Hardware versus software implementations of fast image processing algorithms

In this paper the construction of fast algorithms on sequential machines, the art of programming dedicated parallel machines and a design philosophy for dedicated image processing hardware is treated and their trade-off is discussed. Examples of different image processing systems used and built in our group are given and evaluated. The conclusions are that for each specific class of image processing operations, effort has to be put in the choice of the optimum data representation and in the means for fast access of the data structure incorporated with that representation. This holds for software as well as for hardware. For special hardware designs, effort has to be put in the interface with the host processor and in a joint programming environment of the special purpose machine and its host system.

[1]  H. T. Kung.  Why systolic architectures? , 1982, Computer.

[2]  Frans C. A. Groen,et al.  Distance Transforms with Data Flow Techniques , 1987, ASST.

[3]  Kendall Preston,et al.  Digital processing of biomedical images , 1976 .

[4]  Frans A. Gerritsen,et al.  Design and use of DIP-1: A fast, flexible and dynamically microprogrammable pipelined image processor , 1981, Pattern Recognit..

[5]  Frans C. A. Groen,et al.  Image processing with personal computers , 1988 .

[6]  Jim Piper.  Efficient implementation of skeletonisation using interval coding , 1985, Pattern Recognit. Lett..

[7]  Robert P. W. Duin,et al.  A Hardware Design For Fast 2-D Percentile Filtering , 1986, Other Conferences.

[8]  C. J. Hilditch,et al.  Linear Skeletons From Square Cupboards , 1969 .

[9]  Josef Kittler,et al.  Image processing system architectures , 1985 .

[10]  King-Sun Fu,et al.  Special Computer Architectures for Pattern Processing , 1982 .

[11]  Kendall Preston,et al.  Modern Cellular Automata , 1984, Advanced Applications in Pattern Recognition.

[12]  Frans C. A. Groen,et al.  A fast algorithm for cellular logic operations on sequential machines , 1984, Pattern Recognit. Lett..

[13]  William K. Pratt,et al.  Review Of Machine Vision Architectures , 1987, Photonics West - Lasers and Applications in Science and Engineering.

[14]  Stefano Levialdi,et al.  Parallel thinning of binary pictures , 1975 .

[15]  S. Eiho,et al.  Processing of RI-Angiocardiographic Images , 1976 .

[16]  Robert P. W. Duin,et al.  Fast percentile filtering , 1986, Pattern Recognit. Lett..

[17]  Gösta H. Granlund,et al.  Computer Architectures for Image Processing. , 1985 .

[18]  Frans C. A. Groen,et al.  Design and Implementation of a Cellular-Logic VME Processor Module , 1986, Other Conferences.

[19]  Thomas S. Huang,et al.  A fast two-dimensional median filtering algorithm , 1979 .

[20]  Ian Watson,et al.  A Practical Data Flow Computer , 1982, Computer.