Humans perceive the three-dimensional structure of the world with apparent ease. However, despite all of the recent advances in computer vision research, the dream of having a computer interpret an image at the same level as a two-year old remains elusive. Why is computer vision such a challenging problem and what is the current state of the art? Computer Vision: Algorithms and Applications explores the variety of techniques commonly used to analyze and interpret images. It also describes challenging real-world applications where vision is being successfully used, both for specialized applications such as medical imaging, and for fun, consumer-level tasks such as image editing and stitching, which students can apply to their own personal photos and videos. More than just a source of recipes, this exceptionally authoritative and comprehensive textbook/reference also takes a scientific approach to basic vision problems, formulating physical models of the imaging process before inverting them to produce descriptions of a scene. These problems are also analyzed using statistical models and solved using rigorous engineering techniques Topics and features: structured to support active curricula and project-oriented courses, with tips in the Introduction for using the book in a variety of customized courses; presents exercises at the end of each chapter with a heavy emphasis on testing algorithms and containing numerous suggestions for small mid-term projects; provides additional material and more detailed mathematical topics in the Appendices, which cover linear algebra, numerical techniques, and Bayesian estimation theory; suggests additional reading at the end of each chapter, including the latest research in each sub-field, in addition to a full Bibliography at the end of the book; supplies supplementary course material for students at the associated website, http://szeliski.org/Book/. Suitable for an upper-level undergraduate or graduate-level course in computer science or engineering, this textbook focuses on basic techniques that work under real-world conditions and encourages students to push their creative boundaries. Its design and exposition also make it eminently suitable as a unique reference to the fundamental techniques and current research literature in computer vision.

/pdf/computer-vision-algorithms-and-applications-25dn6wu83j.pdf

Computer Vision: Algorithms and Applications

This thesis describes a general method for automatic reconstruction of accurate, concise, piecewise smooth surfaces from unorganized 3D points. Instances of surface reconstruction arise in numerous scientific and engineering applications, including reverse-engineering--the automatic generation of CAD models from physical objects.
Previous surface reconstruction methods have typically required additional knowledge, such as structure in the data, known surface genus, or orientation information. In contrast, the method outlined in this thesis requires only the 3D coordinates of the data points. From the data, the method is able to automatically infer the topological type of the surface, its geometry, and the presence and location of features such as boundaries, creases, and corners.
The reconstruction method has three major phases: (1) initial surface estimation, (2) mesh optimization, and (3) piecewise smooth surface optimization. A key ingredient in phase 3, and another principal contribution of this thesis, is the introduction of a new class of piecewise smooth representations based on subdivision. The effectiveness of the three-phase reconstruction method is demonstrated on a number of examples using both simulated and real data.
Phases 2 and 3 of the surface reconstruction method can also be used to approximate existing surface models. By casting surface approximation as a global optimization problem with an energy function that directly measures deviation of the approximation from the original surface, models are obtained that exhibit excellent accuracy to conciseness trade-offs. Examples of piecewise linear and piecewise smooth approximations are generated for various surfaces, including meshes, NURBS surfaces, CSG models, and implicit surfaces.

/pdf/surface-reconstruction-from-unorganized-points-4377nlb52v.pdf

Surface reconstruction from unorganized points

We present a method of recovering high dynamic range radiance maps from photographs taken with conventional imaging equipment. In our method, multiple photographs of the scene are taken with different amounts of exposure. Our algorithm uses these differently exposed photographs to recover the response function of the imaging process, up to factor of scale, using the assumption of reciprocity. With the known response function, the algorithm can fuse the multiple photographs into a single, high dynamic range radiance map whose pixel values are proportional to the true radiance values in the scene. We demonstrate our method on images acquired with both photochemical and digital imaging processes. We discuss how this work is applicable in many areas of computer graphics involving digitized photographs, including image-based modeling, image compositing, and image processing. Lastly, we demonstrate a few applications of having high dynamic range radiance maps, such as synthesizing realistic motion blur and simulating the response of the human visual system.

/pdf/recovering-high-dynamic-range-radiance-maps-from-photographs-zk2yr4fcgd.pdf

Recovering high dynamic range radiance maps from photographs

We present a new technique for the display of high-dynamic-range images, which reduces the contrast while preserving detail. It is based on a two-scale decomposition of the image into a base layer,...

/pdf/fast-bilateral-filtering-for-the-display-of-high-dynamic-43a1vkjsrr.pdf

Fast bilateral filtering for the display of high-dynamic-range images

We present a new technique for the display of high-dynamic-range images, which reduces the contrast while preserving detail. It is based on a two-scale decomposition of the image into a base layer, encoding large-scale variations, and a detail layer. Only the base layer has its contrast reduced, thereby preserving detail. The base layer is obtained using an edge-preserving filter called the bilateral filter. This is a non-linear filter, where the weight of each pixel is computed using a Gaussian in the spatial domain multiplied by an influence function in the intensity domain that decreases the weight of pixels with large intensity differences. We express bilateral filtering in the framework of robust statistics and show how it relates to anisotropic diffusion. We then accelerate bilateral filtering by using a piecewise-linear approximation in the intensity domain and appropriate subsampling. This results in a speed-up of two orders of magnitude. The method is fast and requires no parameter setting.

/pdf/fast-bilateral-filtering-for-the-display-of-high-dynamic-fprymiweg7.pdf

We propose a complete image-based process that facilitates recovery of both gross-scale geometry and local surface structure to create highly detailed 3D models of building facades from photographs. We approximate both albedo and sufficient local geometric structure to compute complex self-shadowing effects, and fuse this with a gross-scale 3D model. Our approach yields a perceptually high-quality model, imparting the illusion of measured reflectance. The requirements of our approach are that image capture must be performed under diffuse lighting and surfaces in the images must be predominantly Lambertian. Exemplars of materials are obtained through surface depth hallucination, and our novel method matches these with multi-view image sequences that are also used to automatically recover 3D geometry. In this paper we illustrate the effictiveness of our approach through visually realistic reconstructions of historic buildings in two test cases, together with a table showing the breakdown of effort required to reconstruct each.

Relightable Buildings from Images

Global illumination researchers tend to think in terms of mesh density and sampling frequency, and their software reflects this in its user interface. Advanced rendering systems are rife with long command lines and parameters for tuning the sample densities, thresholds and other algorithm-specific variables, and the novice user is quickly lost in a sea of possibilities. This paper details a successful effort of making one such global illumination system usable by people who understand their problems, even if they do not understand the methods needed to solve them, through an assisted oracle approach. A single program is introduced to map a small set of intuitive control variables to the rendering commands and parameter settings needed to produce the desired output in a reasonable time. This new executive program then serves as the basis for a graphical user interface that is both friendly in its appearance and reliable in its performance. Finally, we conclude with some future directions for improving this interface.

/pdf/making-global-illumination-user-friendly-2tl4wb9yyp.pdf

Making global illumination user-friendly

Using daylighting in commercial buildings may significantly reduce electric lighting requirements if appropriate photoelectric controls are used to adjust the electric lighting output according to the available daylight Prior analysis and results from monitored buildings and scale-model measurements suggest that the selection, placement, and installation of the control photosensor is a difficult task, even with simple non-operable fenestration systems, since the daylight contributions from sun, sky, and ground change continuously The problem becomes even more complex for fenestration systems that incorporate operable shading devices, because every adjustment changes the system's optical properties This paper presents results from measurements in a scale model under real skies, designed to better understand the problem of integrating fenestration and lighting controls The scale model represented a typical office space and was equipped with motorized venetian blinds Three control photosensors mounted on the ceiling were considered for the operation of the electric lighting system, and two control strategies were considered for the operation of the venetian blinds Two ground-plane reflectances and two window orientations were examined Results indicate that the signal from a ceiling-mounted control photosensor shielded from direct light from the window shows the best correlation with daylight work-plane illuminance, regardless of ground planemore » reflectance or venetian blind slat angle for all slat angles that do not allow penetration of direct solar radiation Results also indicate that the control strategies of the venetian blinds that were considered for the purposes of this study may result in significantly different slat angles and thus different daylighting work-plane illuminances and electric lighting requirements, especially when the ground-plane reflectance is high 3 refs, 31 figs, 1 tab« less

/pdf/the-integration-of-operable-shading-systems-and-electric-29f9u1wl73.pdf

The integration of operable shading systems and electric lighting controls

Ce projet a pour but d'elaborer des outils informatiques permettant de concevoir, d'etudier et d'evaluer des systemes de gestion de lumiere naturelle. Il s'inscrit dans le cadre du programme de recherche interdisciplinaire LUMEN, visant a une utilisaiton optimale de la lumiere naturelle dans les bâtiments afin de reduire la consommation d'energie electrique de ces derniers.

Outils informatiques en lumière naturelle

The� visual� properties� of� surfaces� reveal� many� things� including� a� floor�适s� cleanliness� and� a� car�适s� age.� These� judgments� of� material� are� based� on� the� spread� of� light� reflected� from� a� surface.� The� bidirectional� reflectance� distribution� function� (BRDF)� quantifies� the� pattern� of� spread� and� how� it� depends� on� the� direction� of� incident� light,� surface� shape,� and� surface� material.� Two� extremes� are� Lambertian� and� mirrored� surfaces,� which� respectively� have� uniform� and� delta�괂Ġ function� BRDFs.� Most� surfaces� have� more� complicated� BRDFs� and� we� examined� many� of� them� using� the� Ward� model� as� an� approximation� for� real� surfaces.� Reflections� are� generally� view� dependent.� This� dependence� creates� a� difference� between� the� binocular� disparities� of� a� reflection� and� the� surface� itself.� It� also� creates� focus� differences� between� the� reflection� and� physical� surface.� In� simulations� we� examined� how� material� type� affects� retinal� images.� We� calculated� point�괂Ġspread� functions� (PSFs)� for� reflections� off� different� materials� as� a� function� of� the� eye�适s� focus� state.� When� surface� roughness� is� zero,� the� reflection� PSF� changes� dramatically� with� focus� state.� With� greater� roughness,� the� PSF� change� is� reduced� until� there� is� no� effect� of� focus� state� with� sufficiently� rough� surfaces.� The� reflection� PSF� also� has� a� dramatic� effect� on� the� ability� to� estimate� disparity.� We� next� examined� people�适s� ability� to� distinguish� surface� markings� from� reflections� and� to� identify� different� types� of� material.� We� used� a� unique� volumetric� display� that� allows� us� to� present� nearly� correct� focus� cues� along� with� more� traditional� depth� cues� such� as� disparity.� With� binocular� viewing,� we� observed� a� clear� effect� of� the� disparity� of� reflections� on� these� judgments.� We� also� found� that� disparity� provided� less� useful� information� with� rougher� materials.� With� monocular� viewing,� we� observed� a� small� but� consistent� effect� of� the� reflection�适s� focal� distance� on� judgments� of� markings� vs.� reflections� and� on� identification� of� material.�

/pdf/the-perception-of-surface-material-from-disparity-and-focus-2bbcf9rdvg.pdf

Gregory J. Ward

Papers

Relightable Buildings from Images

Making global illumination user-friendly

The integration of operable shading systems and electric lighting controls

Outils informatiques en lumière naturelle

The Perception of Surface Material from Disparity and Focus Cues