Low-level vision on Warp and the apply programming model. Technical report
In the course of implementing low-level (image to image) vision algorithms on Warp, the authors have understood the mapping of this class of algorithms well enough so that the programming of these algorithms is now a straightforward and stereotypical task. The partitioning method used is input partitioning, which provides an efficient, natural implementation of this class of algorithms. A special programming language call Apply was developed, which reduces the problem of writing the algorithm for this class of programs to the task of writing the function to be applied to a window around a single pixel. Apply provides a method for programming Warp in these applications that is easy, consistent, and efficient. Apply is application-specific, but machine independent, it is possible to implement versions of Apply that run efficiently on a wide variety of computers. Implementations are described of Apply on Warp, UNIX and the Hughes HBA, and implementation is sketched on bit-serial processor arrays and distributed-memory machines.
- Research Organization:
- Carnegie-Mellon Univ., Pittsburgh, PA (USA). Robotics Inst.
- OSTI ID:
- 5794366
- Report Number(s):
- AD-A-184330/9/XAB; CMU-RI-TR-87-17
- Country of Publication:
- United States
- Language:
- English
Similar Records
Framework for Extensible, Asynchronous Task Scheduling (FEATS) in Fortran
VLSI architecture for computer vision based on neurobiological principles of organization