A 2D-3D Object Detection System for Updating Building Information Models with Mobile Robots

Ferguson, Max; Law, Kincho H.

doi:10.1109/wacv.2019.00149

Cited by 17 publications

(13 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…RGB-D data can either be transformed to a point cloud or kept in its natural form of four channels. Therefore, we can distinguish between an RGB-D (3D) representation (e.g., Chen et al, 2018;Tang and Lee, 2019;Ferguson and Law, 2019), which exploits the depth information in its spatial form of a point cloud, and an RGB-D (2D) representation (e.g., Chen et al, 2015;He et al, 2017;Li et al, 2019c;Rahman et al, 2019;Luo et al, 2020), which holds an additional 2D depth map in an image format.…”

Section: Rgb-d Front Viewmentioning

confidence: 99%

Survey and Systematization of 3D Object Detection Models and Methods

Moritz¹,

Friederich²,

Egger³

et al. 2022

Preprint

View full text Add to dashboard Cite

This paper offers a comprehensive survey of recent developments in 3D object detection covering the full pipeline from input data, over data representation and feature extraction to the actual detection modules. We include basic concepts, focus our survey on a broad spectrum of different approaches arising in the last ten years and propose a systematization which offers a practical framework to compare those approaches on the methods level.

show abstract

Section: Rgb-d Front Viewmentioning

confidence: 99%

Survey and Systematization of 3D Object Detection Models and Methods

Moritz¹,

Friederich²,

Egger³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Based on these advances, the use of BIM has been extended to robotic navigation. For instance, object recognition-based navigation was proposed in [20]. In that research, real-life pictures were segmented and compared with a digital 3D object repository.…”

Section: Related Workmentioning

confidence: 99%

BIM-Integrated Collaborative Robotics for Application in Building Construction and Maintenance

et al. 2020

View full text Add to dashboard Cite

The application of robotics in construction is hindered by the site environment, which is unstructured and subject to change. At the same time, however, buildings and corresponding sites can be accurately described by Building Information Modeling (BIM). Such a model contains geometric and semantic data about the construction and operation phases of the building and it is already available at the design phase. We propose a method to leverage BIM for simple yet efficient deployment of robotic systems for construction and operation of buildings. With our proposed approach, BIM is used to provide the robot with a priori geometric and semantic information on the environment and to store information on the operation progress. We present two applications that verify the effectiveness of our proposed method. This system represents a step forward towards an easier application of robots in construction.

show abstract

“…In (8), we find Z i from two observations i, j using optical expansion, i.e., the change in bounding box scale ( wi wj ) relative to z-axis camera motion (C Zj − C Zi ). To measure the change in box scale using height, we replace wi wj with hi hj .…”

Section: Depth From Motion and Detection Modelmentioning

confidence: 99%

“…With the progression of high-quality datasets and subsequent methods, our community has seen remarkable advances in image segmentation [27,54], video object segmentation [35,38], and object detection [4,16,32], sometimes with a focus on a specific application like driving [5,11,43]. However, applications like autonomous vehicles and robotics require a three-dimensional (3D) understanding of the environment, so they frequently rely on 3D sensors (e.g., LiDAR [10] or RGBD cameras [8]). Although 3D sensors are great for identifying free space and motion planning, classifying and understanding raw 3D data is a challenging and ongoing area of research [25,28,33,39].…”

Section: Introductionmentioning

confidence: 99%

Learning Object Depth from Camera Motion and Video Object Segmentation

Griffin

Corso

2020

Lecture Notes in Computer Science

View full text Add to dashboard Cite

This paper addresses the problem of learning to estimate the depth of detected objects given some measurement of camera motion (e.g., from robot kinematics or vehicle odometry). We achieve this by 1) designing a recurrent neural network (DBox) that estimates the depth of objects using a generalized representation of bounding boxes and uncalibrated camera movement and 2) introducing the Object Depth via Motion and Detection Dataset (ODMD). ODMD training data are extensible and configurable, and the ODMD benchmark includes 21,600 examples across four validation and test sets. These sets include mobile robot experiments using an end-effector camera to locate objects from the YCB dataset and examples with perturbations added to camera motion or bounding box data. In addition to the ODMD benchmark, we evaluate DBox in other monocular application domains, achieving state-of-the-art results on existing driving and robotics benchmarks and estimating the depth of objects using a camera phone.

show abstract

A 2D-3D Object Detection System for Updating Building Information Models with Mobile Robots

Cited by 17 publications

References 20 publications

Survey and Systematization of 3D Object Detection Models and Methods

Survey and Systematization of 3D Object Detection Models and Methods

BIM-Integrated Collaborative Robotics for Application in Building Construction and Maintenance

Learning Object Depth from Camera Motion and Video Object Segmentation

Contact Info

Product

Resources

About