Nouranics

Nouranics

About this blog

Here you can find information about the things I work with, details of publications, projects, my resume, and software I have written as well slides from the subjects I used to teach.

FFT-based Terrain Segmentation for Underwater Mapping

PublicationsPosted by navid nourani Fri, July 13, 2012 05:36:21

Authors

B. Douillard, N. Nourani-Vatani, M. Johnson-Roberson, S. Williams, C. Roman, O. Pizarro, I. Vaughn, G. Inglis

Publication Date

July 2012

Conference

Robotics: Science and Systems, Sydney, Australia

Abstract

A method for segmenting three-dimensional scans of underwater unstructured terrains is presented. Individual terrain scans are represented as an elevation map and analysed using fast Fourier transform (FFT). The segmentation of the ground surface is performed in the frequency domain. The lower frequency components represent the slower varying undulations of the underlying ground whose segmentation is similar to de-noising / low pass filtering. The cut-off frequency, below which ground frequency components are selected, is automatically determined using peak detection. The user can specify a maximum admissible size of objects (relative to the extent of the scan) to drive the automatic detection of the cut-off frequency. The points above the estimated ground surface are clustered via standard proximity clustering to form object segments. The approach is evaluated using ground truth hand labelled data. It is also evaluated for registration error when the segments are fed as features to an alignment algorithm. In both sets of experiments, the approach is compared to three other segmentation techniques. The results show that the approach is applicable to a range of different terrains and is able to generate features useful for navigation.

RSS ePoster Presentation

Paper

  • Comments(0)//robotics.nourani.dk/#post15

Correlation-based Visual Odometry for Ground Vehicles

PublicationsPosted by navid nourani Fri, January 27, 2012 09:56:53

DOI: 10.1002/rob.20407

Authors

Navid Nourani-Vatani and Paulo VK Borges

Publication Date

September 2011

Journal

Journal of Field Robotics

Abstract

Reliable motion estimation is a key component for autonomous vehicles. We present a visual odometry method for ground vehicles using template matching. The method uses a downward facing camera perpendicular to the ground and estimates the motion of the vehicle by analyzing the image shift from frame to frame. Specifically, an image region (template) is selected and using correlation we find the corresponding image region in the next frame. We introduce the use of multi-template correlation matching and suggest template quality measures for estimating the suitability of a template for the purpose of correlation. Several aspects of the template choice are also presented. Through an extensive analysis we derive the expected theoretical error rate of our system and show its dependence on the template window size and image noise. We also show how a linear forward prediction filter can be used to limit the search area to significantly increase the computation performance. Using a single camera and assuming Ackerman-steering model, the method has been implemented successfully on a large industrial forklift and a 4x4 vehicle. Over 6km of field trials from our industrial test site, an off-road area and an urban environment are presented illustrating the applicability of the method as an independent sensor for large vehicle motion estimation at practical velocities.

Video

  • Comments(0)//robotics.nourani.dk/#post10

Topological Localization Using Optical Flow Descriptors

PublicationsPosted by navid nourani Fri, January 27, 2012 09:50:52

Authors

Navid Nourani-Vatani, Paulo VK Borges, Jonathan Roberts and Mandyam Srinivasan

Publication Date

November 2011

Conference Name & Place

IEEE International Conference on Computer Vision (ICCV) Workshop on Challenges and Opportunities in Robot Perception, Barcelona, Spain

Abstract

We propose a topological localization method based on optical flow information. We analyse the statistical characteristics of the optical flow signal and demonstrate that the flow vectors can be used to identify and describe key locations in the environment. The key locations (nodes) correspond to significant scene changes and depth discontinuities. Since optical flow vectors contain position, magnitude and angle information, for each node, we extract low and high order statistical moments of the vectors and use them as descriptors for that node. Once a database of nodes and their corresponding optical flow features is created, the robot can perform topological localization by using the Mahalanobis distance between the current frame and the database. This is supported by field trials, which illustrate the repeatability of the proposed method for detecting and describing key locations in indoor and outdoor environments in challenging and diverse lighting conditions.

Video




  • Comments(0)//robotics.nourani.dk/#post9

Vision-Based Detection of Unusual Patient Activity

PublicationsPosted by navid nourani Fri, January 27, 2012 09:38:03

Authors

Paulo VK Borges and Navid Nourani-Vatani

Publication Date

August 2011

Conference Name & Place

Health Informatics Conference, Brisbane, Australia

Abstract

Automated patient monitoring in hospital environments has gained increased attention in the last decade. An important problem is that of behaviour analysis of psychiatric patients, where adequate monitoring can minimise the risk of harm to hospital staff, property and to the patients themselves. For this task, we perform a preliminary investigation on visual-based patient monitoring using surveillance cameras. The proposed method uses statistics of optical flow vectors extracted from the patient movements to identify dangerous behaviour. In addition, the method also performs foreground segmentation followed by blob tracking in order to extract shape and temporal characteristics of blobs. Dangerous behaviour includes attempting to break out of safe-rooms, self-harm and fighting. The features considered include a temporal and multi-resolution analysis of blob coarseness, blob area, movement speed and position in the room. This information can also be used to normalise the other features according to estimated position of the patient in the room. In this preliminary study, experiments in a real hospital scenario illustrate the potential applicability of the method.


  • Comments(0)//robotics.nourani.dk/#post8

Scene Change Detection for Topological Mapping and Localization

PublicationsPosted by navid nourani Fri, January 27, 2012 09:33:19

Authors

Navid Nourani-Vatani and Cedric Pradalier

Publication Date

October 2010

Conference Name & Place

IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Taipei, Taiwan

Abstract

A method for detecting changes in the environment using only vision sensors is presented. We demonstrate that optical flow can be used to detect these changes at key locations in outdoor scenarios in difficult and varying lighting conditions. These key locations are used as nodes in a topological mapping and localization framework. To close the loop we employ a bag-of-words methodology. We show that bag-of-words methods can be used in real-time on a standard computer to detect loop closures in sparse topological maps. Experimental results from field trials using our quad-rotor UAV demonstrate the capability of the proposed scene change detection method.

Video


  • Comments(0)//robotics.nourani.dk/#post7

Practical Visual Odometry for Car-like Vehicles

PublicationsPosted by navid nourani Fri, January 27, 2012 09:26:30

Authors

Navid Nourani-Vatani, Jonathan Roberts and Mandyam Srinivasan

Publication Date

May 2009

Conference Name & Place

IEEE International Conference on Robotics and Automation (ICRA), Kobe, Japan

Abstract

A method for calculating visual odometry for ground vehicles with car-like kinematic motion constraints similar to Ackerman's steering model is presented. By taking advantage of this non-holonomic driving constraint we show a simple and practical solution to the odometry calculation by clever placement of a single camera. The method has been implemented successfully on a large industrial forklift and a Toyota Prado SUV. Results from our industrial test site is presented demonstrating the applicability of this method as a replacement for wheel encoder-based odometry for these vehicles.


  • Comments(0)//robotics.nourani.dk/#post6

IMU aided 3D visual odometry for car-like vehicles

PublicationsPosted by navid nourani Fri, January 27, 2012 09:23:16

Authors

Navid Nourani-Vatani, Jonathan Roberts and Mandyam Srinivasan

Publication Date

December 2008

Conference Name & Place

Australasian Conference on Robotics and Automation (ACRA), Canberra, Australia

Abstract

We present a method for calculating odometry in three-dimensions for car-like ground vehicles with an Ackerman-like steering model. In our approach we use the information from a single camera to derive the odometry in the plane and fuse it with roll and pitch information derived from an on-board IMU to extend to three-dimensions, thus providing odometric altitude as well as traditional x and y translation. We have mounted the odometry module on a standard Toyota Prado SUV and present results from a car-park environment as well as from an off-road track.

Video


  • Comments(0)//robotics.nourani.dk/#post5

Automatic Camera Exposure Control

PublicationsPosted by navid nourani Fri, January 27, 2012 09:19:16

Authors

Navid Nourani-Vatani and Jonathan Roberts

Publication Date

December 2007

Conference Name & Place

Australasian Conference on Robotics and Automation (ACRA), Brisbane, Australia

Abstract

It is commonplace to use digital video cameras in robotic applications. These cameras have built-in exposure control but they do not have any knowledge of the environment, the lens being used, the important areas of the image and do not always produce optimal image exposure. Therefore, it is desirable and often necessary to control the exposure off the camera. In this paper we present a scheme for exposure control which enables the user application to determine the area of interest. The proposed scheme introduces an intermediate transparent layer between the camera and the user application which combines the information from these for optimal exposure production. We present results from indoor and outdoor scenarios using directional and fish-eye lenses showing the performance and advantages of this framework.



  • Comments(0)//robotics.nourani.dk/#post4
« PreviousNext »