Multi-Pipeline Architecture For Real-Time Low-Level Image Processing

Z. Chen; C. C. Chang; S. Y. Lin; R. H . Hwang

doi:10.1117/12.942813

Translator Disclaimer

19 February 1988 Multi-Pipeline Architecture For Real-Time Low-Level Image Processing

Z. Chen, C. C. Chang, S. Y. Lin, R. H . Hwang

Author Affiliations +

Proceedings Volume 0848, Intelligent Robots and Computer Vision VI; (1988) https://doi.org/10.1117/12.942813
Event: Advances in Intelligent Robotics Systems, 1987, Cambridge, CA, United States

Abstract

We shall present two kinds of multi-pipeline architecture for real-time low-level image processing. The first one is to be used in the general local operations of image processing with a fixed 3*3 window size. The second one is to be used in a less general local operations with a dynamic window size. The image processing functions executable in these architectures include Sobel edge operator, Laplacian high-pass filter, smoothing operation, erosion, and dilation, etc. The raw image is inputted into the frame buffer of the parallel processor. Different schemes are designed to set up the data flow path among the processing elements or cells.

Citation Download Citation

Z. Chen, C. C. Chang, S. Y. Lin, and R. H . Hwang "Multi-Pipeline Architecture For Real-Time Low-Level Image Processing", Proc. SPIE 0848, Intelligent Robots and Computer Vision VI, (19 February 1988); https://doi.org/10.1117/12.942813

ACCESS THE FULL ARTICLE