Mappping of Image Processing Algorithms on an Adaptable Pipeline.

Date of Submission

December 1988

Date of Award

Winter 12-12-1989

Institute Name (Publisher)

Indian Statistical Institute

Document Type

Master's Dissertation

Degree Name

Master of Technology

Subject Name

Computer Science


Electronics and Communication Sciences Unit (ECSU-Kolkata)


Majumder, D. Dutta (ECSU-Kolkata; ISI)

Abstract (Summary of the Work)

The size of. the images dealt with in computer vision and image processing problems is usually large, the sizes varying from 128 x 128 to 1024 x 1024 pixols. Many real time applications such as target identification in missile defense systoms and robot vision require images to be processed at a very high speed. Conventional computers such as VAX-11/780 wi th a speed of 1 MFLOPS are grosly inadequate for such applications [1], nečessitating the, use of parallel processing architectures for high speed execution of these algorithms.Use of a parallel processing systems for image processing requires mapping of algorithms onto the system. Mapping is an important aspect as the speed of execution is mostly determined by the efficiency with the mapping is done.A parallel processing system which can be used in image processing is the Adaptable Pipeline whose architectural details are given in [2]. It consists of a linear array of cells all of which can be connected to a Host Computer. The cells are made up of commercially available 32-bit Transputer chips [3] which facilitate an easy and cost-effective imple- mentation of the system. In thie dissertation it is shown how image processing algorithms can be mapped onto the above syetem exploiting the parallelism in the algorithms to the maximum extent possible.The basio characteristice of image processing algorithms are studied in Chapter 2. In Chapter 3, a critical survey is made of the existing parallel processing architectures for irage processing applications. In Chapter 4, the architec- tural details of Adaptable Pipeline are discussed together with the features which make it suitable for executing image piocessing algorithms. Finally in Chapter 5 it is shown how itiage processing applications can be efficiently mapped onto such an Adaptable Pipeline.


ProQuest Collection ID:

Control Number


Creative Commons License

Creative Commons Attribution 4.0 International License
This work is licensed under a Creative Commons Attribution 4.0 International License.


This document is currently not available here.