Search Results

Improvements in computer image processing and identification capability have led to programs that can rapidly perform calculations and model the three-dimensional spatial characteristics of objects simply from photographs or video frames. This process, known as structure-from-motion or image based scanning, is a photogrammetric technique that analyzes features of photographs or video frames from multiple angles to create dense surface models or point clouds. Concurrently, unmanned aircraft systems have gained widespread popularity due to their reliability, low-cost, and relative ease of use. These aircraft systems allow for the capture of video or still photographic footage of subjects from unique perspectives. This paper explores the efficacy of using a point cloud created from unmanned aerial vehicle video footage with traditional single-image photogrammetry methods to recreate physical evidence at a crash scene.

Assessing the ability of a driver to see objects, pedestrians, or other vehicles at night is a necessary precursor to determining if that driver could have avoided a nighttime crash. The visibility of an object at night is largely due to the luminance contrast between the object and its background. This difference depends on many factors, one of which is the amount of illumination produced by a vehicle’s headlamps. This paper focuses on a method for digitally modeling a vehicle headlamp, such that the illumination produced by the headlamps can be evaluated. The paper introduces the underlying concepts and a methodology for simulating, in a computer environment, a high-beam headlamp using a computer generated light cluster. In addition, the results of using this methodology are evaluated by comparing light values measured for a real headlamp to a simulated headlamp.

This paper examines the use of camera-matching video analysis techniques to quantify the vehicle dynamics and deformation for a dolly rollover test run in accordance with the SAE Recommended Practice J2114. The method presented enables vehicle motion data and deformation measurements to be obtained without the use of the automated target tracking employed by existing motion tracking systems. Since it does not rely on this automated target tracking, the method can be used to analyze video from rollover tests which were not setup in accordance with the requirements of these automated motion tracking systems. The method also provides a straightforward technique for relating the motion of points on the test vehicle to the motion of the vehicle's center-of-mass. This paper, first, describes the specific rollover test that was utilized. Then, the camera-matching method that was used to obtain the vehicle motion data and deformation measurements is described.

When reconstructing vehicle accidents, it is often important to record the accident scene roadway geometry. Survey techniques such as the plane surveying method1 and the coordinate and triangulation method2, have been the most widely used recording methods and are generally accepted in the industry. This paper proposes to introduce video tracking photogrammetry as a new tool in gathering accident scene and roadway data. The video tracking photogrammetry technique makes it possible to use accident scene video to record the accident scene geometry by incorporating photogrammetry principles. This paper will also report on the accuracy of the video tracking photogrammetry technique and discuss its strengths and limitations.