Title :
Architectures for generalized 2D FIR filtering using separable filter structures
Author_Institution :
LOGIC Devices Inc., Sunnyvale, CA, USA
Abstract :
The problem of generalized 2D FIR filtering for large filter kernel sizes can be computationally prohibitive when required in real-time, such as with video applications. In this paper, we describe architectures and design methods for generalized 2D FIR filtering employing LOGIC Devices´ LF33xx (HV filter) family of devices. The LF33xx family of devices is designed to perform dimensionally separate filtering (row/column) along orthogonal axes (horizontal direction and vertical direction for video applications). Additionally, we will briefly review 2D filter design and kernel separability theory
Keywords :
FIR filters; singular value decomposition; two-dimensional digital filters; video signal processing; HV filter; LOGIC Devices LF33xx family; SVD; design methods; dimensionally separate filtering; filter architectures; generalized 2D FIR filtering; kernel separability theory; orthogonal axes; separable filter structures; video applications; Computer architecture; Design methodology; Eigenvalues and eigenfunctions; Filtering theory; Finite impulse response filter; Frequency response; Kernel; Logic devices; Matrix decomposition; Singular value decomposition;
Conference_Titel :
Acoustics, Speech, and Signal Processing, 1999. Proceedings., 1999 IEEE International Conference on
Conference_Location :
Phoenix, AZ
Print_ISBN :
0-7803-5041-3
DOI :
10.1109/ICASSP.1999.758376
