| OCR Text |
Show 25 This approach combines the geometry and value-based approaches, e.g., the volume seedlings (described in detail in Chapter 3) approach. We are extracting a particular structure based on both the data values in that structure and geometry, such as a seed location. Matte volumes can also be viewed in terms of being procedural or explicit. Explicit matte volumes are the ones in which we store a value at every point in the volume. Thus, they are an explicit array of values. Procedural matte volumes are generated by procedural definition which returns a value at every point in the data volume. The procedural definitions can depend upon the location of the data voxel, its value, and the value of its neighbors. The seed enhancement operation can be viewed as a matte volume, where the matte is defined by equation 3.7. Now the matte volume is an opacity function dependent on the seed location s and the radius r: M=a(s,r) We can take this idea further for seedling growth process. The seed location is provided by the priority function fp(t). We create a matte volume at this location. Seedling growth process is a union of series of new matte volumes defined at seed locations returned by the priority function. So the seedling growth process over time t0 to tn can be described as M=a(s,r) At each time step the priority function JP returns a new voxel. We have seen that matte volumes can be geometric as well as functions of position and data values. This provides us with a number of avenues to explore for better visualization of the scientific data, specially for datasets in which the |
