The Meaning of Accuracy

The following is intended to help MicroScan users better understand their application requirements

As a matter of fact every scanner in the world has its typical range of applications and its limitations. There is no scanner which fits every requirement one can think of. As for the MicroScan in particular the point is that it offers high (although not ultimate) accuracy at an excellent price-performance, thus making it attractive for almost every company working in 3D. Clearly, the larger the number of companies who can afford the system, the larger the number of customers and applications.
  
Many customers who would need a scanner but only have limited budgets now can think of purchasing one. This statement is valid across all market segments as listed on the backside of the MicroScan flyer.

Coming back to accuracy: When deciding for a 3D (point cloud) scanner accuracy has always been the most important issue for customers - and for many of them still remains the only one. However, looking closer at particular applications reveals that the number of tasks which really require high end precision is considerably smaller than expected.

For a particular scanner, let the resolution of the scanned point cloud be higher than the density of shape variations of the captured surface. In that case the scanned surface is ruled by the global behaviour of the point cloud, not by individual measurement points since they only have very little influence. The determining factors for the quality of the final surface are the post-processing software and the statistics of the measurement process rather than the scanner’s tolerance. In a real point cloud the distribution of points lying below the ideal surface equals the distribution of points lying above.

As a consequence, the model surface after post-processing is very close to the real shape of the scanned object. This is true for the 3D model in polygon representation (with carefully chosen smoothing parameters) and for the NURBS model as well (in case a surface reconstruction was needed, e.g. for reverse engineering purposes). Only in areas where the surface curvature variations are in the range of the point-to-point distance accuracy becomes important. With a point density of around 0.1 mm MicroScan can capture objects with a high level of detail. In that context, detail accuracy means that features which are sized down to the specified value are still represented in the measured data. MicroScan software uses sophisticated algorithms to align the scanned profiles which themselves have a higher accuracy than the underlying MicroScribe. As a result, overall system accuracy and resolution equals or even exceeds the MicroScribe’s characteristics. Apart from that MicroScan can also be used together with the new MX generation of MicroScribe CMMs which are accurate to 0.1 mm and better. Also important to know: No point cloud scanner in the world can capture sharp features like edges or corners since there is no a priori knowledge of the objects topology. If needed for use in subsequent applications like reverse engineering or rapid prototyping these features have to be remodelled in software anyway.

Tests prove what is evident from theory: In many cases the quality of surfaces created from scanned data does not differ significantly when scanners with accuracies and resolutions within the same order of magnitude or more are compared.

The answer seems clear. MicroScribe and MicroScan set new much lower boundaries for price performance in 3D scanning. You can take prototypes to product with the highest performance, lowest cost laser scanning solution available in the world today.