Close-Range Photogrammetry with Laser Scan Point Clouds 2012-01-0607
Close range photogrammetry involves the determination of spatial information about objects in photographs within 300 meters of the camera's focal plane. The application of close range photogrammetry to vehicular accident reconstruction has grown exponentially in the past two decades because of its obvious advantage: it offers a cost and time effective, nonintrusive means to document the three-dimensional data pertinent to an accident. Using high resolution digital images in conjunction with control points and constraints from three-dimensional survey data has proven to be an effective way to reconstruct vehicle damage and scene evidence when the physical inspections are not viable. Advances in land surveying technology have introduced High Definition Laser Scanning (HDS) - a process that results in vast data sets of millions of points, referred to as point clouds. Individual points can be colored based on a variety of criterion, including surface reflectivity spectrum or a photo texture map that the scanner projects onto the data set. Using high resolution accident scene images with common points in the HDS laser scan provides a very rapid method for locating, measuring, and presenting physical evidence that no longer exists on scene, in a 3D photorealistic environment. The accuracy of this technique is comparable to results from widely used photogrammetry software with surveyed data from a total station. A specific hypothetical case was constructed using measurement data and quality reduced reference images to compare and contrast the HDS and total station based methods of photo rectification and evidence location.