Signal modeling for two-dimensional image structures
Journal of Visual Communication and Image Representation
Locally Rotation, Contrast, and Scale Invariant Descriptors for Texture Analysis
IEEE Transactions on Pattern Analysis and Machine Intelligence
Continuous dimensionality characterization of image structures
Image and Vision Computing
Analysis of multiple orientations
IEEE Transactions on Image Processing
Representing local structure using tensors II
SCIA'11 Proceedings of the 17th Scandinavian conference on Image analysis
Non-local adaptive structure tensors
Image and Vision Computing
Adaptive robust structure tensors for orientation estimation and image segmentation
ISVC'05 Proceedings of the First international conference on Advances in Visual Computing
Scale-Space'05 Proceedings of the 5th international conference on Scale Space and PDE Methods in Computer Vision
A theory of multiple orientation estimation
ECCV'06 Proceedings of the 9th European conference on Computer Vision - Volume Part II
The monogenic curvature scale-space
IWCIA'06 Proceedings of the 11th international conference on Combinatorial Image Analysis
Accurate junction detection and characterization in line-drawing images
Pattern Recognition
| Hi-index | 0.00 |
The boundaries of image regions necessarily consist of edges (inparticular, step and roof edges), corners, and junctions.Currently, different algorithms are used to detect each boundarytype separately, but the integration of the results into a singleboundary representation is difficult. Therefore, a method for thesimultaneous detection of all boundary types is needed. We proposeto combine responses of suitable polar separable filters into whatwe will call the boundary tensor. The trace of this tensor is ameasure of boundary strength, while the small eigenvalue and itsdifference to the large one represent corner / junction and edgestrengths respectively. We prove that the edge strength measurebehaves like a rotationally invariant quadrature filter. A numberof examples demonstrate the properties of the new method andillustrate its application to image segmentation.