Sum-box technique for fast linear filtering

Authors:
Jun Shen;Wei Shen;Serge Castan;Tianxu Zhang
Affiliations:
EGID Institute, Image Laboratory, Bordeaux-3 University, Allee F. Daguin, 33607 Pessac cedex, France;Department of Electronics, Zhongshan University, Canton, China;IRIT, Paul Sabatier University, 31062 Toulouse cedex, France;Institute of IR&AI, Huazhong University of Sciences and Technology, Wuhan, China
Venue:
Signal Processing
Year:
2002

Citing 6
Cited 2

Efficient Synthesis of Gaussian Filters by Cascaded Uniform Filters

IEEE Transactions on Pattern Analysis and Machine Intelligence
Filtering by repeated integration

SIGGRAPH '86 Proceedings of the 13th annual conference on Computer graphics and interactive techniques
Random signals and systems

Random signals and systems
A real-time audio frequency cubic spline interpolator

Signal Processing
Polynomial preserving algorithm for digital image interpolation

Signal Processing
Learning and approximation capabilities of adaptive spline activation function neutral networks

Neural Networks

Wheeze detection based on time-frequency analysis of breath sounds

Computers in Biology and Medicine
Iris recognition based on non-local comparisons

SINOBIOMETRICS'04 Proceedings of the 5th Chinese conference on Advances in Biometric Person Authentication

Quantified Score

Hi-index	0.08

Visualization

Abstract

Linear filters are widely used in signal processing and in image processing, and the fast realization of large kernel filters has been an important research subject. Traditional fast algorithms try to decompose the initial kernel into the convolution of smaller kernels, and a lot of multiplications are still needed to realize all the convolutions with the small kernels. Box technique can realize the Gaussian filters with little multiplications, but it cannot realize linear filters other than the Gaussians. In the present paper, by use of the analysis on the scaled Spline functions, we propose the sum-box filter technique to approximately realize a given large kernel linear filter (even non Gaussian type) to a factor by the sum of the translated outputs of sum-box filters, requiring no multiplications. This original sum-box technique presented opens thus a new orientation for linear filter fast realization: realizing to a factor the convolution with a large filter kernel by additions only. From the viewpoint of computational complexity, this method is highly efficient in image and signal processing, in particular, for large filter kernels. Experimental results are reported.