NeRF

We present a method that achieves state-of-the-art results for synthesizing novel views of complex scenes by optimizing an underlying continuous volumetric scene function using a sparse set of input views. Our algorithm represents a scene using a fully connected (nonconvolutional) deep network, whose input is a single continuous 5D coordinate (spatial location ( x , y , z ) and viewing direction ( θ, ϕ )) and whose output is the volume density and view-dependent emitted radiance at that spatial location. We synthesize views by querying 5D coordinates along camera rays and use classic volume rendering techniques to project the output colors and densities into an image. Because volume rendering is naturally differentiable, the only input required to optimize our representation is a set of images with known camera poses. We describe how to effectively optimize neural radiance fields to render photorealistic novel views of scenes with complicated geometry and appearance, and demonstrate results that outperform prior work on neural rendering and view synthesis.

Combining multiple shaded reliefs with hypsometric colouring and digital orthophotos using free and open-source software

In this paper we combined layers created from several terrain rendering techniques, namely a shaded relief rendered in the free and open-source 3D computer graphics software Blender, a hillshade created in the free and opensource Geographic Information System (GIS) software QGIS, a hypsometric coloured Digital Elevation Model (DEM) and a draped digital orthophoto. Following a recent trend in the cartographic community towards using Blender, we tried to improve the standard relief visualization in common GIS software by blending it with a shaded relief rendered in Blender. Using different QGIS blending modes and opacity values we achieved different graphic visualizations. To compare and evaluate the suitability of different rendering techniques we chose national park Risnjak located in Croatia because of its specific and diverse terrain landforms. After comparing different input layers and parameter sets, we selected the blending combination which is best suited for visualizing terrain characteristics of all Croatian national parks. The result is a shaded relief created for every national park which is combined from a shaded relief rendered in Blender, a hillshade created in QGIS, a hypsometric coloured DEM and a draped digital orthophoto.

Photorealism in Mixed Reality: A Systematic Literature Review

In Augmented Reality systems, virtual objects are combined with real objects, both three dimensional, interactively and at run-time. In an ideal scenario, the user has the feeling that real and virtual objects coexist in the same space and is unable to differentiate the types of objects from each other. To achieve this goal, research on rendering techniques have been conducted in recent years. In this paper, we present a Systematic Literature Review aiming to identify the main characteristics concerning photorealism in Mixed and Augmented Reality systems to find research opportunities that can be further exploited or optimized. The objective is to verify if exists a definition of photorealism in Mixed and Augmented Reality. We present a theoreticalfundamental over the most used methods concerning realism in Computer Graphics. Also, we want to identify the most used methods and tools to enable photorealism in Mixed and Augmented Reality systems.

Structure-Aware Trail Bundling for Large DTI Datasets

Creating simplified visualizations of large 3D trail sets with limited occlusion and preservation of the main structures in the data is challenging. We address this challenge for the specific context of 3D fiber trails created by DTI tractography. For this, we propose to jointly simplify trails in both the geometric space (by extending and adapting an existing bundling method to handle 3D trails) and in the image space (by proposing several shading and rendering techniques). Our method can handle 3D datasets of hundreds of thousands of trails at interactive rate, has parameters for the most of which good preset values are given, and produces visualizations that have been found, in a small-scale user study involving five medical professionals, to be better in occlusion reduction, conveying the connectivity structure of the brain, and overall clarity than existing methods for the same data. We demonstrate our technique with several real-world public DTI datasets.

Effective volume rendering on mobile and standalone VR headsets by means of a hybrid method

Numerous volume rendering techniques are available to display 3D datasets on desktop computers and virtual reality devices. Recently the spreading of mobile and standalone virtual reality headsets has brought the need for volume visualization on these platforms too. However, the volume rendering techniques that show good performance in desktop environment underachieve on these devices, due to the special hardware conditions and visualization requirements. To speed up the volumetric rendering to an accessible level a hybrid technique is introduced, a mix of the ray casting and 3D texture mapping methods. This technique increases 2-4 times the frame rate of displaying volumetric data on mobile and standalone virtual reality headsets as compared to the original methods. The new technique was created primarily to display medical images but it is not limited only to this type of volumetric data.