Multiple line-template matching with the EM algorithm
Pattern Recognition Letters - special issue on pattern recognition in practice V
Using local planar geometric invariants to match and model images of line segments
Computer Vision and Image Understanding
Topographic Maps and Local Contrast Changes in Natural Images
International Journal of Computer Vision
IEEE Transactions on Pattern Analysis and Machine Intelligence
Stochastic Tracking of 3D Human Figures Using 2D Image Motion
ECCV '00 Proceedings of the 6th European Conference on Computer Vision-Part II
Using Geometric Properties for Correspondence-less Image Alignment
ICPR '98 Proceedings of the 14th International Conference on Pattern Recognition-Volume 1 - Volume 1
Multiple View Geometry in Computer Vision
Multiple View Geometry in Computer Vision
Model-based image matching using location (pattern, recognition)
Model-based image matching using location (pattern, recognition)
Two-dimensional projective point matching
Two-dimensional projective point matching
Efficient cumulative matching for image registration
Image and Vision Computing
Matching Images Using Invariant Level-line Primitives under Projective Transformation
CRV '10 Proceedings of the 2010 Canadian Conference on Computer and Robot Vision
Mixed color/level lines and their stereo- matching with a modified Hausdorff distance
Integrated Computer-Aided Engineering
Automatic image search based on improved feature descriptors and decision tree
Integrated Computer-Aided Engineering
Establishing the correspondence between control points in pairs of mammographic images
IEEE Transactions on Image Processing
Fast computation of a contrast-invariant image representation
IEEE Transactions on Image Processing
Projection-based image registration in the presence of fixed-pattern noise
IEEE Transactions on Image Processing
Human automatic detection and tracking for outdoor video
Integrated Computer-Aided Engineering
| Hi-index | 0.00 |
Planar geometric transformations approximately model differences between images from a moving camera. Our registration technique consists of finding and matching featured primitives, invariant in the observed scene. The invariant shape to be maintained constrains the kind of approximation. Primitives stem from level-lines, inheriting their robustness towards contrast changes. The registration still improves it through efficient cumulative matching based on a multi-stage primitive election procedure. This paper is a continuation of a preliminary work on the simplest geometric transform, similarity, constructing two-segment-primitives. Our contribution is to validate further geometric transforms, completing the path "similarity, affine, projective". Transformations are stable when they are computed on planar objects or from scenes which contain many coplanar facets and elements. Our approach works with cluttered images, and even if the estimation is done globally while the apparent displacement is not small and there are several different unknown motions in the scene. Results obtained with selected shapes of primitives are shown and compared in the corresponding sections. A gauge of transform goodness is elaborated, based on an assumption of spatial ergodicity incentive to estimate a conditional probability of finding similar pixel values in a neighborhood of the original and transformed images respectively.