Investigation of Dynamic Structured
Light (DSL) 3-D Data Artifacts
for Geological Analog Modeling, 10-R9647
Principal Investigators
Michael P. Rigney
Michael J. Magee
Ian R. Meinzen
Darrell W. Sims
Inclusive Dates: 07/18/06 – 11/20/06
Background - The Dynamic Structured Light (DSL) 3-D measurement technique is a unique, non-contact method for making 3-D measurements of a surface. The technique has many advantages, including use of inexpensive components and scalability achieved by changing the lenses and the camera or projector geometry. One promising application for DSL 3-D measurements is the physical analog modeling of geological structures currently under way at SwRI. Physical analog modeling is a well-established laboratory technique for reproducing the developmental sequence and overall geometry of geologic structures. In the models, rock strata are represented by table-top scale analog layers such as sand, claycake or silicon putty, which are deformed in ways consistent with the modeled tectonic setting. The DSL 3-D measurement technique has been used with the physical analog modeling several times, but did not provide completely satisfactory results. This project was undertaken to investigate and develop corrections for data artifacts that were noticeable in rendered 3-D surfaces of geologic analog models.
Approach - Tests were designed to investigate potential error sources and characterize the conditions that strengthen or attenuate the amplitude of the error. The DSL measurement technique incorporates several variable parameters whose effect on measurement precision was evaluated. A spreadsheet model of the signal analysis process was developed to investigate if the error artifact can be observed when processing ideal, synthetic data. Modified signal analysis procedures were developed and tested to demonstrate that some error was an artifact from the original signal analysis algorithm. Gratings used in the DSL system are characterized as "high precision" by the vendor, but flaws observed in some gratings have made them unsuitable for our use. Grating specifications were analyzed in the context of the DSL projection geometry to evaluate if grating tolerances could be an error cause. A grating from an alternate supplier was procured and tested to evaluate this potential error source.
Accomplishments - Tests conducted under this project demonstrated that various DSL configuration and parameter variables had significant effects on the rms error level of the gridline data, and that reduced rms errors could be obtained using parameters different from those normally used. In addition, modifications to DSL analysis code were developed that provided further significant reductions to gridline rms error. Using a combination of these results, rms error below 0.0005 can be achieved, which is better than a 50 percent reduction in error level relative to the previous nominal operating configuration.
