Clesio Luis Tozzi

Shepe from Shading with perspective projection and camera calibration

Abstract The majority of algorithms making use of the Shape from Shading technique to reconstruct a 3-D surface from its image employ the orthographic projection system which, in the case of real images, always causes deformation on the reconstructed surface. To remove this constraint a simultaneous procedure (Hybrid Model) is proposed which uses the Shape from Shading (SFS) technique combined with the photogrammetric technique and the perspective projection to reconstruct the surface and determine the camera calibration parameters. The procedure is based on an explicit mathematical model and an iterative process is adopted to solve the resulting system of equations. The Least Square Method (LSM) is adopted and the Taylors Series Lianearization technique is applied due to the non-linearity of the model. The proposed procedure differs from the procedures found in the literature in that it adopts the simultaneous computation of the tridimensional coordinates for surface point and the camera calibration parameters while employing perspective projection. In addition to that, solving the resulting system of equations does not demand any restriction on the surface, as it happens in other related procedures.

Towards 3D reconstruction of endoscope images using shape from shading

The 3D reconstruction problem from a single endoscope image of a smooth object is studied in the context of the shape from shading methods and considering a single light source at the camera projection center. Based on a curve expansion shape from shading algorithm, a spherical projection model for the endoscope camera and a dichromatic model for the surface reflectance, an approach to solve practical problems, namely the endoscope image distortion and the removal of the image specular reflection component, is presented. Results obtained from application of this approach to synthetic and real images are presented.

Surface reconstruction using multiple light sources and perspective projection

This paper presents a method to recover 3D geometry of Lambertian surfaces by using multiple images taken from the same view point and with the scene illuminated from different positions. This approach differs from stereo photometry in that it considers the light source at a finite distance from the object and the perspective projection in image formation. The proposed model allows local solution and recovery of 3D coordinates, in addition to surface orientation. A procedure to calibrate the light sources is also presented. Results of the application of the algorithm to synthetic images are shown.

Prediction of binding hot spot residues by using structural and evolutionary parameters

In this work, we present a method for predicting hot spot residues by using a set of structural and evolutionary parameters. Unlike previous studies, we use a set of parameters which do not depend on the structure of the protein in complex, so that the predictor can also be used when the interface region is unknown. Despite the fact that no information concerning proteins in complex is used for prediction, the application of the method to a compiled dataset described in the literature achieved a performance of 60.4%, as measured by F-Measure, corresponding to a recall of 78.1% and a precision of 49.5%. This result is higher than those reported by previous studies using the same data set.

Visual evaluation of the length of artificially generated electrical discharges by 3D-snakes

The secondary arc extinction is the main concern in single-phase auto reclosure (SPAR) switching studies, since it dictates whether the reclosure will be successful or not. In this context, a function that describes the arc length variation over time (l(t)) is an important information not only for dynamic response but also for the analysis of transient arc processes. Given that purely electrotechnical approach only gives a profile of the curve l(t), the application of techniques based on image analysis has been considered for obtaining such function. In this paper, a new approach based on 3D-reconstruction and the tracking of an electrical arc is proposed for performing a computer-based evaluation of variations in its length. It was applied a three-dimensional active contour named 3D-snake that is geometrically represented by a B-spline, which evolves in 3D space. This new model is much less dependent on the determination of homologous points than the ad hoc approaches found in the literature for recovering 3D geometry of electrical discharges based on image analysis. In addition, the proposal presented is capable of tracking the evolution of the electrical discharge taking into account the time dependence between consecutive pairs of frames in two videos providing the required function l(t).

COMPARACIÓN DE TÉCNICAS DE CALIBRACIÓN DE CÁMARAS DIGITALES

This work presents a comparison of the calibration techniques proposed by Tuceryan et al. [15], Trucco et al. [13] and Zhang [20], suggesting different ways to determine the parameters, starting from known 3D points and their corresponding points in the 2D image plane. The correspondence is manually extracted, being the obtained parameter accuracies (extrinsic and intrinsic) the most important. The passive patterns used consider the calibration techniques with different homographs, depending on the existing process planes. Zhang [20] proposes an approximation based on patterns located on a 2D planar surface, where the camera or the pattern are displaced without knowledge of the motion. Tuceryan et al. [15] and Trucco et al. [13] base their procedure in a known 3D model and its correspondence on a 2D image plane, the difference being in the consideration of the intrinsic parameters that must be determined or considered in the final camera matrix. The algorithms were implemented using Matlab 6.0, according to the interpretation of the works and optimization techniques extracted from the cited literature.

Prediction of Protein-Protein Binding Hot Spots: A Combination of Classifiers Approach

In this work we approach the problem of predicting protein binding hot spot residues through a combination of classifiers. We consider a comprehensive set of structural and chemical properties reported in the literature for characterizing hot spot residues. Each component classifier considers a specific set of properties as feature set and their output are combined by the mean rule. The proposed combination of classifiers achieved a performance of 56.6%, measured by the F-Measure with corresponding Recall of 72.2% and Precision of 46.6%. This performance is higher than those reported by Darnel et al. [4] for the same data set, when compared through a t-test with a significance level of 5%.

Discrimination of natural contours by means of time-scale-frequency decompositions

This paper evaluates the discriminative potential of time-scale-frequency decompositions for contour-based recognition of natural shapes. Specifically, it provides the analysis and comparison of descriptors derived from the Fourier Transform, the Short-Time Fourier Transform, the Wavelet Transform and the Multi-Resolution Fourier Transform. Linear Discriminant Analysis and Backward Sequential Selection are employed for dimensionality reduction and selection of the most significant features, respectively. A Bayesian Classifier is used for class discrimination. To improve discrimination, a hierarchical classification is adopted. The approaches are analyzed and compared considering experiments developed over digitalized leaves.

TOWARDS A VISUAL COMPUTING AND COMMUNICATION REFERENCE MODEL

Abstract In this paper, a reference model aiming to contribute to the discussion and formalization of the issues related to the integration of Computer Graphics (CG), Image Processing (IP), and Computer Vision (CV) is proposed. This model takes into account that CG, IP, and CV are overlapping areas, and that CG and CV have inverse objectives. Regarding existing and forthcoming standards, the proposed model addresses the problems of data representation, data reduction, and communication. The model is based on three main issues: abstraction of data into layers, user dialog, and information interchange. It fulfills the requirements from various application areas including multimedia.

Image registration and image interpolation in volume rendering of cross sections

We outline a method for registration of images of cross sections using the concept of generalized Hough transform (GHT). The approach may be useful in situations where automation should be a concern. To overcome known problems of noise of traditional GHT we have implemented a slight modified version of the basic algorithm. The modification consists of eliminating points of no interest in the process before the application of the accumulation step of the algorithm. This procedure minimizes the amount of accumulation points while reducing the probability of appearing of spurious peaks. Also, we apply image warping techniques to interpolate images among cross sections. This is needed where the distance of samples between sections is too large. Then it is suggested that the step of registration with GHT can help the interpolation automation by simplifying the correspondence between points of images. Some results are shown.