There currently exist several techniques for selecting and combining images from a digital image library into a single image so that the result meets certain prespecified visual criteria. Image mosaic methods, first explored by Connors and Trivedi[18], arrange library images according to some tiling arrangement, often a regular grid, so that the combination of images, when viewed as a whole, resembles some input target image. Other techniques, such as Autocollage of Rother et al.[78], seek only to combine images in an interesting and visually pleasing manner, according to certain composition principles, without attempting to approximate any target image. Each of these techniques provide a myriad of creative options for artists who wish to combine several levels of meaning into a single image or who wish to exploit the meaning and symbolism contained in each of a large set of images through an efficient and easy process. We first examine the most notable and successful of these methods, and summarize the advantages and limitations of each. We then formulate a set of goals for an image collage system that combines the advantages of these methods while addressing and mitigating the drawbacks. Particularly, we propose a system for creating photocollages that approximate a target image as an aggregation of smaller images, chosen from a large library, so that interesting visual correspondences between images are exploited. In this way, we allow users to create collages in which multiple layers of meaning are encoded, with meaningful visual links between each layer. In service of this goal, we ensure that the images used are as large as possible and are combined in such a way that boundaries between images are not immediately apparent, as in Autocollage. This has required us to apply a multiscale approach to searching and comparing images from a large database, which achieves both speed and accuracy. We also propose a new framework for color post-processing, and propose novel techniques for decomposing images according to object and texture information.

Dendritic or branching structures are commonly seen in natural phenomena such as lightning, cracking and vegetal growth. They are also often used for artistic or decorative purposes, ranging from ornamentation to decorative ceramics. Existing procedural methods for modeling these structures remain very limited in terms of control and flexibility. As a result, these objects tend to be modeled individually, which is a painstaking and costly process.<p>We present a new procedural method for modeling dendritic structures based on a path planning approach. Our method includes the implementation of a partial non-scalar distance metric that gives us effective and flexible control handles over the evolving dendritic structure. These control handles are demonstrated by guiding the growth of dendritic structures using input images, allowing us to create a form of stylistic dendritic halftoning and to embed hidden images in dendritic trees to create pareidolia effects. These applications demonstrate the vast diversity of structures that can easily be modeled by our process a flexibility that existing methods definitely lack. We also demonstrate the application of the partial non-scalar distance metric to the context of texture synthesis from example, and show how it holds promise for many other contexts.

Dendritic or branching structures are commonly seen in natural phenomena such as lightning, cracking and vegetal growth. They are also often used for artistic or decorative purposes, ranging from ornamentation to decorative ceramics. Existing procedural methods for modeling these structures remain very limited in terms of control and flexibility. As a result, these objects tend to be modeled individually, which is a painstaking and costly process.<p>We present a new procedural method for modeling dendritic structures based on a path planning approach. Our method includes the implementation of a partial non-scalar distance metric that gives us effective and flexible control handles over the evolving dendritic structure. These control handles are demonstrated by guiding the growth of dendritic structures using input images, allowing us to create a form of stylistic dendritic halftoning and to embed hidden images in dendritic trees to create pareidolia effects. These applications demonstrate the vast diversity of structures that can easily be modeled by our process a flexibility that existing methods definitely lack. We also demonstrate the application of the partial non-scalar distance metric to the context of texture synthesis from example, and show how it holds promise for many other contexts.

There currently exist several techniques for selecting and combining images from a digital image library into a single image so that the result meets certain prespecified visual criteria. Image mosaic methods, first explored by Connors and Trivedi[18], arrange library images according to some tiling arrangement, often a regular grid, so that the combination of images, when viewed as a whole, resembles some input target image. Other techniques, such as Autocollage of Rother et al.[78], seek only to combine images in an interesting and visually pleasing manner, according to certain composition principles, without attempting to approximate any target image. Each of these techniques provide a myriad of creative options for artists who wish to combine several levels of meaning into a single image or who wish to exploit the meaning and symbolism contained in each of a large set of images through an efficient and easy process. We first examine the most notable and successful of these methods, and summarize the advantages and limitations of each. We then formulate a set of goals for an image collage system that combines the advantages of these methods while addressing and mitigating the drawbacks. Particularly, we propose a system for creating photocollages that approximate a target image as an aggregation of smaller images, chosen from a large library, so that interesting visual correspondences between images are exploited. In this way, we allow users to create collages in which multiple layers of meaning are encoded, with meaningful visual links between each layer. In service of this goal, we ensure that the images used are as large as possible and are combined in such a way that boundaries between images are not immediately apparent, as in Autocollage. This has required us to apply a multiscale approach to searching and comparing images from a large database, which achieves both speed and accuracy. We also propose a new framework for color post-processing, and propose novel techniques for decomposing images according to object and texture information.

In this dissertation I examine how research in non-photorealistic rendering, human perception, and game-based learning can be combined to produce illustrative simulations of different visual systems that effectively convey information about vision to unprimed observers. The Visual Differences Simulation (VDS) methodology and framework that I propose is capable of producing simulations of animal visual systems based on how they relate to human vision, and can represent differences in color vision, hyperspectral sensitivity, visual acuity, light sensitivity, field of view, motion sensitivity, and eye construction. The simulations produced by the VDS framework run in real time, allowing users to explore computer-generated environments from `behind the eyes' of an animal in an interactive and immersive manner. I also examine how cognitive principles and game-based learning can be leveraged to demonstrate and enhance the educational impact of the simulations produced by the VDS framework. Two case studies are presented, where simulations of the cat and the bee visual systems are used as the basis to design educational games, and are evaluated to show that embedding the simulations in educational games is an effective and engaging way to convey information about vision to unprimed observers. / Graduate

The rapid proliferation of 3D volume data, including MRI and CT scans, is prompting the search within computer graphics for more effective volume visualisation techniques. Partially because of the traditional association with medical subjects, concepts borrowed from the domain of scientific illustration show great promise for enriching volume visualisation. This thesis describes the first general system dedicated to creating user-directed, variable-detail, scientific illustrations directly from volume data. In particular, using volume segmentation for explicit abstraction in non-photorealistic volume renderings is a new concept. The unique challenges and opportunities of volume data require rethinking many non-photorealistic algorithms that traditionally operate on polygonal meshes. The resulting 2D images are qualitatively different from but complementary to those normally seen in computer graphics, and inspire an analysis of the various artistic implications of volume models for scientific illustration.

Modeling of natural phenomena has been of particular interest in the graphics ommunity in recent years. This thesis will explore a method for creating and animating orb webs using a coupled spring-mass system. Using a spring-mass system for creating the orb web is ideal as we can represent each web strand using coupled spring-mass pairs. This allows the orb web simulator to be physically based, i.e., the simulation follows the laws that act on objects in the real world. This in turn simplifies the process of animating the web, as the animation emerges from the simulator without anyone having to set it up explicitly. Since this model is physically based, it would allow for realistic visualization of effects such as observing an orb web under a wind. In the children's book ``Charlotte's Web', the spider creates orb webs with words inscribed on them. Charlotte's web is used as an inspiration, in this thesis, to create webs which no real world spider could possibly create, while keeping the model physically based. This involves modifying the orb web such that the target text shows up on the orb web while keeping the web looking as natural as possible.

Black-and-white imagery is a popular and interesting depiction technique in the visual arts, in which varying tints and shades of a single colour are used. Within the realm of black-and-white images, there is a set of black-and-white illustrations that only depict salient features by ignoring details, and reduce colour to pure black and white, with no intermediate tones. These illustrations hold tremendous potential to enrich decoration, human communication and entertainment. Producing abstract black-and-white illustrations by hand relies on a time consuming and difficult process that requires both artistic talent and technical expertise. Previous work has not explored this style of illustration in much depth, and simple approaches such as thresholding are insufficient for stylization and artistic control. I use the word wholetoning to refer to illustrations that feature a high degree of shape and tone abstraction. In this thesis, I explore computer algorithms for generating wholetoned illustrations. First, I offer a general-purpose framework, “artistic thresholding”, to control the generation of wholetoned illustrations in an intuitive way. The basic artistic thresholding algorithm is an optimization framework based on simulated annealing to get the final bi-level result. I design an extensible objective function from our observations of a lot of wholetoned images. The objective function is a weighted sum over terms that encode features common to wholetoned illustrations. Based on the framework, I then explore two specific wholetoned styles: papercutting and representational calligraphy. I define a paper-cut design as a wholetoned image with connectivity constraints that ensure that it can be cut out from only one piece of paper. My computer generated papercutting technique can convert an original wholetoned image into a paper-cut design. It can also synthesize stylized and geometric patterns often found in traditional designs. Representational calligraphy is defined as a wholetoned image with the constraint that all depiction elements must be letters. The procedure of generating representational calligraphy designs is formalized as a “calligraphic packing” problem. I provide a semi-automatic technique that can warp a sequence of letters to fit a shape while preserving their readability.

The rapid proliferation of 3D volume data, including MRI and CT scans, is prompting the search within computer graphics for more effective volume visualisation techniques. Partially because of the traditional association with medical subjects, concepts borrowed from the domain of scientific illustration show great promise for enriching volume visualisation. This thesis describes the first general system dedicated to creating user-directed, variable-detail, scientific illustrations directly from volume data. In particular, using volume segmentation for explicit abstraction in non-photorealistic volume renderings is a new concept. The unique challenges and opportunities of volume data require rethinking many non-photorealistic algorithms that traditionally operate on polygonal meshes. The resulting 2D images are qualitatively different from but complementary to those normally seen in computer graphics, and inspire an analysis of the various artistic implications of volume models for scientific illustration.

Black-and-white imagery is a popular and interesting depiction technique in the visual arts, in which varying tints and shades of a single colour are used. Within the realm of black-and-white images, there is a set of black-and-white illustrations that only depict salient features by ignoring details, and reduce colour to pure black and white, with no intermediate tones. These illustrations hold tremendous potential to enrich decoration, human communication and entertainment. Producing abstract black-and-white illustrations by hand relies on a time consuming and difficult process that requires both artistic talent and technical expertise. Previous work has not explored this style of illustration in much depth, and simple approaches such as thresholding are insufficient for stylization and artistic control. I use the word wholetoning to refer to illustrations that feature a high degree of shape and tone abstraction. In this thesis, I explore computer algorithms for generating wholetoned illustrations. First, I offer a general-purpose framework, “artistic thresholding”, to control the generation of wholetoned illustrations in an intuitive way. The basic artistic thresholding algorithm is an optimization framework based on simulated annealing to get the final bi-level result. I design an extensible objective function from our observations of a lot of wholetoned images. The objective function is a weighted sum over terms that encode features common to wholetoned illustrations. Based on the framework, I then explore two specific wholetoned styles: papercutting and representational calligraphy. I define a paper-cut design as a wholetoned image with connectivity constraints that ensure that it can be cut out from only one piece of paper. My computer generated papercutting technique can convert an original wholetoned image into a paper-cut design. It can also synthesize stylized and geometric patterns often found in traditional designs. Representational calligraphy is defined as a wholetoned image with the constraint that all depiction elements must be letters. The procedure of generating representational calligraphy designs is formalized as a “calligraphic packing” problem. I provide a semi-automatic technique that can warp a sequence of letters to fit a shape while preserving their readability.