FIGURE. DETAIL PERSPECTIVE WITH ONLY TWO SURVEYS, ONE SURVEY (LEFT), BOTH SURVEYS (MIDDLE), THE OTHER SURVEY (RIGHT)
 
FIGURE. ANOTHER DETAIL PERSPECTIVE FROM THE SAME DATA SET, ENTIRE SURVEY (LEFT) AND DETAIL WITH OUTLIERS (IN RED, TO THE RIGHT)

The two Figures above  show, in addition to the surface, a distance field computed by comparing the surface to the initial data points. This functionality is included in the workflows for elevation models from point cloud data. Also this functionality scales well in the cloud.  It compares one or more data sets with a LR B-splines surface of the area covered by the data sets. In the workflows, the point data are the ones used to produce the surface representation. Alternatively, other data from the same area can be compared to the surface giving for instance an indication of whether the sea floor has change in time. 

The Figure shows a detail from a data set with 255 survey blocks, which sum up to 1.5 GB. We look at two overlapping surveys. Both surveys contain a substantial number of dense points, but the pattern differs. The points are coloured according to distances, green points lie within a specified threshold while red and blue points are outside the threshold, blue points are above and red points below the surface. Even though the second survey only covers parts of the area covered by the first one, it can be seen that the distance field patterns coincide and both surveys can be used for the final surface generation giving a good basis for the computation. The second Figure above shows another detail from the same data set. Here, the distance field provides the necessary information to identify an outlier in one of the input surveys.