On the Computational Complexity of 2D Filtering by Winograd Method

Abstract

Increasing the speed of digital image processing is an important problem in various fields. This paper proposes a way to increase the speed of 2D image filtering using the Winograd method. New algorithms for image processing based on the Winograd method have been proposed. The theoretical analysis results showed that an increase in the number of pixels in the resulting image fragment leads to a decrease in multiplications and the computational complexity of digital filtering. Winograd methods reduced multiplications to 4 times by increasing additions to 2.88 times using filter 3{x}3. The results obtained can be used in various image processing fields that require real-time processing using modern microelectronic devices.