Rendering of light shaft and shadow for indoor environments enhancing technique

Abstract

The ray marching methods have become the most attractive method to provide realism in rendering the effects of light scattering in the participating media of numerous applications. This has attracted significant attention from the scientific community. Up-sampling of ray marching methods is suitable to evaluate light scattering effects such as volumetric shadows and light shafts for rendering realistic scenes, but suffers of cost a lot for rendering. Therefore, some encouraging outcomes have been achieved by using down-sampling of ray marching approach to accelerate rendered scenes. However, these methods are inherently prone to artifacts, aliasing and incorrect boundaries due to the reduced number of sample points along view rays. This study proposed a new enhancing technique to render light shafts and shadows taking into consideration the integration light shafts, volumetric shadows, and shadows for indoor environments. This research has three major phases that cover species of the effects addressed in this thesis. The first phase includes the soft volumetric shadows creation technique called Soft Bilateral Filtering Volumetric Shadows (SoftBiF-VS). The soft shadow was created using a new algorithm called Soft Bilateral Filtering Shadow (SBFS). This technique was started by developing an algorithm called Imperfect Multi-View Soft Shadows (IMVSSs) based on down-sampling multiple point lights (DMPLs) and multiple depth maps, which are processed by using bilateral filtering to obtain soft shadows. Then, down-sampling light scattering model was used with (SBFS) to create volumetric shadows, which was improved using cross-bilateral filter to get soft volumetric shadows. In the second phase, soft light shaft was generated using a new technique called Realistic Real-Time Soft Bilateral Filtering Light Shafts (realTiSoftLS). This technique computed the light shaft depending on down-sampling volumetric light model and depth test, and was interpolated by bilateral filtering to gain soft light shafts. Finally, an enhancing technique for integrating all of these effects that represent the third phase of this research was achieved. The performance of the new enhanced technique was evaluated quantitatively and qualitatively a measured using standard dataset. Results from the experiment showed that 63% of the participants gave strong positive responses to this technique of improving realism. From the quantitative evaluation, the results revealed that the technique has dramatically outpaced the stateof- the-art techniques with a speed of 74 fps in improving the performance for indoor environments.