Fast Image Edge Detection based on Faber Schauder Wavelet and Otsu Threshold

Area of Science:

• Computer Vision
• Image Processing
• Signal Processing

Background:

• Edge detection is crucial for computer vision tasks like image segmentation and object detection.
• Existing methods often struggle with precision and computational complexity.
• Developing efficient and accurate edge detection techniques remains an active research area.

Purpose of the Study:

• To propose a novel, low-complexity, and precise edge detection method.
• To leverage the properties of Faber Schauder Wavelet (FSW) and Otsu thresholding for multi-scale edge detection.
• To demonstrate the method's suitability for real-time applications.

Main Methods:

• Image smoothing using a bilateral filter based on noise estimation.
• Selection of FSW extrema coefficients using Otsu thresholding.
• Edge point linking via a predictive edge linking algorithm.

Main Results:

• The proposed method effectively detects edges in images.
• Experimental results show the method is faster than state-of-the-art approaches.
• The technique demonstrates potential for real-time computer vision applications.

Conclusions:

• The FSW and Otsu thresholding method offers an efficient solution for edge detection.
• The approach achieves a balance between precision and low complexity.
• This method is a viable candidate for real-time edge detection in computer vision systems.