fatcat:i6bgd7zgqzgslgjjyuibla5riu

Individual Tree Detection in Urban ALS Point Clouds with 3D Convolutional Networks

Stefan Schmohl, Alejandra Narváez Vallejo, Uwe Soergel
2022 Remote Sensing  
Some existing works have employed 2D object detection networks for this purpose. However, these have used three-dimensional information only in the form of projected feature maps.  ...  Although deep-learning-based object detection networks are the state-of-the-art in computer vision, their adaptation to individual tree detection in urban areas has scarcely been studied.  ...  AIRBUS Defence and Space Pléiades-1 satellite images were accessed as an ESA Third Party Mission within the ESA TPM project research id 49634.  ... 
doi:10.3390/rs14061317 fatcat:26nlpv6q5zhfffr5wvrw4uaseu

Association of Moving Objects Across Visual Sensor Networks

Muhammad J. Mirza, Nadeem Anjum
2012 Journal of Multimedia  
projection of T i n (x, y, t) and H n,G is the homography matrix.  ...  It detects the duplicated region directly, for SIFT feature is local feature, when searching all of matched feature points, the duplicated regions can be decided.  ...  Abstract-This paper proposes an OWL-S framework for distributed CAD system based on the combination of semantic web service and CAD technology.  ... 
doi:10.4304/jmm.7.1.2-8 fatcat:4cc7sala7ffixohi3ccvothajm

Automated Edge-Aware Refinement of Anatomical Atlases in 3D and Its Application to the Developing Mouse Brain [article]

David M Young, Siavash Fazel Darbandi, Grace Schwartz, Zachary Bonzell, Deniz Yuruk, Mai Nojima, Laurent Gole, John Rubenstein, Weimiao Yu, Stephan J Sanders
2020 biorxiv/medrxiv   pre-print
To test the refined 3D atlas, we generated whole-brain nuclei-resolution images from 15 mice at postnatal day 0, identifying 35 million nuclei per brain.  ...  AbstractAnatomical atlases are valuable resources for investigating organogenesis, anatomy, pathology, and therapeutic targeting.  ...  We are deeply indebted to Hwee Kuan Lee from the Agency for Science, Technology and Research (A*STAR) for invaluable imaging informatics advice.  ... 
doi:10.1101/2020.04.01.017665 fatcat:o66s5hocezcmfdw2debywyqthi

Mobile Visual Location Recognition

Georg Schroth, Robert Huitl, David Chen, Mohammad Abu-Alqumsan, Anas Al-Nuaimi, Eckehard Steinbach
2011 IEEE Signal Processing Magazine  
[WKIS09] bundle SIFT features within MSER regions [MCUP04] for partial-duplicate web image search.  ...  The camera pose of the reference image is denoted by a homogeneous 4 × 4 matrix T ref , the pose of the virtual image is denoted by T virt .  ... 
doi:10.1109/msp.2011.940882 fatcat:7vzjij4qvjb5hcc2znqp7ginuy

TOOLKIT FOR IMMERSIVE VIRTUAL SIMULATORS WITH BIOFEEDBACK

V. V. Gribova, V. O. Strekalev
2020 Vestnik komp iuternykh i informatsionnykh tekhnologii  
based on the state of a person and the state of objects are determined within its framework.  ...  The architecture of the instrumental complex for the creation and reproduction of immersive virtual simulators using biological feedback is proposed.  ...  The matrix T eye held the camera pose, whereas T vp and T can contained the perspective projection and transformed the projected point into image coordinates.  ... 
doi:10.14489/vkit.2020.08.pp.019-028 fatcat:nyet4p3qkvecliahuktcjbkwbu

De novo assembly, annotation, and comparative analysis of 26 diverse maize genomes [article]

Matthew B Hufford, Arun S Seetharam, Margaret R Woodhouse, Kapeel M. Chougule, Shujun Ou, JIANING LIU, William A Ricci, Tingting Guo, Andrew Olson, Yinjie Qiu, Rafael Della Coletta, Silas Tittes (+34 others)
2021 bioRxiv   pre-print
We report de novo genome assemblies, transcriptomes, annotations, and methylomes for the 26 inbreds that serve as the founders for the maize nested association mapping population.  ...  Finally, we document variation at the level of DNA methylation, and demonstrate that unmethylated regions are enriched for cis-regulatory elements that overlap QTL and contribute to changes in gene expression  ...  Duplicate reads were removed with Sambamba (version 0.7.1) (120) , with the parameters markdup --remove-duplicates and reads were filtered for MAPQ scores of 30 or higher with sort -F "mapping_quality  ... 
doi:10.1101/2021.01.14.426684 fatcat:veliuslqerayxm43hycqbsoqcu

3D Point Cloud Compression

Chao Cao, Marius Preda, Titus Zaharia
2019 The 24th International Conference on 3D Web Technology on - Web3D '19  
Any combination of such transformations can be described by a single four-by-four affine transformation matrix, under homogeneous coordinates.  ...  Moreover, the proposed segmentation also introduced more duplicated points when they are projected onto the 2D images.  ...  Project the error vector 𝐸(𝑖, 𝑗) along the normal direction 𝑁 𝑗 and get a new error vector 𝐸 ̂(𝑖, 𝑗).  ... 
doi:10.1145/3329714.3338130 dblp:conf/vrml/CaoPZ19 fatcat:ypzqrhcifza2rlqmrn5l3bs6du
« Previous Showing results 1 — 15 out of 1,211 results