Radim Šára

Finding the Largest Unambiguous Component of Stereo Matching

Stereo matching is an ill-posed problem for at least two principal reasons: (1) because of the random nature of match similarity measure and (2) because of structural ambiguity due to repetitive patterns. Both ambiguities require the problem to be posed in the regularization framework. Continuity is a natural choice for a prior model. But this model may fail in low signal-to-noise ratio regions. The resulting artefacts may then completely spoil the subsequent visual task.A question arises whether one could (1) find the unambiguous component of matching and, simultaneously, (2) identify the ambiguous component of the solution and then, optionally, (3) regularize the taskfor the ambiguous component only. Some authors have already taken this view. In this paper we define a new stability property which is a condition a set of matches must satisfy to be considered unambiguous at a given confidence level. It turns out that for a given matching problem this set is (1) unique and (2) it is already a matching. We give a fast algorithm that is able to find the largest stable matching. The algorithm is then used to show on real scenes that the unambiguous component is quite dense (10-80%) and error-free (total error rate of 0.3-1.4%), both depending on the confidence level chosen.

Efficient Sampling of Disparity Space for Fast And Accurate Matching

A simple stereo matching algorithm is proposed that visits only a small fraction of disparity space in order to find a semi-dense disparity map. It works by growing from a small set of correspondence seeds. Unlike in known seed-growing algorithms, it guarantees matching accuracy and correctness, even in the presence of repetitive patterns. This success is based on the fact it solves a global optimization task. The algorithm can recover from wrong initial seeds to the extent they can even be random. The quality of correspondence seeds influences computing time, not the quality of the final disparity map. We show that the proposed algorithm achieves similar results as an exhaustive disparity space search but it is two orders of magnitude faster. This is very unlike the existing growing algorithms which are fast but erroneous. Accurate matching on 2-megapixel images of complex scenes is routinely obtained in a few seconds on a common PC from a small number of seeds, without limiting the disparity search range.

Specularities Reduce Ambiguity of Uncalibrated Photometric Stereo

Lambertian photometric stereo with uncalibrated light directions and intensities determines the surface normals only up to an invertible linear transformation. We show that if object reflectance is a sum of Lambertian and specular terms, the ambiguity reduces into a 2dof group of transformations (compositions of isotropic scaling, rotation around the viewing vector, and change in coordinate frame handedness).Such ambiguity reduction is implied by the consistent viewpoint constraint which requires that all lights reflected around corresponding specular normals must give the same vector (the viewing direction). To employ the constraint, identification of specularities in images corresponding to four different point lights in general configuration suffices. When the consistent viewpoint constraint is combined with integrability constraint, binary convex/concave ambiguity composed with isotropic scaling results. The approach is verified experimentally.We observe that an analogical result applies to the case of uncalibrated geometric stereo with four affine cameras in a general configuration observing specularities from a single distant point light source.

Image texture analysis of sonograms in chronic inflammations of thyroid gland

The current practice in assessing sonographic findings of chronic inflamed thyroid tissue is mainly qualitative, based just on a physicians experience. This study shows that inflamed and healthy tissues can be differentiated by automatic texture analysis of B-mode sonographic images. Feature selection is the most important part of this procedure. We employed two selection schemes for finding recognition-optimal features: one based on compactness and separability and the other based on classification error. The full feature set included Muzzolinis spatial features and Haralicks co-occurrence features. These features were selected on a set of 2430 sonograms of 81 subjects, and the classifier performance was evaluated on a test set of 540 sonograms of 18 independent subjects. A classification success rate of 100% was achieved with as few as one optimal feature among the 129 texture characteristics tested. Both selection schemes agreed on the best features. The results were confirmed on the independent test set. The stability of the results with respect to sonograph setting, thyroid gland segmentation and scanning direction was tested.

On occluding contour artifacts in stereo vision

We study occluding contour artifacts in area-based stereo matching: they are false responses of the matching operator to the occlusion boundary and cause the objects extend beyond their true boundaries in disparity maps. Most of the matching methods suffer from these artifacts; the effect is so strong that it cannot be ignored. We show what gives rise to the artifacts and design a matching criterion that accommodates the presence of occlusions as opposed to methods that identify and remove the artifacts. This approach leads to the problem of measurement contamination studied in statistics. We show that such a problem is hard given finite computational resources, unless more independent measurements directly related to occluding contours is available. What can be achieved is a substantial reduction of he artifacts, especially for large matching templates. Reduced artifacts allow for easier hierarchical matching and for easy fusion of reconstructions from different viewpoints into a coherent whole.

Fish-scales: representing fuzzy manifolds

We address the problem of automatically reconstructing m-manifolds of unknown topology from unorganized points in metric p-spaces obtained from a noisy measurement process . The point set is first approximated by a collection of oriented primitive fuzzy sets over a range of resolutions. Hierarchical multiresolution representation is then computed based on the relation of relative containment defined on the collection. Finally, manifold structure is recovered by establishing connectivity between these primitives based on proximity, compatibility of position and orientation and local topological constraints. The method has been successfully applied to the problem of surface reconstruction from polynocular-stereo data with many outliers.

Unambiguous determination of shape from photometric stereo with unknown light sources

Photometric stereo with uncalibrated lights determines surface orientations ambiguously up to any regular transformation. If the surface reflectance model is separable with respect to the illumination and viewing directions, its inherent symmetries enable to design two previously unrecognized constraints on normals that reduce this ambiguity. The two constraints represent projections of normals onto planes perpendicular to the viewing and illumination directions, respectively. We identify the classes of transformations that leave each constraint invariant. We construct the constraints using polarization measurement under the assumption of separable reflectance model for smooth dielectrics. We verify that applying the first constraint together with the integrability constraint results in bas-relief ambiguity, while application of the second constraint on integrable normals reduces the ambiguity to convex/concave ambiguity. Importantly, the latter result is also obtained when the first and second constraints alone are combined.

Stratified Dense Matching for Stereopsis in Complex Scenes.

Local joint image modeling in stereo matching brings more discriminable and stable matching features. Such features reduce the need for strong prior models (continuity) and thus algorithms that are less prone to false positive artefacts in general complex scenes can be applied. One of the principal quality factors in area-based dense stereo is the matching window shape. As it cannot be selected without having any initial matching hypothesis we propose a stratified matching approach. The window adapts to high-correlation structures in disparity space found in pre-matching which is then followed by final matching. In a rigorous ground-truth experiment we show that Stratified Dense Matching is able to increase matching density 3×, matching accuracy 1.8×, and occlusion boundary detection 2× as compared to a fixed-size rectangular windows algorithm. Performance on real outdoor complex scenes is also evaluated.

Spatial Pattern Templates for Recognition of Objects with Regular Structure

We propose a method for semantic parsing of images with regular structure. The structured objects are modeled in a densely connected CRF. The paper describes how to embody specific spatial relations in a representation called Spatial Pattern Templates (SPT), which allows us to capture regularity constraints of alignment and equal spacing in pairwise and ternary potentials.

3D geometry from uncalibrated images

We present an automatic pipeline for recovering the geometry of a 3D scene from a set of unordered, uncalibrated images. The contributions in the paper are the presentation of the system as a whole, from images to geometry, the estimation of the local scale for various scene components in the orientation-topology module, the procedure for orienting the cloud components, and the method for dealing with points of contact. The methods are aimed to process complex scenes and non-uniformly sampled, noisy data sets.