Wael Abdelrahman

Augmented optometry training simulator with multi-point haptics

Training of optometrists is traditionally achieved under close supervision of peers and superiors. With the rapid advancement in technology, medical procedures are performed more efficiently and effectively, resulting in faster recovery times and less trauma to the patient. However, application of this technology has made it difficult to effectively demonstrate and teach these manual skills as the education is now a combination of not only the medical procedure but also the use of the technology. In this paper we propose to increase the training capabilities of optometry students through haptically-enabled single-point and multi-point training tools as well as augmented reality techniques. Haptics technology allows a human to touch and feel virtual computer models as though they are real. Through physical connection to the operator, haptic devices are considered to be personal robots that are capable of improving the human-computer interaction with a virtual environment. These devices have played an increasing role in developing expertise, reducing instances of medical error and reducing training costs. A haptically-enabled virtual training environment, integrated with an optometry slit lamp instrument can be used to teach cognitive and manual skills while the system tracks the performance of each individual. These interactions would ideally replicate every aspect of the real procedure, consequently preparing the trainee for every possible scenario, without risking the health of a real patient.

A comparative study of supervised learning techniques for data-driven haptic simulation

This paper focuses on the choice of a supervised learning algorithm and possible data preprocessing in the domain of data-driven haptic simulation. This is done through a comparison of the performance of different supervised learning techniques with and without data preprocessing. The simulation of haptic interactions with deformable objects using data-driven methods has emerged as an alternative to parametric methods. The accuracy of the simulation depends on the empirical data and the learning method. Several methods were suggested in the literature and here we provide a comparison between their performance and applicability to this domain. We selected four examples to be compared: singular learning mechanism which is artificial neural networks (ANN), attribute selection followed by ANN learning process, ensemble of multiple learning techniques, and attribute selection followed by the learning ensemble. These methods performance was compared in the domain of simulating multiple interactions with a deformable object with nonlinear material behavior.

Extracting 3D Mesh Skeletons Using Antipodal Points Locations

Finding the skeleton of a 3D mesh is an essential task for many applications such as mesh animation, tracking, and 3D registeration. In recent years, new technologies in computer vision such as Microsoft Kinect have proven that a mesh skeleton can be useful such as in the case of human machine interactions. To calculate the 3D mesh skeleton, the mesh properties such as topology and its components relations are utilized. In this paper, we propose the usage of a novel algorithm that can efficiently calculate a vertex antipodal point. A vertex antipodal point is the diametrically opposite point that belongs to the same mesh. The set of centers of the connecting lines between each vertex and its antipodal point represents the 3D mesh desired skeleton. Post processing is completed for smoothing and fitting centers into optimized skeleton parts. The algorithm is tested on different classes of 3D objects and produced efficient results that are comparable with the literature. The algorithm has the advantages of producing high quality skeletons as it preserves details. This is suitable for applications where the mesh skeleton mapping is required to be kept as much as possible.

Data-based dynamic haptic interaction model with deformable 3D objects

The data-based modeling of the haptic interaction simulation is a growing trend in research. These techniques offer a quick alternative to parametric modeling of the simulation. So far, most of the use of the data-based techniques was applied to static simulations. This paper introduces how to use data-based model in dynamic simulations. This ensures realistic behavior and produce results that are very close to parametric modeling. The results show that a quick and accurate response can be achieved using the proposed methods.

Wireless haptic rendering for mobile platforms

Computer haptics has so far been performed on a personal computer (PC). Off the shelf haptic devices provide only PC interfaces and software drivers for control and communication. The new wave of high capable tablet PCs and high end smart phones introduced new platforms for haptic applications. The major problem was to communicate wirelessly to provide user convenience and support mobility which is an essential feature for these platforms. In this paper we provide a wireless layered communication protocol and a hardware setup that enables off the shelf haptic devices to communicate wirelessly with a mobile device. The layers in the protocol enable the change of any hardware components without affecting the data flow. However, the adoption of the wireless interface instead of the wired one comes with the price of speed. Haptic refresh loops require a relatively high refresh rate of 1000 Hz compared to graphics loop which require between 30 and 60 only. An interpolation algorithm was demonstrated to compensate the latency and secure a stable user experience. The introduced setup was tested against portable environments and the users could perform similar functionalities to what are available on a wired setup to a PC.

Physical modeling of heterogeneous embedded deformable object deformation

Simulation of the interactions with deformable models is important in many applications such as medical training and tissue engineering. To physically model the 3D object, both the inner and outer segments need to be considered. This implies dealing with different materials and hence different deformation behavior. Thus, a physically-based simulation needs to augment the behavior of embedded materials when the materials are in direct physical contact, and produce a plausible net result in both visual and haptic cues.

Towards a parameterless 3D mesh segmentation

This paper proposes a novel technique for 3D mesh segmentation using multiple 2D pose footprints. Such a problem has been targeted many times in the literature, but still requires further development especially in the area of automation. The proposed algorithm applies cognition theory and provides a generic technique to form a 3D bounding contour from a seed vertex on the 3D mesh. Forming the cutlines is done in both 2D and 3D spaces to enrich the available information for the search processes. The main advantage of this technique is the possibility to operate without any object-dependent parameters. The parameters that can be used will only be related to the used cognition theory and the seeds suggestion, which is another advantage as the algorithm can be generic to more than one theory of segmentation or to different criterion. The results are competitive against other algorithms, which use object-dependent or tuning parameters. This plus the autonomy and generality features, provides an efficient and usable approach for segmenting 3D meshes and at the same time to reduce the computation load.

Physically based simulation of heterogeneous deformable models using XFEM

This paper addresses the problem of heterogeneous deformable model accuracy using the finite element methods (FEM). Classic FEM uses predefined shape functions for interpolation and does not account easily for regions of discontinuities. Extended finite element methods (XFEM) use enrichment functions to compensate for the change in an element degrees of freedom (DoFs) in deformable objects. The XFEM is an accurate and fast method as no remeshing is required. In this study we investigate the performance of XFEM and demonstrate how it may be applied to discontinuities of materials that exist in heterogeneous (piece-wise homogeneous) models. The results show realistic stress prediction compared to modeling the same objects with classic FEM.

Automated 3D mesh segmentation using 2D footprints

Segmenting 3D meshes into distinct components is vital and necessary for more efficient processing and usability. The smaller segments are usually easier to process and can be associated with with semantics or geometric features. This can be used in 3D parametrization, 3D database creation, animation, deformation transfer and many other 3D graphics applications. However, automating such a process is challenging due to the variety and complexity of the input.

Automatic deformation transfer for data-driven haptic rendering

This paper addresses a major challenge in data-driven haptic modeling of deformable objects. Data-driven modeling is done for specific objects and is difficult to generalize for nearly isometric objects that have similarities in semantics or topology. This limitation prevents the wide use of the data-driven modeling techniques when compared with parametric methods such as finite element methods. The proposed solution is to incorporate deformation transfer methods when processing similar instances. The contributions of this work are focused on the novel automatic shape correspondence method that overcomes the problems of symmetry and semantics presence requirement. The results shows that the proposed algorithm can efficiently calculate the correspondence and transfer deformations for a range of similar 3D objects.