Shape from Shading

Our work focuses on the problem of automatic generation of viewpoint invariant geometric surface models from range data to be used in CAD/CAM systems. We developed two methods for recovering the surface. First is on the base of Shape from Shading method and second is on the base of Facet method. A comparison between the results obtained using the 2 methods is made. We also offer new techniques for the representation and CNC processing of the surfaces.

How do humans see three-dimensional shape based on two-dimensional shading? Much research has assumed that a 'light from above' bias solves the ambiguity of shape from shading. Counter to the 'light from above' bias, studies of Bayesian priors have found that such a bias can be swayed by other light cues. Despite the persuasive power of the Bayesian models, many new studies and books cite the original 'light from above' findings. Here I present a version of the Bayesian result that can be experienced. The perception of shape-from-shading was found here to be influenced by an external light source, even when the light was obstructed and did not directly illuminate a two-dimensional stimulus. The results imply that this effect is robust and not low-level in nature. The perception of shape from shading is not necessarily based on a hard-wired internal representation of lighting direction, but rather assesses the direction of lighting in the scene adaptively. Here, for the first time, is an experiential opportunity to see what the Bayesian models have supported all along.

Relief inversion effect is a perceptual phenomenon that leads to an inverted perception of convex and concave shapes. This perceptual inversion occurs in scenes where the shading/shadows act as the main depth cue. In visuospatial displays, such as shaded relief maps, the positioning of the shadows in the northern slopes, thus when light source placed broadly in south, mislead the cognitive system based on the 'light from above prior' assumption (Mamassian and Goutcher 2001). Thus, assuming the light must come from above, our mind creates an illusion, and we perceive the landforms incorrectly. To judge the 3D spatial relationships in terrain representations correctly, the relief inversion effect must be avoided. Cartographic convention against this effect is to place the light source at northwest (NW), whereas a recent study demonstrated that north-north-west (NNW), or even north yields more precise results . Since this finding goes against decades of convention, to establish its validity further, we attempted replicating the results with a different sample in South Africa. In this paper, we present our findings, which broadly confirm that the NNW (or also N) is better than NW against the relief inversion effect.

In this work, we propose a new method to enhance text in document-image. Firstly, we introduce a classical model and a way to solve it by means of a non-convex optimization problem. So, a simoultaneaous estimation of the reflectance and the luminance is obtained when the non uniform illumination (also called luminance) is a smooth function and the reflectance is a function of bounded variation. We give an analyse of this problem and some conditions of existence and unicity. Then, we consider the "log" of the classical model. A new pde's model is proposed. This method is based on the resolution of an original partial differential equation (PDE) estimating the log of the luminance. We assume that the luminance is enough smooth and is the solution of a non classical second order's PDE.Then we deduce the reflectance from the estimated luminance and the acquired image. The effectiveness and the robustness of the proposed process are shown on numerical examples in real-world situation (images acquired from cameraphones). Then, we illustrate the ability of this method to improve an Optical Character Recognition (OCR) in text recognition.

In this paper we analyze a specific problem within the context of recovering the geometric shape of an unknown surface from multiple noisy shading patterns generated by consecutive parallel illuminations by different light-sources. Shading-based single-view shape recovery in computer vision often leads to vector fields (i.e. estimated surface normals) which have to be integrated for calculations of height or depth maps. We present an algorithm for enforcing the integrability condition of a given non-integrable vector field which ensures a global suboptimal solution by local optimizations. The scheme in question relies neither on a priori knowledge of boundary conditions nor on other global constraints imposed on the sofar derived noise contaminated gradient integration techniques. The discussion is supplemented by examples illustrating algorithm performance.

Existing Photometric Stereo methods provide reasonable surface reconstructions unless the irradiance image is corrupted with noise and effects of digitisation. However, in real world situations the measured image is almost always corrupted, so an efficient method must be formulated to denoise the data. Once noise is added at the level of the images the noisy Photometric Stereo problem with a least squares estimate is transformed into a non-linear discrete optimization problem depending on a large number of parameters. One of the computationally feasible methods of performing this non-linear optimization is to use many smaller local optimizations to find a minimum (called 2D Leap-Frog). However, this process still takes a large amount of time using a single processor, and when realistic image resolutions are used this method becomes impractical. This paper presents a parallel implementation of the 2D Leap-Frog algorithm in order to provide an improvement in the time complexity. While the focus of this research is in the area of shape from shading, the iterative scheme for finding a local optimum for a large number of parameters can also be applied to any optimization problems in Computer Vision. The results presented herein support the hypothesis that a high speed up and high efficiency can be achieved using a parallel method in a distributed shared memory environment.

In 3-source photometric stereo, a Lambertian surface is illuminated from 3 known independent light-source directions, and photographed to give 3 images. The task of recovering the surface reduces to solving systems of linear equations for the gradients of a bivariate function u whose graph is the visible part of the surface [9], [16], [17], [24]. In the present paper we consider the same task, but with slightly more realistic assumptions: the photographic images are contaminated by Gaussian noise, and light-source directions may not be known. This leads to a non-quadratic optimization problem with many independent variables, compared to the quadratic problems resulting from addition of noise to the gradient of u and solved by linear methods in [6], [10], [20], [21], [22], [25]. The distinction is illustrated in Example 1 below. Perhaps the most natural way to solve our problem is by global Gradient Descent, and we compare this with the 2-dimensional . For this we review some mathematical results of [23] and describe an implementation in sufficient detail to permit code to be written. Then we give examples comparing the behaviour of Leap-Frog with Gradient Descent, and explore an extension of ) to estimate light source directions when these are not given, as well as the reflecting surface.

A smooth object depicted in a monochrome image will often exhibit brightness variation, or shading. Of interest in computer vision is the inverse problem of how object shape may be recovered from such an image. This is referred to as the shape-from-shading problem. When the imaging conditions are such that an overhead point-source illuminates a smooth Lambertian surface, the problem may be formulated mathematically as that of finding a solution to an eikonal equation. In this paper, we seek images for which there are no corresponding object shapes. We are therefore concerned with the nonexistence of (bounded) solutions to certain eikonal equations. Specifically, we give a necessary and sufficient condition for a circularly-symmetric eikonal equation to admit exclusively unbounded solutions. In addition, we give a sufficient condition for an eikonal equation to have no solution. Examples are presented that elucidate the significance of these results for computer vision. 1. Introduction. A monochrome photograph of a smooth object will typically exhibit brightness variation, or shading. Of interest to researchers in computer vision has been the inverse problem of how object shape may be extracted from image shading. This shape-from-shading problem has been shown by Horn ([5]; see also [8, pp. 123-172]) to correspond to that of solving a first-order partial differential equation.

A smooth object depicted in a monochrome image will often exhibit brightness variation, or shading. A problem much studied in computer vision has been that of how object shape may be recovered from image shading. When the imaging conditions are such that an overhead point-source illuminates a smooth Lambertian surface, the problem may be formulated as that of finding a solution to an eikonal equation. This article will focus on the existence and uniqueness of such solutions, reporting recent results obtained. With regard to existence, shading patterns are exhibited for which there is no corresponding object shape. Specifically, a necessary and sufficient condition is presented for a circularly symmetric eikonal equation to admit exclusively unbounded solutions; additionally, a sufficient condition is given for an eikonal equation to have no solution whatsoever. In connection with uniqueness, we consider eikonal equations, defined over a disc, such that the Euclidean norm of the gradient of any solution is circularly symmetric, vanishes exactly at the disc center, and diverges to infinity as the circumference of the disc is approached. Contrary to earlier influential work, a class of such eikonal equations is shown to possess simultaneously circularly symmetric and noncircularly symmetric bounded smooth solutions.

We describe two psychophysical experiments testing predictions of the square difference mechanism we have previously proposed for intensity-based stereo. Experiment 1 assesses the relative contributions of disparity and contrast to intensity-based stereo by measuring detection thresholds. The product of disparity and contrast at threshold is shown to be constant. In experiment 2, we measure quantitatively the global depth position perceived in stereograms of curved, smoothly shaded surfaces. The results show that disparity averaging over the surface involves a contrastdependent weighting function. The results from both experiments are consistent with predictions derived from the square difference mechanism. The relation of this mechanism to feature correspondence stereopsis and shape-from-shading is discussed and a general framework for assessing the modularity of stereopsis is presented.

In this paper, we propose a new optimization approach for the simultaneous computation of optical flow and edge detection therein. Instead of using an Ambrosio-Tortorelli type energy functional, we reformulate the optical flow problem as a multidimensional control problem. The optimal control problem is solved by discretization methods and large-scale optimization techniques. The edge detector can be immediately built from the control variables. We provide three series of numerical examples. The first shows that the mere presence of a gradient restriction has a regularizing effect, while the second demonstrates how to balance the regularizing effects of a term within the objective and the control restriction. The third series of numerical results is concerned with the direct evaluation of a TV-regularization term by introduction of control variables with sign restrictions.

This paper proposes a camera projection model for an oblique‐viewing endoscope, intended for application to an extended virtual reality system. The oblique‐viewing endoscope differs from the forward‐viewing endoscope by the fact that the view field can be shifted by rotating the scope cylinder. The projection model proposed in this study describes the systematic relation between the rotation angle of the scope cylinder and the external parameters of the camera model. A calibration method is also proposed for estimating the parameters of the projection model, and its accuracy is investigated. The oblique‐viewing projection model is derived as an extension of the general pinhole camera model by adding two axes of rotation as parameters and using the angle of rotation of the scope cylinder as the input. In the experiment, an abdominal endoscope is used as the real equipment. A rotary encoder with an angular resolution of 0.25° is used to input the rotation angle of the scope cylinder. ...

In this paper we propose a multi-scale Azimuthal Projection Distance Image scheme for analysing 3D texture and use it in classifying wrinkles and pores on high resolution facial normal maps. Azimuthal Projection Distance Images (APDI) have been widely used for mapping normals' orientation to scalar values on a tangent plane. The multiscale scheme we introduce computes APDIs at different resolutions of the normal map. We propose a geodesic-based method of scaling normal maps and modify the original formulation of the APDI which only take into consideration polar orientation. We finally apply the approach in classifying facial skin conditions (large pores, wrinkles). Local Binary Patterns feature are extracted from the multi-scale APDI pyramid and put into a Multi-Layer Perceptron. The results we achieve are promising even though the dataset used is limited.

Figure 1: First 3 principal components of our statistical diffuse (left) and specular (middle) albedo models. Both are visualised in linear sRGB space. Right: rendering of the combined model under frontal illumination in nonlinear sRGB space.

Abstract. This study employed the use of Cellular Neural Networks (CNN) for Edge Detection in images due to its high operational speed. The process of edge detection is unavoidable in many image processing tasks such as obstacle detection and satellite picture processing . ...

Topographic variations affect the reflectance properties of the Earth's surface and are often removed in remote sensing studies, especially when significant terrain variations exist. In this study, however, we show that shading effects assessed by Landsat can be treated as a signal that stores important topographic information, especially when the spectral characteristics of a surface are homogenous. The coastal transverse dunes of the Ashdod area, and the desert linear dunes of Nizzana (both located in Israel), were selected to investigate the abovementioned idea. The dune heights in these areas are 10 m on average (relative to their surroundings) and have maximum slopes of 33j. An innovative method for extracting slope, aspect, and height data for sand dunes using Landsat Thematic Mapper (TM) and Enhanced Thematic Mapper Plus (ETM+) images was developed, based on the regularity and periodicity of dunes' landscapes. Using two Landsat images representing different sun zenith and azimuth angles, reflectance values of each image were converted to cos(i) values (i = incident angle between the surface normal and the solar beam radiation), applying histogram matching methods. The slope and aspect of each pixel were determined as those that give the best prediction of the observed value of cos(i). Height profiles were then extracted, using simple trigonometric relationships. The accuracies of heights and slopes along selected profile lines were to the order of 1 m and 3j, respectively (at a spatial resolution of 15 m). Best results were obtained when the images included one from the summer and the other from the winter, corresponding to maximum difference in solar zenith and azimuth angles. Errors in heights were attributed to surface heterogeneity (e.g., presence of biogenic soil crusts in the rainy season), geometric correction errors, cast shadows, and Bidirectional Reflectance Distribution Function (BRDF) effects. Comparison to Advanced Thermal Emission and Reflection Radiometer (ASTER) 3D information showed that the proposed method is better in representing the topographic variation of the area than the digital elevation model (DEM) produced by ASTER.

We introduce a normal-based bas-relief generation and stylization method which is motivated by the recent advancement in this topic. Creating bas-relief from normal images has successfully facilitated bas-relief modeling in image space. However, the use of normal images in previous work is often restricted to certain type of operations only. This paper is intended to extend normal-based methods and construct bas-reliefs from normal images in a versatile way. Our method can not only generate a new normal image by combining various frequencies of existing normal images and details transferring, but also build bas-reliefs from a single RGB image and its edge-based sketch image. In addition, we introduce an auxiliary function to represent a smooth base surface and generate a layered global shape. To integrate above considerations into our framework, we formulate the bas- relief generation as a variational problem which can be solved by a screened Poisson equation. Some advantages of our...

Automatic bas-relief generation from 2D photographs potentially has applications to coinage, commemorative medals and souvenirs. However current methods are not yet ready for real use in industry due to insufficient artistic effect, noticeable distortion, and unbalanced contrast. We previously proposed a shape-from-shading (SFS) based method to automatically generate bas-reliefs from single frontal photographs of human faces; however, suppression of unwanted details remained a problem. Here, our experimental results show how incorporating non-photorealistic rendering (NPR) into our previous framework enables us to both suppress unwanted detail, and yet also emphasise important detail. We have also experimented with an alternative approach to recovering relief shape, using photometric stereo instead of SFS for surface orientation estimation. This can effectively reduce the computational time.

Univer sit é Cl er m o nt -F er r a nd II -B l a ise P a sc a l E c o l e Do c t o r a l e S c ienc es P o ur l'Ing énieur d e Cl er m o nt -F er r a nd T H E S E P ré se n té e p a r Renaud GAGLIONE P o u r o b te n ir le g ra d e d e Do c t eur d 'Univer sit é S p éc ia l it é : M a té ria u x e t c o m p o sa n ts p o u r l'é le c tro n iq u e Ele c tro n iq u e d 'ac q u isitio n d 'u n e g am m a-c am é ra S o u te n u e p u b liq u e m e n t le 3 n o v e m b re 2 0 0 5 d e v a n t le J u ry : M. Alain B a l d it P ré sid e nt M. P atrick l e Dû R ap p o rte u r M. B e rnard Il l e R ap p o rte u r M. L au re nt L uo ng E x am inate u r M. J ac q u e s L ec o q C o -d ire c te u r d e T h è se M. G é rard M o nt a r o u C o -d ire c te u r d e T h è se M. G é rard B o h ner Inv ité

Capturing detailed surface geometry currently requires specialized equipment such as laser range scanners, which despite their high accuracy, leave gaps in the surfaces that must be reconciled with photographic capture for relighting applications. Using only a standard digital camera and a single view, we present a method for recovering models of predominantly diffuse textured surfaces that can be plausibly relit and viewed from any angle under any illumination. Our multiscale shape-from-shading technique uses diffuse-lit/flash-lit image pairs to produce an albedo map and textured height field. Using two lighting conditions enables us to subtract one from the other to estimate albedo. In the absence of a flash-lit image of a surface for which we already have a similar exemplar pair, we approximate both albedo and diffuse shading images using histogram matching. Our depth estimation is based on local visibility. Unlike other depth-from-shading approaches, all operations are performed...

Complex reflectance phenomena such as specular reflections confound many vision problems since they produce image 'features' that do not correspond directly to intrinsic surface properties such as shape and spectral reflectance. A common approach to mitigate these effects is to explore functions of an image that are invariant to these photometric events. In this paper we describe a class of such invariants that result from exploiting color information in images of dichromatic surfaces. These invariants are derived from illuminant-dependent 'subspaces' of RGB color space, and they enable the application of Lambertian-based vision techniques to a broad class of specular, non-Lambertian scenes. Using implementations of recent algorithms taken from the literature, we demonstrate the practical utility of these invariants for a wide variety of applications, including stereo, shape from shading, photometric stereo, material-based segmentation, and motion estimation.

External morphology of egg of Eurygaster maura (LINNAEUS, 1758) was investigated by light (LM) and scanning electron microscope (SEM). The females were collected from Ankara, Ayaş, Asartepe and maintained under the laboratory conditions. Each female was shown to deposit generally 14 green eggs in a mass. The spherical shaped eggs were 1.04 mm in length and 0.96 mm in width. The first external evidence of embryonic development was the appearance of two red eye spots opposite each other beneath the aero-micropylar processes followed by appearance of a blackish T-shaped egg burster between the eye spots. Egg surface are covered with polygons. They have 20 to 22 aero-micropyles in a shape of truncated cones between polygons.

We discuss a statistical shape-from-shading framework for images of general and unknown illumination. To overcome arbitrary illumination, the framework makes use of the fact that general lighting can be expressed using low-order spherical harmonics for convex Lambertian objects. We cast the classical shape-from-shading equation as a Partial Least Squares (PLS) regression problem, which allows for the rapid computation of the solution. Results show accurate shape recovery with respect to ground truth data.

For the creation of a realistic 3 meter-sized relief globe of the Moon, a detailed height map of the entire lunar surface is required. Available height measurements of the Moon's surface are too coarse by a factor of 15 for this purpose. The only publicly available source of high-resolution information are photographic images from the Lunar Orbiter IV mission in 1967. We present a shape-from-shading approach to plausibly increase the resolution of existing low-resolution height data, based on a single high-resolution photographic mosaic image of the Moon. The presented reconstruction approach is designed to be robust with respect to frequent imperfections of the photographic imagery. Aside from the automatic reconstruction of a complete detailed lunar surface height map, we give a qualitative validation by the reconstruction of lunar surface details from close-up photographs of the Apollo 15 landing site.

This paper proposed a shape recovery approach from Endoscope Image using Eikonal Equation. Photometric constraint equation derived from the Lambert reflectance and geometrical constraint equation derived from the relationship between the neighboring points are used and these equations can make a new approximation equation of Eikonal equation under the point light source illumination and perspective projection. The original endoscope image is transformed and generated to the Lambertian image by removing the specular reflectance. Framework of Fast Marching Method using the derived Eikonal Equation can recover the 3D shape from endoscope image. Usefulness was confirmed using simulation and experiments.

In this paper a multiple classifier systems for face verification is proposed based on the study of scores fusion for four face authentication systems. Extraction features is realized by the Gabor wavelets phases, Principal Component Analysis (PCA) with the Enhanced Fisher linear discriminant Model (EFM) are used as a method of reducing data space. For the study of fusion of scores we used two approaches, the first based on the classification of scores using Fisher statistical method, Support Vector Machine (SVM) and artificial neural networks (MLP) and the second is based on combinations of scores by the weighted sum and fuzzy logic.

A smooth object depicted in a monochrome image will often exhibit brightness variation, or shading. Of interest in computer vision is the problem of how object shape may be recovered from such an image. When the imaging conditions are such that an overhead point-source illuminates a smooth Lambertian surface, the problem may be formulated as that of finding a solution to an eikonal equation. This article will focus on the existence and uniqueness of such solutions, reporting recent results obtained. With regard to existence, shading patterns are exhibited for which there is no corresponding object shape. Specifically, a necessary and sufficient condition is presented for a circularly-symmetric eikonal equation to admit exclusively unbounded solutions; additionally, a sufficient condition is given for an eikonal equation to have no solution. In connection with uniqueness, we consider eikonal equations, defined over a disc, such that the Euclidean norm of the gradient of any solution is circularly-symmetric, vanishes exactly at the disc centre, and diverges to infinity as the circumference of the disc is approached. Contrary to earlier expectations in the area, a class of such eikonal equations is shown to possess simultaneously circularly-symmetric and non-circularly-symmetric bounded smooth solutions.

Psychophysical studies have shown that human observers resolve shape-from-shading ambiguities by assuming that light is coming from above-left. Using event-related potentials (ERPs), we measured the processing time of the perception of an ambiguous shaded pattern. We found that the N2 component followed the change of perceived shape with stimulus orientation.We also found that the P1 component in occipital and temporal areas was corre-lated with the observers' idiosyncratic bias for light source position. The precocity of the correlated ERP components suggests that the light source is represented early in the visual system. Altogether, our results indicate that shape-from-shading is a mostly bottom-up mechanism. NeuroReport 14 :971^975 c 2003 Lippincott Williams & Wilkins.

It is a h clear that there are m a l dBerent approaches to the problem of how to integrate visual cuea. Let us list some of the obvious poasibilities: 1 There is no active integration of visual processes. Their individual outputs ate "integrated" at the stage at which they are used, for example by a navigation system. This is the approach advocated by Brooks [1987]. While it makes sense for automatic, insectlike, visuo-motor tasks such as tracking a target or avoiding obstacles (for example, the fly' s visuo-motor system Wchardt & Poggio 1976]), it seems quite unlikely for visual perception in the wider sense. 2 The visual modules ate so tightly coupled that it is impossible to consider visual modules as separate, even to a first order approximation. This view is unattractive on epistemological, engineering, and psychophysid grounds. 3 The visual modules are coupled to each other and to the image data in a parallel fashion--each pro-represented as an array coupled to the arrays associated with the other processes. This point of view is in the tradition of Marr's 2 i-D sketch, and especially of the "intrinsic images" of Barrow and Tenenbaum (19781. Our present scheme is of this type, and exploits the machinery of Markov Random Field (MRF) models. 4 Integration of different vision modalities is taking place in a taskdependent way at specific locations--not over the whole image-and when it is needed-therefore not at all times. This approach is suggested by wychoplu.sical data on visual attentiin and by the idea of visual routines [UIlman 19841 (see also , Mahoney 119871, and ~uithoff and Mallot 119881). We are presently exploring the third of these approaches. We believe that the last two approaches are compatible with each other. In particular, visual routines may operate on maps of discontinuities such as thwe delivered by the present Vision Machine, and therefore be h a t e d after a paraUel, automatic integration stage. In real life, of course, it may be more a matt e of coexistence. We believe, in fact, that a control structure baaed on specific knowledge about the properties of the various modules, the s p d c scene and the specific task will be needed in a later version of the V i o n Machine to overview and control the MRF integration stage itself and its parameters. It is poasible that the integration stage should he much more goal-directed than what o w present methods (MRF baaed) allow. The main goal of our work is to find out whether this is true. The V i n Machine project has a number of other goals. It provides a focus for developing parallel vision algorithms and for studying how to Chapter 42 The MIT Vision Machine

Standard regularization methods can be used to solve satisfactorily several problems in early vision, including edge detection, surface reconstruction, the computation of motion and the recovery of color. In this paper, we suggest (a) that the quadratic variational principles corresponding to standard regularization methods are equivalent to a linear regularizing operator acting on the data and (b) that this operator can be synthesized through associative learning. The synthesis of the regularizing operator involves the computation of the pseudoinverse of the data. The pseudoinverse can be computed by iterative methods, that can be implemented in analog networks. Possible implications for biological visual systems are also discussed.

Computer vision has developed algorithms for several early vision processes, such as edge detection, stereopsis, motion, texture, and color, which give separate cues as to the distance from the viewer of three-dimensional sur-'Ibmam Po@o k The S t. & important locations in a s m e. Third, we have argued that discontinuities are ideal for integrating idormation from different visual cues.

Dual-frequency Precipitation Radar (DPR) on the Global Precipitation Mission (GPM) core satellite started its precipitation measurement in March 2014, and the DPR standard products are available in public since September 2014. DPR consists of two radars, KuPR (13.6GHz) and KaPR (35.5GHz), and the DPR standard product consists of KuPR single-frequency product, KaPR single-frequency product, and dual-frequency product. Drop size distribution (DSD) and precipitation rates estimated by dual-frequency product are expected to be better than those by single-frequency product. However, the dual-frequency product has not been validated well. In this study, DSD and precipitation rate estimates by the dual-frequency method are compared with those in KuPR single-frequency product with surface reference technique. They are not agreed well partly because attenuation correction is not well applied for KaPR. For heavy precipitation, as attenuation correction is difficult, dual-frequency method may not be expected to work well, but the dual-frequency method should be applied mainly for weak and medium precipitation.

In this paper, we first discuss some general problems and issues concerning Shape from Shading (SfS) from Synthetic Aperture Radar (SAR) images and present the state of the art. Starting from this general discussion, we propose a new perspective and introduce a novel approach to SAR SfS based on fractals. In particular, fractal geometry is used to model the shape of natural surfaces, which are here of concern. The scattering mechanisms are described through solutions suitable for fractal surface models; in particular, the Small Perturbation Method (SPM) is used. Considering a simple SAR image model, the forward model is then inverted via an appropriate and extremely lowcomputational inversion approach and the underlying topography is estimated. The proposed SAR SfS technique is tested and numerically evaluated using an actual SAR image.

In this paper we show that the introduction of an attenuation factor in the brightness equations relative to various perspective Shape-from-Shading models allows to make the corresponding differential problems well-posed. We propose a unified approach based on the theory of viscosity solutions and we show that the brightness equations with the attenuation term admit a unique viscosity solution. We also discuss in detail the possible boundary conditions that we can use for the Hamilton-Jacobi equations associated to these models.

This paper deals with a possible application of computer vision techniques in the field of Cultural Heritage. These techniques allow an effective integration of data from different sources. Particular consideration will be given to an accurate geometric analysis of the zone under study in order to detect degradation damage in historical building-stone materials. To this end, we employ a computer vision technique, known as the Shape from Shading method, for which a photographic image of the stone material under consideration is the only input data required. By using this method, it is possible to recover, under some constraints, the shape of the three-dimensional surface of the object from the photographic image. In order to improve the results, in this paper, we implement this technique by using an optimization approach which allows a suitable integration of photographic and spatial data, the latter of which is obtained by a topographical device. We outline the potentialities of the method which mainly consist of two relevant capabilities. The first one is the geometric shape reconstruction of the surface material at a resolution much higher than the one allowed by topographical acquisition. The second one is the correction of the lightinginduced distortions in the photographic image. Such a correction is relevant for further image-based analysis of the degradation of the stone material. Experimental results, obtained by using both photographic and spatial data relative to a pudding stone with degradation, are presented and discussed.

: Wavelet bases are used to generate spaces of approximation for the resolution of bidimensional elliptic and parabolic problems. Under some specific hypotheses relating the properties of the wavelets to the order of the involved operators, it is shown that an approximate solution can be built. This approximation is then stable and converges towards the exact solution. It is designed such that fast algorithms involving biorthogonal multi resolution analyses can be used to resolve the corresponding numerical problems. Detailed algorithms are provided as well as the results of numerical tests on partial differential equations defined on the bidimensional torus.

We report on the reversal of asymmetry in visual-search tasks with shaded items. Previous studies have suggested that the target of a bottom-lit disk among distractors of top-lit disks is detected in a rapid and parallel manner, but not vice versa. However, in this study, we have shown that the compound items of top-lit disks were searched more quickly than those composed of bottom-lit disks where the items had to be segregated from their background. By modulating the inter-element distances, we confirmed that the reversal of search asymmetry cannot be due to the grouping of items. Further, we showed that the regions of the top-lit disks were perceived as figure more consistently than those of bottom-lit disks. The results indicate that the boundary assignment to the compound items of the top-lit disks enhances the segregation of search items from the background, and that the search mechanism may access the relatively higher representation that includes figure – ground relations.

We present a parallelized algorithm for accelerating the velocity extension calculations in a level-set method, which is essential for surface velocity based topography simulations, such as etching or deposition simulations for nanopatterning applications. The proposed algorithm improves the prevailing fast marching method by optimizing the heap data structure and efficiently reordering the calculations. We implemented the algorithm into Silvaco's Victory Process simulator, which is utilized for evaluating our algorithm with a three-dimensional simulation of an ion beam etching process used for spin-transfer torque magnetoresistive random access memory devices. Our results show a significant serial speed-up by a factor of at least 1.4 and a total speed-up by a factor of up to 8 using 8 threads for the velocity extension.

There are three reasons for illumination to vary within a scene. First, a light source may be visible from some surfaces but not from others. Second, because of linear perspective, the shape and size of a finite source may be different when viewed from different points in a scene. Third, the brightness of a source may be non-uniform. These variations are captured by a new computational model of spatially varying illumination. Two types of source are described: a distant hemispheric source such as the sky in which light converges onto a scene, and a proximal source such as a lamp in which fight diverges into a scene. Either type of source may have a non-uniform brightness function. We show how to render surfaces using this model, and how to compute shape from shading under it.

Our work focuses on the problem of automatic generation of viewpoint invariant geometric surface models from range data to be used in CAD/CAM systems. We developed two methods for recovering the surface. First is on the base of Shape from Shading method and second is on the base of Facet method. A comparison between the results obtained using the 2 methods is made. We also offer new techniques for the representation and CNC processing of the surfaces.

Given any two images taken under different illumination conditions, there always exist a physically realizable object which is consistent with both the images even if the lighting in each scene is constrained to be a known point light source at infinity [10]. In this paper, we show that images are much less ambiguous for the class of bilaterally symmetric Lambertian objects. In fact, the set of such objects can be partitioned into equivalence classes such that it is always possible to distinguish between two objects belonging to different equivalence classes using just one image per object. The conditions required for two objects to belong to the same equivalence class are very restrictive, thereby leading to the conclusion that images of symmetric objects are hardly ambiguous. The observation leads to an illumination-invariant matching algorithm to compare images of bilaterally symmetric Lambertian objects. Experiments on real data are performed to show the implications of the theoretical result even when the symmetry and Lambertian assumptions are not strictly satisfied. * Partially funded by an NSF-ITR Grant 03-25119 show that unlike general objects, it is almost always possible to distinguish between two bilaterally symmetric objects using just one image per object.

Object recognition systems relying on local descriptors are increasingly used because of their perceived robustness with respect to occlusions and to global geometrical deformations. Descriptors of this type-based on a set of oriented Gaussian derivative filters-are used in our recognition system. In this paper, we explore a multiview 3D object recognition system that does not use explicit geometrical information. The basic idea is to find discriminant features to describe an object across different views. A boosting procedure is used to select features out of a large feature pool of local features collected from the positive training examples. We describe experiments on face images with excellent recognition rate.

In this paper, we propose an original framework for three dimensional face representation and similarity matching. Basic traits of a face are encoded by extracting convex and concave regions from the surface of a face model. A compact graph representation is then constructed from these regions through an original modeling technique capable to quantitatively measure spatial relationships between regions in a three dimensional space and to encode this information in an attributed relational graph. In this way, the structural similarity between two face models is evaluated by matching their corresponding graphs. Experimental results on a 3D face database show that the proposed solution attains high retrieval accuracy and is reasonably robust to facial expression and pose changes.