Abstract
Industrial photogrammetry provides the means to measure large-volume parts accurately and rapidly. Like all measurement systems photogrammetry has limitations. One of the limitations of photogrammetry is error that results from measuring curved surfaces. The curvature of a surface causes an applied self-adhesive photogrammetry target to distort form a circle to an ellipse. This distortion results in an error in the as-measured target data. This paper and presentation will study the effect of curvature of cylindrical parts of different radiuses on photogrammetry data. The effect of radius of a part will be quantified by comparing photogrammetry data of a part to scanned data of the same part. Cylinders of different radiuses will be measured on a surface plate that has common, stationary targets that can be measured by the photogrammetry system and a portable CMM arm and line scanner. Using the common targets as well as scanned and photogrammetry data the data sets will be compared in the same coordinate frame. The resulting data will be used to determine the variance of the photogrammetry data to the scanned data which will quantify the error caused by the photogrammetry targets being applied to a curved surface. This process will be repeated on cylinders of varying radiuses to find the maximum radius that this error appears and if there is anything the user can do to minimize or eliminate this error.
Cite this paper
Robert P. Elliott, Jr. Lockheed Martin, Sunnyvale, CA. Presented at the Coordinate Metrology Systems Conference 2014, North Charleston, South Carolina, July 21-25, 2014. Coordinate Metrology Society, reference CMS-2014-001.