Parallel Image Processing with Image Algebra on SIMD Mesh-Connected Computers

Abstract

Publisher Summary This chapter discusses the relationship between image algebra and parallel image processing algorithms; it discusses how well single instruction multiple data (SIMD) mesh-connected computers are suited for image algebra. A group of image algebra primitives useful for parallel image processing is selected, and efficient algorithms to implement these image algebra primitives on SIMD mesh-connected computers are developed. Image algebra treats images as primary operands and addresses implicitly the parallelism in image processing. The chapter demonstrates that image algebra can serve as a good model for parallel image processing. This is accomplished by using image algebra to describe several highly parallel algorithms developed for various image processing tasks. The chapter describes an efficient algorithm for the Abingdon Cross image processing benchmark. A new binary image component shrinking algorithm is presented in the chapter. The chapter describes and analyzes several local image component labeling algorithms, one of which positively answers an open question. A special class of image-template operations that prove useful for computing properties of image components is defined and an efficient general algorithm for them is developed.