Structure from Motion using the Extended Kalman Filter

Civera, J. Ibáñez; Davison, Andrew J.; Montiel, José María

doi:10.1007/978-3-642-24834-4

Cited by 26 publications

(22 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The flow chart has four major realms that have both RPAS and engineering geological units: (i) preparation, (ii) field survey, (iii) data processing and calculation and (iv) interpretation. The RPAS line is described in detail in the next part of the paper, but is also linked to previous publications providing overview of image acquisition, image processing and interpretation (Civera et al, 2012;Westoby et al, 2012;Remondino et al, 2014). The engineering geological part of the flow chart is also explained below and has strong links to publications describing the application of RPAS to landslide characterization and rock slope stability assessment (Niethammer et al, 2012;Tannant, 2015).…”

Section: Methodsmentioning

confidence: 99%

Slope stability and rockfall assessment of volcanic tuffs using RPAS with 2-D FEM slope modelling

Török

Barsi

Bögöly

et al. 2018

Nat. Hazards Earth Syst. Sci.

View full text Add to dashboard Cite

Abstract. Steep, hardly accessible cliffs of rhyolite tuff in NE Hungary are prone to rockfalls, endangering visitors of a castle. Remote sensing techniques were employed to obtain data on terrain morphology and to provide slope geometry for assessing the stability of these rock walls. A RPAS (Remotely Piloted Aircraft System) was used to collect images which were processed by Pix4D mapper (structure from motion technology) to generate a point cloud and mesh. The georeferencing was made by Global Navigation Satellite System (GNSS) with the use of seven ground control points. The obtained digital surface model (DSM) was processed (vegetation removal) and the derived digital terrain model (DTM) allowed cross sections to be drawn and a joint system to be detected. Joint and discontinuity system was also verified by field measurements. On-site tests as well as laboratory tests provided additional engineering geological data for slope modelling. Stability of cliffs was assessed by 2-D FEM (finite element method). Global analyses of cross sections show that weak intercalating tuff layers may serve as potential slip surfaces. However, at present the greatest hazard is related to planar failure along ENE-WSW joints and to wedge failure. The paper demonstrates that RPAS is a rapid and useful tool for generating a reliable terrain model of hardly accessible cliff faces. It also emphasizes the efficiency of RPAS in rockfall hazard assessment in comparison with other remote sensing techniques such as terrestrial laser scanning (TLS).

show abstract

Section: Methodsmentioning

confidence: 99%

Slope stability and rockfall assessment of volcanic tuffs using RPAS with 2-D FEM slope modelling

Török

Barsi

Bögöly

et al. 2018

Nat. Hazards Earth Syst. Sci.

View full text Add to dashboard Cite

show abstract

“…Semantic information can also be added to the map by object-template matching. Civera et al [17] match the map points created using a monocular SLAM system against a known database of objects, which upon recognition using a feature based methods, can be inserted into the map. This creates more complete maps and allows for scale resolution.…”

Section: A Semantic Mappingmentioning

confidence: 99%

Meaningful maps with object-oriented semantic mapping

Sünderhauf

Pham

Latif

et al. 2017

2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

212

136

View full text Add to dashboard Cite

Fig. 1: We demonstrate object-oriented semantic mapping using RGB-D data that scales from small desktop environments (left) to offices (middle) and whole labs (right). The pictures show 3D map structures with objects colored according to their semantic class. We do not merely project semantic labels for individual 3D points, but rather maintain objects as the central entity of the map, freeing it from the requirement for a-priori 3D object models in [1]. To achieve this, our system creates and extends 3D object models while continuously mapping the environment. Object detection and classification is performed using a Convolutional Network, while an unsupervised 3D segmentation algorithm assigns a segment of 3D points to every object detection. These segmented object detections are then either fused with existing objects, or added as a new object to the map. ORB-SLAM2 provides a global SLAM solution that enables us to reconstruct a 3D model of the environment that contains both non-object structure and objects of various types.Abstract-For intelligent robots to interact in meaningful ways with their environment, they must understand both the geometric and semantic properties of the scene surrounding them. The majority of research to date has addressed these mapping challenges separately, focusing on either geometric or semantic mapping. In this paper we address the problem of building environmental maps that include both semantically meaningful, object-level entities and point-or mesh-based geometrical representations. We simultaneously build geometric point cloud models of previously unseen instances of known object classes and create a map that contains these object models as central entities. Our system leverages sparse, feature-based RGB-D SLAM, image-based deep-learning object detection and 3D unsupervised segmentation.

show abstract

“…When landmarks are static, the prediction involves estimating the camera and its covariance with cross-covariance [4]. As we transfer the camera state directly in to the reduced state x t , camera parameters can be predicted in the usual manner while keeping x y t unchanged to get the predicted state x t|(t−1) .…”

Section: Prediction In the Reduced Spacementioning

confidence: 99%

Reduced Dimensionality Extended Kalman Filter for SLAM

Gamage¹,

Drummond²

2013

Procedings of the British Machine Vision Conference 2013

View full text Add to dashboard Cite

Computational complexity of the Kalman filter grows at least quadratically with the number of dimensions in the filter. This is a particular problem for applications like monocular simultaneous localization and mapping (SLAM) where it is not possible to run a single filter on a large map with many thousands of landmarks.This paper presents a method for dramatically reducing the computational complexity of the Kalman filters by reducing the dimensionality as information is acquired. We prove the validity of our method by applying it to monocular SLAM, where there is a large number of dimensions in the filter that are not subject to process noise (the landmark locations). This has the effect of reducing the cost of running a filter or allowing a single filter to process a much larger set of landmarks.Our approach also has a role to play within modern efficient sparse matrix approaches to SLAM where local information is coalesced into keyframes using Kalman filters. It also has general applicability to filtered measurement of static quantities where there are large numbers of dimensions that are not subject to process noise.

show abstract

Structure from Motion using the Extended Kalman Filter

Abstract: The use of general descriptive names, registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use.

Cited by 26 publications

References 0 publications

Slope stability and rockfall assessment of volcanic tuffs using RPAS with 2-D FEM slope modelling

Slope stability and rockfall assessment of volcanic tuffs using RPAS with 2-D FEM slope modelling

Meaningful maps with object-oriented semantic mapping

Reduced Dimensionality Extended Kalman Filter for SLAM

Contact Info

Product

Resources

About