A Computational Approach to Edge Detection
IEEE Transactions on Pattern Analysis and Machine Intelligence
IEEE Transactions on Pattern Analysis and Machine Intelligence
Sphere-packings, lattices, and groups
Sphere-packings, lattices, and groups
Characterization of Signals from Multiscale Edges
IEEE Transactions on Pattern Analysis and Machine Intelligence
Directional and time-scale wavelet analysis
SIAM Journal on Mathematical Analysis
A filter bank for the directional decomposition of images: theoryand design
IEEE Transactions on Signal Processing
Shiftable multiscale transforms
IEEE Transactions on Information Theory - Part 2
Singularity detection and processing with wavelets
IEEE Transactions on Information Theory - Part 2
Multidirectional and multiscale edge detection via M-band wavelet transform
IEEE Transactions on Image Processing
The single-pass perceptual embedded zero-tree coding implementation on DSP
Computers & Mathematics with Applications
Hi-index | 0.09 |
The standard 2D wavelet transform (WT) has been an effective tool in image processing. In recent years, many new transforms have been proposed successively, such as curvelets, bandlets, directional wavelet transform etc, which inherit the merits of the standard WT, and are more adequate at the 2D image processing tasks. Intuitively, it seemed that applying these novel tools to edge detection should acquire finer performance. In this paper, we propose an edge detection approach based on directional wavelet transform which retains the separable filtering and the simplicity of computations and filter design from the standard 2D WT. In addition, the corresponding gradient magnitude is redefined and a new algorithm for non-maximum suppression is described. The experimental results of edge detection for several test images are provided to demonstrate our approach.