http://www.ri.cmu.edu/pub_files/2011/6/2011-antonio-isarc-3d-modeling-v6.pdf

Automatic Creation of Semantically Rich 3D Building Models from Laser Scanner Data

Laser scanners are often used to create 3D models of buildings for civil engineering applications. The current manual process is time-consuming and error-prone. This paper presents a method for using laser scanner data to model predominantly planar surfaces, such as walls, floors, and ceilings, despite the presence of significant amounts of clutter and occlusion, which occur frequently in natural indoor environments. Our goal is to recover the surface shape, detect and model any openings, and fill in the occluded regions. Our method identifies candidate surfaces for modeling, labels occluded surface regions, detects openings in each surface using supervised learning, and reconstructs the surface in the occluded regions. We evaluate the method on a large, highly cluttered data set of a building consisting of forty separate rooms.

/pdf/3d-reconstruction-of-interior-wall-surfaces-under-occlusion-uj5yapvydp.pdf

3D Reconstruction of Interior Wall Surfaces under Occlusion and Clutter

Laser scanners are increasingly used to create semantically rich 3D models of buildings for civil engineering applications such as planning renovations, space usage planning, and building maintenance. Currently these models are created manually – a time-consuming and error-prone process. This paper presents a method to automatically convert the raw 3D point data from a laser scanner positioned at multiple locations throughout a building into a compact, semantically rich model. Our algorithm is capable of identifying and modeling the main structural components of an indoor environment (walls, floors, ceilings, windows, and doorways) despite the presence of significant clutter and occlusion, which occur frequently in natural indoor environments. Our method begins by extracting planar patches from a voxelized version of the input point cloud. We use a conditional random field model to learn contextual relationships between patches and use this knowledge to automatically label patches as walls, ceilings, or floors. Then, we perform a detailed analysis of the recognized surfaces to locate windows and doorways. This process uses visibility reasoning to fuse measurements from different scan locations and to identify occluded regions and holes in the surface. Next, we use a learning algorithm to intelligently estimate the shape of window and doorway openings even when partially occluded. Finally, occluded regions on the surfaces are filled in using a 3D inpainting algorithm. We evaluated the method on a large, highly cluttered data set of a building with forty separate rooms yielding promising results.

Radio frequency identification (RFID) technology has been widely used in the field of construction during the last two decades. Basically, RFID facilitates the control on a wide variety of processes in different stages of the lifecycle of a building, from its conception to its inhabitance. The main objective of this paper is to present a review of RFID applications in the construction industry, pointing out the existing developments, limitations and gaps. The paper presents the establishment of the RFID technology in four main stages of the lifecycle of a facility: planning and design, construction and commission and operation and maintenance. Concerning this last stage, an RFID application aiming to facilitate the identification of pieces of furniture in scanned inhabited environments is presented. Conclusions and future advances are presented at the end of the paper.

/pdf/evolution-of-rfid-applications-in-construction-a-literature-443zs2v8dr.pdf

Evolution of RFID Applications in Construction: A Literature Review.

We introduce a technique to fill large holes in LiDAR data sets. We combine concepts from patch-based image inpainting and gradient-domain image editing to simultaneously fill both texture and structure in a LiDAR scan. We discuss the problems with directly inpainting a depth image, and present a solution to this problem based on inpainting the depth gradients. Once the inpainted depth gradients are obtained, we use an image reconstruction technique to obtain the final 3D scene structure. We present several real-world examples of this technique with excellent results.

/pdf/filling-large-holes-in-lidar-data-by-inpainting-depth-2yjm08t641.pdf

Filling large holes in LiDAR data by inpainting depth gradients

This paper presents a review of the most relevant current techniques that deal with hole-filling in 3D models. Contrary to earlier reports, which approach mesh repairing in a sparse and global manner, the objective of this review is twofold. First, a specific and comprehensive review of hole-filling techniques as a relevant part in the field of mesh repairing is carried out. We present a brief summary of each technique with attention paid to its algorithmic essence, main contributions and limitations. Second, a solid comparison between 34 methods is established. To do this, we define 19 possible meaningful features and properties that can be found in a generic hole-filling process. Then, we use these features to assess the virtues and deficiencies of the method and to build comparative tables. The purpose of this review is to make a comparative hole-filling state-of-the-art available to researchers, showing pros and cons in a common framework.

/pdf/a-comparison-of-hole-filling-methods-in-3d-3qo2fk73xu.pdf

A comparison of hole-filling methods in 3D

This work analyzes a new method for object recognition in complex scenes combining vision-based techniques applied to the 3-D data obtained using range sensors and object identification coming from radio frequency tags (radio frequency identification (RFID) technology). Three-dimensional vision-based algorithms for object recognition have many restrictions in practical applications, i.e., uncertainty, incapability for real-time tasks, etc., but they work well for pose determination once the object is recognized. On the other hand, RFID technology allows us to detect the presence of specific objects in a scene, but it cannot provide their localization, at least not with the accuracy required in applications such as ours. In this paper, we present a new and powerful recognition method obtained by fusing both techniques. The phases of the method are described, and abundant experimentation results are included. An in-depth performance analysis has been carried out to demonstrate the recognition improvements achieved by the algorithm when RFID assistance is considered. It helps to confirm the robustness of this fusion approach and prove its effectiveness. A final discussion is included, concerning what should be the most adequate size of the object database for optimal algorithm exploitation.

Improved Method for Object Recognition in Complex Scenes by Fusioning 3-D Information and RFID Technology

This work describes a method for filling holes in a 3D mesh based on 2D image restoration algorithms. Since these algorithms need an image as input, the first stage of the method concerns a 3D to 2D transformation. By projecting the 3D surface onto a squared plane, a 2D image is generated in such a way that the depth information is stored in each grid. The image restoration algorithms are applied to it. Once the image has been repaired, the inverse transformation 2D to 3D is performed and the repaired 3D surface recovered. To test the method, artificial holes have been generated on a set of 3D surfaces. The goodnessof the results has been measured from the comparison between the 3D original surfaces and the 3D repaired ones. An evaluation with commercial software has been carried out to show thevalidityof themethod. Theimage restorationalgorithms have been applied to 3D cultural heritage modeling with good results. Specifically, sculptures of the collection from the National Museum of Roman Art in Spain have been reconstructed.

/pdf/filling-holes-in-3d-meshes-using-image-restoration-42o5jvh14a.pdf

Santiago Salamanca

Papers

A comparison of hole-filling methods in 3D

Improved Method for Object Recognition in Complex Scenes by Fusioning 3-D Information and RFID Technology

Filling Holes in 3D Meshes using Image Restoration Algorithms

3D scene retrieval and recognition with Depth Gradient Images

Cultural Heritage: A hybrid human-computer approach for recovering incomplete cultural heritage pieces