Affiliation:
1. Smart Engines Service LLC, 117312 Moscow, Russia
2. Faculty of Mechanics and Mathematics, Lomonosov Moscow State University, 119991 Moscow, Russia
3. Federal Research Center “Computer Science and Control”, Russian Academy of Sciences, 119333 Moscow, Russia
Abstract
Gaussian filtering, being a convolution with a Gaussian kernel, is a widespread technique in image analysis and computer vision applications. It is the traditional approach for noise reduction. In some cases, performing the exact convolution can be computationally expensive and time-consuming. To address this problem, approximations of the convolution are often used to achieve a balance between accuracy and computational efficiency, such as with running sums, Bell blur, Deriche approximation, etc. At the same time, modern computing devices support data parallelism (vectorization) via Single Instruction Multiple Data (SIMD) and can process integer numbers faster than floating-point approaches. In this paper, we describe several methods for approximating a Gaussian filter, implement the SIMD and quantized versions, and compare them in terms of speed and accuracy. The experiments were performed on central processing units with a x86_64 architecture using a family of SSE SIMD extensions and an ARMv8 architecture using the NEON SIMD extension. All the optimized approximations demonstrated 10–20× speedup while maintaining the accuracy in the range of 1 × 10−5 or higher. The fastest method is a trivial Stack blur with a relatively high error, so we recommend using the second-order Vliet–Young–Verbeek filter and quantized Bell blur and running sums as more accurate and still computationally efficient alternatives.
Funder
Ministry of Science and Higher Education of the Russian Federation
Reference33 articles.
1. Devi, T.G., Patil, N., Rai, S., and Philipose, C.S. (2023). Gaussian blurring technique for detecting and classifying acute lymphoblastic leukemia cancer cells from microscopic biopsy images. Life, 13.
2. Mewada, H., Al-Asad, J.F., Almalki, F.A., Khan, A.H., Almujally, N.A., El-Nakla, S., and Naith, Q. (2023). Gaussian-Filtered High-Frequency-Feature Trained Optimized BiLSTM Network for Spoofed-Speech Classification. Sensors, 14.
3. Abuya, T.K., Rimiru, R.M., and Okeyo, G.O. (2023). An Image Denoising Technique Using Wavelet-Anisotropic Gaussian Filter-Based Denoising Convolutional Neural Network for CT Images. Appl. Sci., 13.
4. Study of Keypoints Detectors and Descriptors Performance on X-ray Images Compared to the Visible Light Spectrum Images;Chekanov;IEEE Access,2022
5. An accurate approach to real-time machine-readable zone detection with mobile devices;Gayer;Lecture Notes in Computer Science (LNCS),2023