Kevin Glass

Duplicating road patterns in south african informal settlements using procedural techniques

The formation of informal settlements in and around urban complexes has largely been ignored in the context of procedural city modeling. However, many cities in South Africa and globally can attest to the presence of such settlements. This paper analyses the phenomenon of informal settlements from a procedural modeling perspective. Aerial photography from two South African urban complexes, namely Johannesburg and Cape Town is used as a basis for the extraction of various features that distinguish different types of settlements. In particular, the road patterns which have formed within such settlements are analysed, and various procedural techniques proposed (including Voronoi diagrams, subdivision and L-systems) to replicate the identified features. A qualitative assessment of the procedural techniques is provided, and the most suitable combination of techniques identified for unstructured and structured settlements. In particular it is found that a combination of Voronoi diagrams and subdivision provides the closest match to unstructured informal settlements. A combination of L-systems, Voronoi diagrams and subdivision is found to produce the closest pattern to a structured informal settlement.

Minimum spanning trees for valley and ridge characterization in digital elevation maps

Texture synthesis employs neighbourhood matching to generate appropriate new content. Terrain synthesis has the added constraint that new content must be geographically plausible. The profile recognition and polygon breaking algorithm (PPA) [Chang et al. 1998] provides a robust mechanism for characterizing terrain as systems of valley and ridge lines in digital elevation maps. We exploit this to create a terrain characterization metric that is robust, efficient to compute and is sensitive to terrain properties. Terrain regions are characterized as a minimum spanning tree derived from a graph created from the sample points of the elevation map which are encoded as weights in the edges of the graph. This formulation allows us to provide a single consistent feature definition that is sensitive to the pattern of ridges and valleys in the terrain Alternative formulations of these weights provide richer characteristic measures and we provide examples of alternate definitions based on curvature and contour measures. We show that the measure is robust, with a significant portion derived directly from information local to the terrain sample. Global terrain characteristics introduce the issue of over- and under-connected valley/ridge lines when working with sub-regions. This is addressed by providing two graph construction strategies, which respectively provide an upper bound on connectivity as a single spanning tree, and a lower bound as a forest of trees. Efficient minimum spanning tree algorithms are adapted to the context of terrain data and are shown to provide substantially better performance than previous PPA implementations. In particular, these are able to characterize valley and ridge behaviour at every point even in large elevation maps, providing a measure sensitive to terrain features at all scales. The resulting graph based formulation provides an efficient and elegant algorithm for characterizing terrain features. The measure can be calculated efficiently, is robust under changes of neighbourhood position, size and resolution and the hybrid measure is sensitive to terrain features both locally and globally.

Hierarchical rule generalisation for speaker identification in fiction books

This paper presents a hierarchical pattern matching and generalisation technique which is applied to the problem of locating the correct speaker of quoted speech found in fiction books. Patterns from a training set are generalised to create a small number of rules, which can be used to identify items of interest within the text. The pattern matching technique is applied to finding the Speech-Verb, Actor and Speaker of quotes found in fiction books. The technique performs well over the training data, resulting in rule-sets many times smaller than the training set, but providing very high accuracy. While the rule-set generalised from one book is less effective when applied to different books than an approach based on hand coded heuristics, performance is comparable when testing on data closely related to the training set.

Simulating crowd phenomena in African markets

Crowd simulation is an important feature in the computer graphics field. Typical implementations simulate battle scenes, emergency situations, safety issues or add content to virtual environments. The problem stated in this paper falls in the last category. We present a crowd simulation behavioural model which allows us to simulate identified phenomena in popular local African markets such as narrow street flows and crowd formation around street performances. We propose a three-tier architecture model enable to produce intentions, perform path planning and control movement. We demonstrate that this approach produces the desired behaviour associated with crowds in an African market, which includes navigation, flow formation and circle creation.

Constraint-based conversion of fiction text to a time-based graphical representation

This paper presents a method for converting unrestricted fiction text into a time-based graphical form. Key concepts extracted from the text are used to formulate constraints describing the interaction of entities in a scene. The solution of these constraints over their respective time intervals provides the trajectories for these entities in a graphical representation. Three types of entity are extracted from fiction books to describe the scene, namely Avatars, Areas and Objects. We present a novel method for modeling the temporal aspect of a fiction story using multiple time-line representations after which the information extracted regarding entities and time-lines is used to formulate constraints. A constraint solving technique based on interval arithmetic is used to ensure that the behaviour of the entities satisfies the constraints over multiple universally quantified time intervals. This approach is demonstrated by finding solutions to multiple time-based constraints, and represents a new contribution to the field of Text-to-Scene conversion. An example of the automatically produced graphical output is provided in support of our constraint-based conversion scheme.

Mechanisms for multimodality: taking fiction to another dimension

We present methods for automatically constructing representations of fiction books in a range of modalities: audibly, graphically and as 3D virtual environments. The correspondence between the sequential ordering of events against the order of events presented in the text is used to correctly resolve the dynamic interactions for each representation. Synthesised audio created from the fiction text is used to calibrate the base time-line against which the other forms of media are correctly aligned. The audio stream is based on speech synthesis using the text of the book, and is enhanced using distinct voices for the different characters in a book. Sound effects are included automatically. The graphical representation represents the text (as subtitles), identifies active characters and provides visual feedback of the content of the story. Dynamic virtual environments conform to the constraints implied by the story, and are used as a source of further visual content. These representations are all aligned to a common time-line, and combined using sequencing facilities to provide a multimodal version of the original text.