Wavelet-Based Signal Processing for Image Edge Detection
PDF

Keywords

wavelet transform
image edge detection
multi-resolution analysis
discrete wavelet transform (DWT)
image segmentation

Abstract

Edge detection serves as a fundamental process in the domain of image processing, providing essential data for tasks such as object recognition, image segmentation, medical diagnostics, and remote sensing. Traditional methods—such as Sobel, Prewitt, and Canny detectors—utilize gradient and Laplacian operators to highlight intensity discontinuities. However, these approaches often falter in the presence of noise, texture variations, or low contrast regions. Recent advancements in signal processing have led to the integration of wavelet transforms as a robust alternative. Wavelets allow simultaneous localization in both spatial and frequency domains, enabling more accurate multi-resolution edge detection.This study explores the use of Discrete Wavelet Transform (DWT) for generating noise-resistant edge maps. The process involves decomposing the image into different frequency components and reconstructing edges using wavelet modulus maxima and thresholding techniques. Performance evaluations reveal that wavelet-based edge detection provides superior precision, edge continuity, and robustness compared to classical methods. This paper also investigates computational strategies for accelerating wavelet processing and discusses real-world applications in medical imaging, satellite surveillance, and autonomous systems.

PDF
Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 International License.

Copyright (c) 2024 Dr. Elena V. Petrova (Author)