A PRACTICAL APPROACH FOR 3D BUILDING MODELING FROM UNCALIBRATED VIDEO SEQUENCES

2002 ◽  
Vol 02 (02) ◽  
pp. 287-307 ◽  
Author(s):  
YONG LIU ◽  
CHENG-KE WU ◽  
HUNG-TAT TSUI
Keyword(s):  
Structure From Motion ◽  
3D Model ◽  
Simulated Data ◽  
Image Data ◽  
Video Sequences ◽  
Real Image ◽  
Building Modeling ◽  
Critical Configurations ◽  
Motion Problem ◽  
Fundamental Matrices

This paper presents an approach for reconstructing a realistic 3D model of a building from its uncalibrated video sequences taken by a hand-held camera. The novelty of this approach lies in the integration of some prior scene knowledge in the different stages of the Structure From Motion problem (SFM). First, the coplanarity of buildings is considered in the calculation of the fundamental matrices to deal with the critical configurations. Second, the line parallelism and plane orthogonality are transformed to the constraints on the absolute quadric during camera auto-calibration. This makes some critical cases solvable and the reconstruction more Euclidean. The approach is implemented and validated using simulated data and real image data. The experimental results at the end of the paper show the effectiveness of our approach.

An attempt of SfM photogrammetry to narrow and dark underground building heritage

2020 ◽  
Author(s):  
Chiaki Oguchi ◽  
Kisara Shimizu ◽  
Yasuhiko Tamura ◽  
Yuichi Hayakawa ◽  
Takuro Ogura
Keyword(s):  
Structure From Motion ◽  
3D Model ◽  
Layer Structure ◽  
Morphological Characteristics ◽  
Image Data ◽  
3D Models ◽  
Underground Space ◽  
Central Japan ◽  
Strong Light ◽  
Underground Passage

<p>The 3D models creating by SfM (Structure-from-Motion) photogrammetry became one of the important and convenient methods for any kinds of objects on geomorphology, geoheritage, or geoarchaeological fields. These objects are landforms, monuments, buildings, relics and so on. In order to evaluate these objects, it is necessary to collect morphological characteristics, and then proceeding to decide investigating points or areas of these materials.</p><p>The progress of this methods developed significantly, however, there have been still remained difficulties depending on the objects. For example, it is difficult to create 3D models that the object is too flat, too dark, and/or any restricts of combination of target size and focusing distances. The present study attempts to these difficulties by targeting to narrow and dark underground space. The investigating object is an archaeological man-made cave, called Taya Cave, in central Japan. It was excavated in 13 century originally and used as study areas for Buddhists by making Buddhism bas-reliefs. The cave has a total length of 570 m underground passage with a three-layer structure. The cave also has several domes connected by narrow paths. The present study tried to make a 3D model of this complicated, dark and narrow cave by SfM photogrammetry. In order to concur to make 3D models for the whole area of the cave, it is useful making chunks; separating several areas of simple morphology and then compiled. When facing narrow path, it is better to take photographs not by perpendicularly but by inclinedly. Furthermore, it is important to use strong light with attach to camera. After obtained the image data of the whole cave, the accuracy of the created model was evaluated. The results were that the accuracy of horizontal distances are higher than that of vertical distances. </p>

scholarly journals The use of Convolutional Neural Networks for signal-background classification in Particle Physics experiments

2020 ◽  
Vol 245 ◽  
pp. 06003
Author(s):  
Venkitesh Ayyar ◽  
Wahid Bhimji ◽  
Lisa Gerhardt ◽  
Sally Robertson ◽  
Zahra Ronaghi
Keyword(s):  
Neural Networks ◽  
Convolutional Neural Networks ◽  
Particle Physics ◽  
Simulated Data ◽  
Image Data ◽  
Use Case ◽  
Neural Architecture ◽  
Ice Cube ◽  
2D And 3D ◽  
Physics Experiments

The success of Convolutional Neural Networks (CNNs) in image classification has prompted efforts to study their use for classifying image data obtained in Particle Physics experiments. Here, we discuss our efforts to apply CNNs to 2D and 3D image data from particle physics experiments to classify signal from background. In this work we present an extensive convolutional neural architecture search, achieving high accuracy for signal/background discrimination for a HEP classification use-case based on simulated data from the Ice Cube neutrino observatory and an ATLAS-like detector. We demonstrate among other things that we can achieve the same accuracy as complex ResNet architectures with CNNs with less parameters, and present comparisons of computational requirements, training and inference times.

scholarly journals REPRESENTATION AND DIGITALIZATION OF STONE THEATRES IN EASTERN SICILY: THE PALAZZOLO ACREIDE THEATER

2019 ◽  
Vol XLII-2/W11 ◽  
pp. 475-479
Author(s):  
G. Di Gregorio
Keyword(s):  
Structure From Motion ◽  
3D Model ◽  
Southern Italy ◽  
Three Dimensional ◽  
Cultural Value ◽  
Geometric Conditions ◽  
Dense Matching ◽  
Laser Scanners ◽  
First Results ◽  
Quality And Reliability

<p><strong>Abstract.</strong> The ancient theatres in Sicily, in southern Italy and along the countries facing the Mediterranean Sea basin, constitute a reality of incomparable cultural value. Regarding the research on the ancient theatres of eastern Sicily, few studies have been recently dealt with different methodologies. In the last years some practices have been done using 3D laser scanners for the theatres of Syracuse, Taormina and Morgantina, as well as the Syracuse amphitheatre and Taormina Odeon, just obtaining very interesting results. Lately the theatre of Palazzolo Acreide (Syracuse) has been studied, with Structure From Motion (SFM) and Dense Matching methodologies. From these experience, conclusions could be drawn on the quality and reliability of the elaborations realised with the SFM methodologies. We really know that these systems are today representing one of the fastest growing areas of examination, on which several software houses are investing. The study was chosen both for the small size of the building, and for the particular geometric conditions typical of the architecture of ancient theatres. This because their three-dimensional trend varies continually in the three variables X, Y, Z. The purpose of the work was to check whether the latest releases of these systems of survey allow today more than yesterday, a rapid digitalization and representation of the enormous archaeological cultural heritage. Various software were used, to verify the practicality and operation, the choice then fell on the Zephyr of 3DFlow, kindly available by the manufacturer, whose results were quite agreeable. The possibility offered by the program of a graphical tracing of polylines on the textured 3D model, has been a considerable advantage. Therefore the results obtained by modeling and surveying of the Palazzolo Acreide theatre have been compared, with the survey of the Syracuse, Taormina and Morgantina theatre performed using 3D laser scanners. First results of the research are matter of the following work.</p>

scholarly journals Study on Robot Teleperation System using Real-Image Data Record

2004 ◽  
Vol 2004 (0) ◽  
pp. 4
Author(s):  
N Shiroma ◽  
J Kagotani ◽  
M Sugimoto ◽  
M Inami ◽  
F Matsuno
Keyword(s):  
Image Data ◽  
Real Image ◽  
Data Record

scholarly journals Key Issues in Modeling of Complex 3D Structures from Video Sequences

2012 ◽  
Vol 2012 ◽  
pp. 1-17 ◽  
Author(s):  
Shengyong Chen ◽  
Yuehui Wang ◽  
Carlo Cattani
Keyword(s):  
Basic Principle ◽  
Video Analysis ◽  
Structure From Motion ◽  
State Of The Art ◽  
Three Dimensional ◽  
Video Sequences ◽  
3D Structures ◽  
Uncalibrated Camera ◽  
Key Issues ◽  
Scene Object

Construction of three-dimensional structures from video sequences has wide applications for intelligent video analysis. This paper summarizes the key issues of the theory and surveys the recent advances in the state of the art. Reconstruction of a scene object from video sequences often takes the basic principle of structure from motion with an uncalibrated camera. This paper lists the typical strategies and summarizes the typical solutions or algorithms for modeling of complex three-dimensional structures. Open difficult problems are also suggested for further study.

scholarly journals Spatial, Temporal, and Interchannel Image Data Fusion for Long-Distance Terrestrial Observation Systems

2008 ◽  
Vol 2008 ◽  
pp. 1-18 ◽  
Author(s):  
Barak Fishbain ◽  
Leonid P. Yaroslavsky ◽  
Ianir Ideses
Keyword(s):  
Data Fusion ◽  
Visual Analysis ◽  
Real Life ◽  
Super Resolution ◽  
Image Data ◽  
Video Sequences ◽  
Long Distance ◽  
Image Stabilization ◽  
Thermal Range ◽  
Observation Systems

This paper presents methods for intrachannel and interchannel fusion of thermal and visual sensors used in long-distance terrestrial observation systems. Intrachannel spatial and temporal fusion mechanisms used for image stabilization, super-resolution, denoising, and deblurring are supplemented by interchannel data fusion of visual- and thermal-range channels for generating fused videos intended for visual analysis by a human operator. Tests on synthetic, as well as on real-life, video sequences have confirmed the potential of the suggested methods.

A Data-Constrained 3D Model for Material Compositional Microstructures

2008 ◽  
Vol 32 ◽  
pp. 267-270 ◽  
Author(s):  
Sam Yang ◽  
Scott Furman ◽  
Andrew Tulloh
Keyword(s):  
Mathematical Model ◽  
Material Properties ◽  
Reasonable Agreement ◽  
3D Model ◽  
Compositional Data ◽  
Simulated Data ◽  
Measured Data ◽  
Porosity Distribution ◽  
Numerical Predictions ◽  
Tomography Data

A mathematical model has been developed for predicting material compositional microstructures using measured data as constraints. Examples of measured data include 3-D sets of tomography data, 2-D sets of compositional data on surfaces and sections, and material absorption and interaction properties. The model has been partially implemented as a MS-Windows application. Reasonable agreement has been obtained between the numerical predictions from the software and the simulated data. The predicted microstructures could be used to study various material properties such as porosity distribution, diffusion and corrosion.

Sign in / Sign up

Export Citation Format

Share Document