8. Local Filters

  1. Donald G. Bailey

Published Online: 2 JUN 2011

DOI: 10.1002/9780470828519.ch8

Design for Embedded Image Processing on FPGAs

Design for Embedded Image Processing on FPGAs

How to Cite

Bailey, D. G. (2011) Local Filters, in Design for Embedded Image Processing on FPGAs, John Wiley & Sons (Asia) Pte Ltd, Singapore. doi: 10.1002/9780470828519.ch8

Author Information

  1. Massey University, New Zealand

Publication History

  1. Published Online: 2 JUN 2011
  2. Published Print: 1 JUN 2011

ISBN Information

Print ISBN: 9780470828496

Online ISBN: 9780470828519



  • field programmable gate array (FPGA);
  • image processing;
  • linear filters;
  • morphological filters;
  • nonlinear filters;
  • rank filters


Local filters extend point operations by making the output depend not only in the corresponding input pixel value, but also its local context. To implement filters efficiently on a field programmable gate array (FPGA), it is necessary to cache the input values as they are loaded so that each pixel is only loaded once. Linear filters are arguably the most widely used class of local filters. The output of a linear filter is a linear combination of the input pixels within the window. Two important classes of nonlinear filters considered in some detail are rank filters and morphological filters. A range of efficient structures for implementing these filters has been described in some detail. Local filters are an essential part of any image processing application. Therefore, the range of techniques and principles described in this chapter is indispensible for accelerating any embedded image processing application.

Controlled Vocabulary Terms

image processing; nonlinear filters