Catherine Deegan

Assessment of stereo camera calibration techniques for a portable mobile mapping system

Mobile mapping systems that detect and geo-reference road markings almost always consist of a stereo camera system integrated with a global positioning system/inertial navigation system. The data acquired by this navigational system allows features detected in the stereo images to be assigned global co-ordinates. An essential step in this process is the calibration of the cameras, as it relates the pose of the two cameras to each other and a world co-ordinate system. In Europe, road markings must be evaluated from a 35 m range, so the cameras are required to have a wide field of view. Traditional calibration methods supposedly require a calibration object that would fill most of the calibration images. This large field of view would require a calibration object of substantial size that would be impractical for the purposes of this portable system. This study explores the theory of camera calibration and then details two camera calibration techniques (using portable 3D and 2D calibration objects). The accuracy of these methods is then evaluated using a ground-truth experiment.

Changing lanes

This paper describes an approach that allows for a steerable environment to be generated directly from a video for the purpose of integration with a video-based driving simulator. As the range of steering motion in a driving simulator is relatively limited, a pseudo-three-dimensional approach can be taken. This method requires only a single image sequence or video, acquired by any type of imaging system along a road. No three-dimensional, stereo or visual odometry data is acquired or calculated. An experiment is then presented, involving multiple lane change requests, where participants are asked to change from the left-hand lane to the right-hand lane and back again.

Automatic Generation and Population of a Graphics-Based Driving Simulator: Use of Mobile Mapping Data for Behavioral Testing of Drivers

A stereo image and high-accuracy positional data were used to generate and populate a low-fidelity graphics model. The data were acquired by a simple mobile mapping system. The use of positional data made it possible to generate automatically a sparse model consisting of a road, central road marking, a green area, and a skybox. This allowed for several applications, such as the synchronization of the model with the video and the semiautomatic population of road signs into the model data. An experiment was conducted to evaluate the model and the video as viable sources for behavioral testing of drivers. The correlations between driver speed in response to the model and the video are presented; this presentation allows for an examination of the effect of the fidelity of the driving simulators visual cue stream. The study results were used to compare driver speed in a real vehicle with driver speeds in the video and model roads, with correlations of 84.6% (between video and ground truth), 87.3% (between model and ground truth), and 92.8% (between video and model).

The application of digital accelerometers for wired and non-wired Mechanomyography

The objective of this paper is to consider the use of digital accelerometers for Mechanomyographic applications. The digital outputs of the accelerometer require the consideration of additional interfacing hardware for any commercial data acquisition systems being considered. The Arduino open-source platform is shown to meet this requirement. This platform also provides access to set the data registers on the accelerometer to output data at the resolution, speed and format required. Results show that digital accelerometers provide an accurate representation of the MMG signal. The second objective of this work was to extend this digital platform to design a wireless MMG sensor. This has been completed using open-source components and libraries. The wireless sensor can provide an inexpensive accurate representation of the MMG response for distances in excess of 30 meters.

Analysis of glottal source parameters in Parkinsonian speech

Diagnosis and monitoring of Parkinsons disease has a number of challenges as there is no definitive biomarker despite the broad range of symptoms. Research is ongoing to produce objective measures that can either diagnose Parkinsons or act as an objective decision support tool. Recent research on speech based measures have demonstrated promising results. This study aims to investigate the characteristics of the glottal source signal in Parkinsonian speech. An experiment is conducted in which a selection of glottal parameters are tested for their ability to discriminate between healthy and Parkinsonian speech. Results for each glottal parameter are presented for a database of 50 healthy speakers and a database of 16 speakers with Parkinsonian speech symptoms. Receiver operating characteristic (ROC) curves were employed to analyse the results and the area under the ROC curve (AUC) values were used to quantify the performance of each glottal parameter. The results indicate that glottal parameters can be used to discriminate between healthy and Parkinsonian speech, although results varied for each parameter tested. For the task of separating healthy and Parkinsonian speech, 2 out of the 7 glottal parameters tested produced AUC values of over 0.9.

Steering in Video-Based Driving Simulation with Stereo Depth Maps: Dynamic Perspective Corrections

This paper describes a video-based driving simulator that allows the user to steer through a video sequence acquired by a mobile mapping system. This mapping system acquired geotagged stereo video at a rate of 10 Hz. The video data were collected from a standard vehicle, which was driven along a road under normal road conditions. With the use of the stereo image data, a method that allowed a participant to steer through a video was implemented. An experiment to assess a drivers response to this steering is presented. Drivers were found to respond to the geometry of the video sequence by steering left or right when expected.

Short Stereo Baseline Retroreflector Detection Method

This paper describes a short stereo baseline vision system capable of detecting retro reflective surfaces without the requirement of image correspondence.

Comparison of Camera Calibration Techniques for a Portable Mobile Mapping System

A Mobile Mapping System (MMS) has been designed by the Intelligent Transportation Systems (ITS) research group at the Institute of Technology Blanchard town (ITB). This system detects, geo-references and evaluates road delineation data for the purposes of road marking maintenance. It consists of a monochrome stereo camera system integrated with a tightly-coupled Global Positioning System/Inertial Navigation System (GPS/INS), allowing stereo images to be acquired with navigational data, allowing for geo-referencing of detected road markings. An essential step in this process is the calibration of the cameras, and relates the pose of the two cameras to each other and a world co-ordinate system. Road markings must be evaluated from a 35 m range so the cameras are required to have a wide field of view. Traditional calibration methods supposedly require a calibration object that would fill most of the calibration images. This large field of view would require a calibration object of substantial size that would be impractical for the purposes of this portable system. This paper gives an overview of the MMS, explores the theory of camera calibration and then details two camera calibration techniques (using a three-dimensional calibration object and two-dimensional object). The accuracy of these methods is then evaluated using a ground truth experiment.