This paper presents a study of the impact of MMX technology in image processing and machine vision application, which, because of their hard real time constrains, is an undoubtedly challenging task. A comparison with a traditional scalar code and with another parallel SIMD architecture (IMAP-VISION board) is discussed with emphasize of the particular programming strategies for speed optimization. More precisely we discuss the low level and intermediate level image processing algorithms, which are best suited for parallel SIMD implementation. High-level image processing algorithms are more suitable for parallel implementation on MIMD architectures. While the IMAP-Vision system performs better because of the large number of processing elements, the MMX processor remains a good candidate for low-level image processing.
[1]
P.P. Jonker,et al.
Bucket processing: a paradigm for image processing
,
1996,
Proceedings of 13th International Conference on Pattern Recognition.
[2]
I. Coorporation,et al.
Using the rdtsc instruction for performance monitoring
,
1997
.
[3]
Shorin Kyo,et al.
Efficient Implementation of Image Processing Algorithms on Linear Processor Arrays Using the Data Parallel Language IDC
,
1996,
MVA.
[4]
Shin'ichiro Okazaki,et al.
IMAP-VISION: An SIMD Processor with High-Speed On-chip Memory and Large Capacity External Memory
,
1996,
MVA.
[5]
Robert P. W. Duin,et al.
Architectures for Multidimensional Low- and Intermediate Level Image Processing (Invited)
,
1990,
MVA.