Veronica Teichrieb

Poster: Improving motor rehabilitation process through a natural interaction based system using Kinect sensor

In general, the motor rehabilitation process can take advantage of natural interaction based systems, including measurements from patient performance to track its evolution during time and therapy direction. Thus, the aim of this research is the analysis of the use of Kinect sensor as interaction support tool for rehabilitation systems. The Kinect sensor gives three-dimensional information about the user body, recognizing skeleton and joint positions, however does not provide the detection of the body specific movements. This way, the correct description of a rehabilitation movement (shoulder abduction, for instance) was implemented in a system prototype. A scoring mechanism was also developed in order to measure the patient performance, as well as to stimulate his improvement by displaying a positive feedback.

Photorealistic rendering for Augmented Reality: A global illumination and BRDF solution

This paper presents a solution for the photorealistic rendering of synthetic objects into dynamic real scenes, in Augmented Reality applications. In order to achieve this goal, an Image Based Lighting approach is used, where environment maps with different levels of glossiness are generated for each virtual object in the scene at every frame. Due to this, illumination effects, such as color bleeding and specular reflections, can be simulated for virtual objects in a consistent way. A unifying sampling method for the spherical harmonics transformation pass is also used. It is independent of map format and does not need to apply different weights for each sample. The developed technique is combined with an extended version of Lafortune Spatial BRDF, featuring Fresnel effect and an innovative tangent rotation parameterization. The solution is evaluated in various Augmented Reality case studies, where other features like shadowing and lens effects are also exploited.

Massively Parallel Nearest Neighbor Queries for Dynamic Point Clouds on the GPU

We introduce a parallel algorithm to solve approximate and exact nearest neighbor queries on the GPU, exploiting its massively parallel processing power. Both data structure construction and nearest neighbor queries are performed on the GPU, avoiding memory copies from system memory to device memory. This algorithm achieves real-time performance, enabling its usage in dynamic scenarios, by minimizing the sorting comparisons needed for a large K value. The underlying data structure for spatial subdivision handles 3D points and is based on grid spatial hashing. Users can specify the grid size interactively. Comparisons were done with other nearest neighbor algorithms implemented on both CPU and GPU. Our approach clearly surpasses CPU implementations regarding processing time, while it presents a competitive solution to GPU ones. Real-time results were obtained with ANN searches (K = 10) for data sets up to 163K points and the potential of our algorithm is demonstrated through a point-based rendering application.

Teleoperation Using Google Glass and AR, Drone for Structural Inspection

This paper proposes the use of a wearable device for visualization and control in association with an UAV applied to the structural inspection of buildings. More specifically, an AR. Drone is controlled through head positions and gestures performed by the operator wearing a Google Glass, and the images captured by the drone are visualized on Glasss screen. We discuss the problems that arise when such a solution is developed, along with the limitations that come from todays available technology and how to overcome them.

Real Time Ray Tracing for Augmented Reality

This paper introduces a novel graphics rendering pipeline applied to augmented reality, based on a real time ray tracing paradigm. Ray tracing techniques process pixels independently from each other, allowing an easy integration with image-based tracking techniques, contrary to traditional projection-based rasterization graphics systems, e.g. OpenGL. Therefore, by associating our highly optimized ray tracer with an augmented reality framework, the proposed pipeline is capable to provide high quality rendering with real time interaction between virtual and real objects, such as occlusions, soft shadows, custom shaders, reflections and self-reflections, some of these features only available in our rendering pipeline. As proof of concept, we present a case study with the ARToolKitPlus library and the Microsoft Kinect hardware, both integrated in our pipeline. Furthermore, we show the performance and visual results in high definition of the novel pipeline on modern graphics cards, presenting occlusion and recursive reflection effects between virtual and real objects without the latter ones needing to be previously modeled when using Kinect. Furthermore, an adaptive soft shadow sampling algorithm for ray tracing is presented, generating high quality shadows in real time for most scenes.

Challenges in 3D Reconstruction from Images for Difficult Large-Scale Objects: A Study on the Modeling of Electrical Substations

In recent years, 3D reconstruction from images has played a major role in computer vision with a lot of improvements regarding both quality and performance. One of its main uses is the generation of 3D models of objects that are difficult to modeling. In the electrical sector, 3D reconstruction from images shows itself as a candidate to be used in specific scenarios with the advantage of its low price compared to laser scanning techniques. In fact, there are many industrial applications that can use the power of 3D reconstruction from images but no work has focused more deeply on their requirements yet. This paper analyzes the advantages and drawbacks in using 3D reconstruction from image techniques and tools in the uncontrolled environment of an electrical substation from the scenario characteristics and tools limitations points of view. Some representative available tools (commercial and academic) where evaluated and the relationship between scenario/object characteristics and reconstructed model quality could be pointed out for further improvements of the techniques in future work. This results can be used to create industrial applications with large scale difficult objects.

kD-Tree Traversal Implementations for Ray Tracing on Massive Multiprocessors: A Comparative Study

Current GPU computational power enables the execution of complex and parallel algorithms, such as Ray Tracing techniques supported by kD-trees for 3D scene rendering in real time. This work describes in detail the study and implementation of five different kD-Tree traversal algorithms using the parallel framework NVIDIA Compute Unified Device Architecture (CUDA), in order to point their pros and cons regarding adaptation capability to the chosen architecture. In addition, a new algorithm is proposed by the authors based on this analysis, aiming performance improvement. A performance analysis of the implemented techniques demonstrates that two of these algorithms, once adequately adapted to CUDA architecture, are capable of reaching speedup gains up to 15x when compared to former CPU implementations and up to 4x in comparison to existing and optimized parallel ones. As a consequence, interactive frame rates are possible for scenes with 1376x768 pixels of resolution and 1 million primitives.

An Open-Source Framework for Air Guitar Games

This paper presents an open-source framework for developing guitar-based games using gesture interaction. The goal of this work was to develop a robust platform capable of providing seamless real time interaction, intuitive playability and coherent sound output. Each part of the proposed architecture is detailed and a case study is performed to exemplify its easiness of use. Some tests are also performed in order to validate the proposed platform. The results showed to be successful: all tested subjects could reach the objective of playing a simple song during a small amount of time and the most important, they were satisfied with the experience.

Virtual Reality in Brazil 2011: RPR-SORS: Real-time photorealistic rendering of synthetic objects into real scenes

This paper presents a review of the Photorealistic Augmented Reality field and proposes a solution for interactively rendering virtual objects into dynamic real scenes in a photorealistic way. This solution features a rendering pipeline that comprises techniques regarding illumination, reflectance model, shadowing, composition, and camera effects. The techniques are chained in a flexible way, allowing the user to choose which techniques are to be enabled. An environment map generation procedure was developed and allows virtual objects to exhibit coherent effects such as color bleeding and specular reflection, even when the real objects are moved. The range of materials that can be rendered was widened by extending Lafortunes Spatial BRDF. The implemented infrastructure is offered as an authoring toolkit that consists of an API and a material editor tool. The aim of this authoring toolkit is to increase development productivity of Photorealistic Augmented Reality applications. The proposed solution was evaluated by taking into account visual and performance metrics. It allowed consistent rendering of dynamic scenes and photorealistic materials. The frame rate obtained was suitable to Augmented Reality applications when there were few virtual objects in the scene.

Human Body Motion and Gestures Recognition Based on Checkpoints

The computational implementation of human body gestures recognition has been a challenge for several years. Nowadays, thanks to the development of RGB-D cameras it is possible to acquire a set of data that represents a human position in time. Despite that, these cameras provide raw data, still being a problem to identify in real-time a specific pre-defined user movement without relying on offline training. However, in several cases the real-time requisite is critical, especially when it is necessary to detect and analyze a movement continuously, as in the tracking of physiotherapeutic movements or exercises. This paper presents a simple and fast technique to recognize human movements using the set of data provided by a RGB-D camera. Moreover, it describes a way to identify not only if the performed motion is valid, i.e. belongs to a set of pre-defined gestures, but also the identification of at which point the motion is (beginning, end or somewhere in the middle of it). The precision of the proposed technique can be set to suit the needs of the application and has a simple and fast way of gesture registration, thus, being easy to set new motions if necessary. The proposed technique has been validated through a set of tests focused on analyzing its robustness considering a series of variations during the interaction like fast and complex gestures.