johng@OCE.ORST.EDU (John A. Gregor) (02/15/91)
We have a relatively large (several thousand) set of ocean samples in 3-D that we want to visualize. Unfortunately, all the tools we have demand that the data be represented as gridded data or as a connected mesh (either tetrahedrons xor hexahedrons). Any pointers to routines or literature to accomplish this would be greatly appreciated. Some more notes about the data: It's scattered in XY but linear in Z. Also, the number of and distance between Z samples varies among the XY locations. aTdHvAaNnKcSe -John -- John Gregor johng@oce.orst.edu Oregon State University (503) 737-3022 College of Oceanography Corvallis, OR
sdo@cbnewsl.att.com (scott.orshan) (02/16/91)
In article <1991Feb14.190639.26922@lynx.CS.ORST.EDU> johng@OCE.ORST.EDU (John A. Gregor) writes: >We have a relatively large (several thousand) set of ocean samples in >3-D that we want to visualize. Unfortunately, all the tools we have >demand that the data be represented as gridded data or as a connected >mesh (either tetrahedrons xor hexahedrons). Any pointers to routines >or literature to accomplish this would be greatly appreciated. > There's a scientific graphics package (for DOS) called Graftool. It has a built in spreadsheet, and one of its capabilities is that it will convert a list of (x,y,z) points into a grid. This is usually done as a prelude to plotting the surface. It "uses an inverse-distance weighting algorithm to generate a rectangular M by N grid of points over a specified range of the two independent variables, x and y. At each point in the grid, an x, y, and z value is computed ..." You can specify weighting, normalization, and the power used in the inverse-distance weighting to affect the smoothness of the surface generated. This program does quite a lot, and it costs $495. (They have a demo version for $15). It is a product of 3-D Visions Corp, 213-540-8818. I'm not connected with them in any way, except as a customer. Scott Orshan UNIX Systems Labs 908-522-5063 sdo@attunix.att.com
wes@arco.com (Wes Monroe) (02/20/91)
In article <1991Feb14.190639.26922@lynx.CS.ORST.EDU> johng@OCE.ORST.EDU (John A. Gregor) writes: >We have a relatively large (several thousand) set of ocean samples in >3-D that we want to visualize. Unfortunately, all the tools we have >demand that the data be represented as gridded data or as a connected >mesh (either tetrahedrons xor hexahedrons). Any pointers to routines >or literature to accomplish this would be greatly appreciated. > >Some more notes about the data: It's scattered in XY but linear in Z. >Also, the number of and distance between Z samples varies among the XY >locations. > I have similarly configured data, and I also needed to feed my graphics package geometric information as well. In my case, the Z samplings represented horizons or zones, so I just grouped data by zone, and then created triangles and rectangles by traversing across the data and finding the nearest 2 or 3 neighboring points. If what your trying to do is get a feeling for how the data is configured spatially, then I might suggest an application called MacSpin for the Mac. Its reasonably inexpensive (<200$) and will post the data for you. It will also "grid" the data with a simple technique, the nearest n neighbors, but it will not output any grid connection network like what your looking for. Its from D2 Software, Inc, P.O. Box 9546, Austin Texas 78766-9546 (512) 454-7746. Cheers, Wes
sg04@harvey.gte.com (Steven Gutfreund) (02/20/91)
> In article <1991Feb14.190639.26922@lynx.CS.ORST.EDU> johng@OCE.ORST.EDU (John > A. Gregor) writes: > >We have a relatively large (several thousand) set of ocean samples in > >3-D that we want to visualize. Unfortunately, all the tools we have > >demand that the data be represented as gridded data or as a connected > >mesh (either tetrahedrons xor hexahedrons). Any pointers to routines > >or literature to accomplish this would be greatly appreciated. > > > >Some more notes about the data: It's scattered in XY but linear in Z. > >Also, the number of and distance between Z samples varies among the XY > >locations. I use the UNIRAS package for these sorts of operations. It it is especially good for GIS applications, where one has a lot of randomly placed 3D samples and one needs a grid. It is produced by European Software Contractors A/S 376, Gladsaxevej DK-2860 Soborg Denmark Phone +45 1 67 22 88. They have a very fancy package, and routines I just don't have the time to write myself. 1. A weighted moving averages technique with smoothing, and correction based on surrounding gradients (good for large data sets) 2. A Triangularization based approach 3. A fit of 5th order spline surface patch 4. Kringing (a statistical based approach) and highly valued by GIS types. Good for maintaining small intermittent features. A pretty sophisticated algorithm (ie. don't try and implement it yourself in a day). -- -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Yechezkal Shimon Gutfreund sgutfreund@gte.com GTE Laboratories, Waltham MA harvard!bunny!sgutfreund -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
robeson@brahms.udel.edu (Scott M Robeson) (02/21/91)
In article <1991Feb14.190639.26922@lynx.CS.ORST.EDU> johng@OCE.ORST.EDU (John A. Gregor) writes: >We have a relatively large (several thousand) set of ocean samples in >3-D that we want to visualize. Unfortunately, all the tools we have >demand that the data be represented as gridded data or as a connected >mesh (either tetrahedrons xor hexahedrons). A potential problem with using a "canned" routine here is that these data are distributed on the surface of an oblate spheroid: the earth. The interpolation should really be done using spherical trigonometry to compute the distance-weighting (or angles if the interpolator requires such information). There are a couple of papers on spherical interpolation (that I know of): Renka, R. (1984) "Interpolation of data on the surface of a sphere," ACM Trans. Math. Software, 10(4), 417. Willmott, C. et al. (1985) "Small-scale climate maps: A sensitivity analysis of some common assumptions associated with grid-point interpolation and contouring," American Cartographer, 12(1), 5.