Because of its versatility, speed and robustness, shadow mapping has always been a popular algorithm for fast hard shadow generation since its introduction in 1978, first for offline film productions and later increasingly so in real-time graphics. So it is not surprising that recent years have seen an explosion in the number of shadow map related publications. Because of the abundance of articles on the topic, it has become very hard for practitioners and researchers to select a suitable shadow algorithm, and therefore many applications miss out on the latest high-quality shadow generation approaches. The goal of this survey is to rectify this situation by providing a detailed overview of this field. We show a detailed analysis of shadow mapping errors and derive a comprehensive classification of the existing methods. We discuss the most influential algorithms, consider their benefits and shortcomings and thereby provide the readers with the means to choose the shadow algorithm best suited to their needs.

/pdf/a-survey-of-real-time-hard-shadow-mapping-methods-2rnlftcorg.pdf

A Survey of Real-Time Hard Shadow Mapping Methods

Shadow Algorithms Data Miner

Shadows are crucial for enhancing realism and provide important visual cues. In recent years, many important contributions have been made both for hard shadows and soft shadows. Often spurred by the tremendous increase of computational power and capabilities of graphics hardware, much progress has been made concerning visual quality and speed, making high-quality real-time shadows a reachable goal. But with the growing wealth of available choices, it is particularly dicult to pick the right solution and assess shortcomings. Because currently there is no ultimate approach available, algorithms should be selected in accordance to the context in which shadows are produced. The possibilities range across a wide spectrum; from very approximate but really ecient to slower but accurate, adapted only to smaller or only to larger sources, addressing directional lights or positional lights, or involving GPU- or CPU-heavy computations. This course tries to serve as a guide to better understand limitations and failure cases, advantages and disadvantages, and suitability of the algorithms for dierent application scenarios. We will focus on real-time to interactive solutions but also discuss oine approaches if needed for a better understanding.In the future, we will provide you with more information via the course webpage: http://www.mpi-inf.mpg.de/resources/ShadowCourse/.

Casting Shadows in Real Time

Ambient occlusion (AO) provides an effective approximation to global illumination that enjoys widespread use amongst practitioners. In this paper, we present a fast easy-to-implement separable approximation to screen space ambient occlusion. Computing occlusion first along a single direction and then transporting this occlusion into a second pass that is stochastically evaluating the final shading based on the AO estimates proves extremely efficient. Combined with interleaved sampling and geometry-aware blur, visually convincing results close to a non-separable occlusion can be obtained at much higher performance.

/pdf/separable-approximation-of-ambient-occlusion-469p7vkal1.pdf

Separable Approximation of Ambient Occlusion.

Casting shadows in real time

Original bump mapping is only defined for surfaces with a known surface parametrization. In this paper a new method, for the GPU, is presented which does not use such a given parametrization. To compute the perturbed normal, the only inputs used are the surface position, the height value and the original normal. The method decouples bump mapping from the primitive type, thus allowing for a higher degree of proceduralism in generation of both the height value and the surface.

Bump Mapping Unparametrized Surfaces on the GPU

This paper proposes a real-time solution that is a hybrid between standard shadow mapping and any of the current methods based on perspective shadow mapping. Existing methods are known to have aliasing problems. At little extra cost, we combine two methods to minimize aliasing errors. This is done by computing a plane equation that separates the two resulting shadow maps to get the best possible area-based sampling density. An additional property of this separation plane is that objects entirely on one side of the plane only need to be rendered into one map.

/pdf/separating-plane-perspective-shadow-mapping-409si4zift.pdf

Morten S. Mikkelsen

Papers

Bump Mapping Unparametrized Surfaces on the GPU

Separating-Plane Perspective Shadow Mapping