Computer vision: algorithms and architectures

摘要: Ever since computers were used for pattern recognition, image processing, and more generally vision, a number of special-purpose algorithms architectures have been developed. As new reached the construction stage, different classes emerged in order to produce effective efficient solutions new, heavy burdens color moving images, stereo real-time performance. The multiprocessor machines (including variety interconnection patterns, memory organization, control structures, input-output management) stimulated algorithm designers develop suitable data choice appropriate primitive operations, adequate sequencing input data, use concurrency local computations, etc. This chapter presents review some basic image-processing implemented on machines. Algorithms connected component labeling, line detection, matching are considered systolic, mesh, tree, pyramid analysis evaluation these may hopefully lead their real applications cost-effective way.