Abstract
Many image processing applications require fast convolution of an image with one or more 2D filters. Field-Programmable Gate Arrays (FPGAs) are often used to achieve this goal due to their fine grain parallelism and reconfigurability. However, the heterogeneous nature of modern reconfigurable devices is not usually considered during design optimization. This article proposes an algorithm that explores the space of possible implementation architectures of 2D filters, targeting the minimization of the required area, by optimizing the usage of the different components in a heterogeneous device. This is achieved by exploring the heterogeneous nature of modern reconfigurable devices using a Singular Value Decomposition based algorithm, which provides an efficient mapping of filter's implementation requirements to the heterogeneous components of modern FPGAs. In the case of multiple 2D filters, the proposed algorithm also exploits any redundancy that exists within each filter and between different filters in the set, leading to designs with minimized area. Experiments with real filter sets from computer vision applications demonstrate an average of up to 38% reduction in the required area.
- Andrews, M. S. 1999. Architectures for generalized 2d fir filtering using separable filter structures. In Proceedings of the Acoustics, Speech, and Signal Processing IEEE International Conference. Vol. 4. 2215--2218. Google Scholar
Digital Library
- Belhumeur, P., Hespanha, J., and Kriegman, D. 1997. Eigenfaces vs. fisherfaces: recognition using class specific linear projection. IEEE Trans. Patt. Anal. Mach. Intell. 19, 7, 711--720. Google Scholar
Digital Library
- Bouganis, C.-S., Cheung, P. Y. K., and Constantinides, G. A. 2005. Heterogeneity Exploration for Multiple 2D Filter Designs. In Proceedings of the Conference on Field Programmable Logic and Applications. 263--268.Google Scholar
- Bouganis, C.-S., Cheung, P. Y. K., Ng, J., and Bharath, A. A. 2004. A Steerable Complex Wavelet Construction and its Implementation on FPGA. In Proceedings of the Conference on Field Programmable Logic and Applications. 394--403.Google Scholar
- Bouganis, C.-S., Constantinides, G. A., and Cheung, P. Y. K. 2005a. A Novel 2D Filter Design Methodology. In Proceedings of the International Symposium in Circuits and Systems. 532--535.Google Scholar
- Bouganis, C.-S., Constantinides, G. A., and Cheung, P. Y. K. 2005b. A Novel 2D Filter Design Methodology For Heterogeneous Devices. In Proceedings of the Conference on Field-Programmable Custom Computing Machines. 1--10. Google Scholar
Digital Library
- Chen, C.-L., Khoo, K.-Y., and A.N. Willson,Jr. 1995. An improved polynomial-time algorithm for designing digital filters with power-of-two coefficients. In Proceedings of the IEEE International Symposium on Circuits and Systems. Vol. 1. 223--226.Google Scholar
- Cmar, R., Rijnders, L., Schaumont, P., Vernalde, S., and Bolsens, I. 1999. A methodology and design environment for DSP ASIC fixed point refinement. In Proceedings of the Design, Automation, and Test in Europe. 271--276. Google Scholar
Digital Library
- Constantinides, G. A., Cheung, P., and Luk, W. 2001. The Multiple Wordlength Paradigm. In Proceedings of the Conference on Field-Programmable Custom Computing Machines. 51--60. Google Scholar
Digital Library
- Constantinides, G. A., Cheung, P. Y. K., and Luk, W. 2003. Wordlength Optimization for Linear Digital Signal Processing. IEEE Trans. Comput.-Aid. Des. Integr. Circ. Syst. 22, 10. Google Scholar
Digital Library
- Dempster, A. and Macleod, M. D. 1995. Use of minimum-adder multiplier blocks in FIR digital filters. IEEE Trans. Circ. Syst. II 42, 569--577.Google Scholar
Cross Ref
- Gong, S., McKenna, S., and Psarrou, A. 2000. Dynamic Vision: From Images to Face Recognition 1st Ed. Imperial College Press. Google Scholar
Digital Library
- Haseyama, M. and Matsuura, D. 2006. A filter coefficient quantization method with genetic algorithm, including simulated annealing. IEEE Signal Process. Lett. 13, 4, 189--192.Google Scholar
Cross Ref
- Hastad, J. 1990. Tensor rank is NP-complete. J. Algo. 11, 4, 644--654. Google Scholar
Digital Library
- Ja'Ja, J. 1978. Optimal evaluation of pairs of bilinear forms. In Proceedings of the 10th Annual ACM Symposium on Theory of Computing. 173--182. Google Scholar
Digital Library
- Kodek, D. 1980. Design of Optimal Finite Wordlength FIR Digital Filters Using Integer Linear Programming Techniques. IEEE Trans. Acoust. Speech Signal Process. 28, 304--308.Google Scholar
Cross Ref
- Koren, I. 2002. Computer Arithmetic Algorithms 2nd Ed. Prentice-Hall Inc. Google Scholar
Digital Library
- Kum, K.-I. and Sung, W. 2001. Combined word-length optimization and high-level synthesis of digital signal processing systems. IEEE Trans. Comput.-Aid. Des. Integr. Circ. Syst. 20, 8, 921--930. Google Scholar
Digital Library
- Li, D., Lim, Y. C., Lian, Y., and Song, J. 2002. A polynomial-time algorithm for designing fir filters with power-of-two coefficients. IEEE Trans. Acoust. Speech Signal Process. 50, 8, 1935--1941.Google Scholar
Digital Library
- Li, Z. 2003. V1 mechanisms and some figure-ground and border effects. J. Physiol. Paris 97, 503--515.Google Scholar
Cross Ref
- Mitra, S. K. 2002. Digital Signal Processing: A Computer-Based Approach 2nd Ed. McGraw-Hill Higher Education. Google Scholar
Digital Library
- Park, I.-C. and Kang, H.-J. 2001. Digital filter synthesis based on minimal signed digit representation. In Proceedings of the Annual ACM IEEE Design Automation Conference. 468--473. Google Scholar
Digital Library
- Pasko, R., Schaumont, P., Derudder, V., Vernalde, S., and Durackova, D. 1999. A new algorithm for elimination of common subexpressions. IEEE Trans. Comput.-Aid. Des. Integra. Circ. Syst. 18, 1, 58--68. Google Scholar
Digital Library
- Press, W., Teukolsky, S., Vetterling, W., and Flannery, B. 1992. Numerical Recipes in C. Cambridge University Press.Google Scholar
- Samueli, H. 1989. An improved search algorithm for the design of multiplierless fir filters with powers-of-two coefficients. IEEE Trans. Circ. Syst. 36, 7, 1044--1047.Google Scholar
Cross Ref
- Shashua, A. and Levin, A. 2001. Linear image coding for regression and classification using the tensor-rank principle. In Proceedings of the Computer Vision and Pattern Recognition Conference. Vol. I. IEEE, 42--49.Google Scholar
- Siohan, P. 1990. An analysis of coefficient inaccuracy for 2-d fir direct form digital filters. IEEE Trans. Circ. Syst. 37, 10, 1308--1313.Google Scholar
Cross Ref
- Strang, G. 1998. Introduction to Linear Algebra, 3rd ed. Wellesley-Cambridge Press.Google Scholar
- Yurdakul, A. 2005. Multiplierless implementation of 2-d fir filters. Integration VLSI J. 38, 4, 597--613. Google Scholar
Digital Library
Index Terms
Synthesis and Optimization of 2D Filter Designs for Heterogeneous FPGAs
Recommendations
An evolvable hardware system in Xilinx Virtex II Pro FPGA
In this paper, a new circuit architecture for image filter evolution is proposed. The evolvable system is based on the implementation of a search algorithm in the PowerPC processor which is available in Xilinx Virtex II Pro Field Programmable Gate ...
VHDL description of a synthetizable and reconfigurable real-time stereo vision processor
ISPRA'05: Proceedings of the 4th WSEAS International Conference on Signal Processing, Robotics and AutomationThis paper describes a reconfigurable digital architecture to compute dense disparity maps at video-rate for stereo vision. The processor architecture is described in synthetizable VHDL and, by means of the reconfigurability, the hardware requirements ...
DRAF: a low-power DRAM-based reconfigurable acceleration fabric
ISCA'16FPGAs are a popular target for application-specific accelerators because they lead to a good balance between flexibility and energy efficiency. However, FPGA lookup tables introduce significant area and power overheads, making it difficult to use FPGA ...








Comments