The conventional computational integral-imaging reconstruction (CIIR) method has some drawbacks of intensity irregularity and brightness degradation in the reconstructed pane object images (POIs) due to its insufficient and uneven overlapping of the projected elemental images. Thus, in this paper, a modified CIIR (M-CIIR) technique is proposed to dramatically enhance the viewing quality of the computationally reconstructed object images. For this, a pixel-count matrix (PCM) is introduced in the CIIR process to calculate the number of overlapped pixels of the projected elemental images in the reconstructed object plane. Each pixel value of the reconstructed POI, then, is determined as the average of all overlapped pixel values. As the pixel values of the reconstructed POI are regulated by the PCM, viewing quality of the reconstructed POIs can be dramatically improved in the proposed method. In order to show the feasibility of the proposed method, experiments are performed, and the results are compared and discussed with those of the conventional method.

Interaction with 3D information is one of the fundamental and most familiar tasks in virtually all areas of engineering and science. Several recent technological advances pave the way for developing hand gesture recognition capabilities available to all, which will lead to more intuitive and efficient 3D user interfaces (3DUI). These developments can unlock new levels of expression and productivity in all activities concerned with the creation and manipulation of virtual 3D shapes and, specifically, in engineering design. Building fully automated systems for tracking and interpreting hand gestures requires robust and efficient 3D imaging techniques as well as potent shape classifiers. We survey and explore current and emerging 3D imaging technologies, and focus, in particular, on those that can be used to build interfaces between the users' hands and the machine. The purpose of this paper is to categorize and highlight the relevant differences between these existing 3D imaging approaches in terms of the nature of the information provided, output data format, as well as the specific conditions under which these approaches yield reliable data. Furthermore we explore the impact of each of these approaches on the computational cost and reliability of the required image processing algorithms. Finally we highlight the main challenges and opportunities in developing natural user interfaces based on hand gestures, and conclude with some promising directions for future research.

The C 2 variable degree splines1---3 have been proven to be an efficient tool for solving the curve shape-preserving interpolation problem in two and three dimensions. Based on this representation, the current paper proposes a Hermite interpolation scheme, to construct C 1 shape-preserving splines of variable degree. After this, a slight modification of the method leads to a C 1 shape-preserving Hermite cubic spline. Both methods can easily be developed within a CAD system, since they compute directly (without iterations) the B-spline control polygon. They have been implemented and tested within the DNV Software CAD/CAE system GeniE.

In this paper we present a novel stereo view synthesis algorithm that is highly accurate with respect to inter-view consistency, thus to enabling stereo contents to be viewed on the autostereoscopic displays. The algorithm finds identical occluded regions within each virtual view and aligns them together to extract a surrounding background layer. The background layer for each occluded region is then used with an exemplar based inpainting method to synthesize all virtual views simultaneously.Our algorithm requires the alignment and extraction of background layers for each occluded region; however, these two steps are done efficiently with lower computational complexity in comparison to previous approaches using the exemplar based inpainting algorithms. Thus, it is more efficient than existing algorithms that synthesize one virtual view at a time.This paper also describes the implementation of a simplified GPU accelerated version of the approach and its implementation in CUDA. Our CUDA method has sublinear complexity in terms of the number of views that need to be generated, which makes it especially useful for generating content for autostereoscopic displays that require many views to operate.An objective of our work is to allow the user to change depth and viewing perspective on the fly. Therefore, to further accelerate the CUDA variant of our approach, we present a modified version of our method to warp the background pixels from reference views to a middle view to recover background pixels. We then use an exemplar based inpainting method to fill in the occluded regions. We use warping of the foreground from the reference images and background from the filled regions to synthesize new virtual views on the fly. Our experimental results indicate that the simplified CUDA implementation decreases running time by orders of magnitude with negligible loss in quality.

According to IEEE 802.15.4a standardization, an ultra wideband PHY symbol is capable of carrying two bits of information and only one pulse shall be transmitted in each symbol. To prevent inter symbol interference caused by multipath effect, the guard interval is present between the possible burst positions. In order to make full use of the symbol for estimating the azimuth and the elevation of target, we can add additional information at the guard interval position, if the pulse used at the possible burst position is different from the pulse used at the guard interval position. Using the orthonormality between the multiple pulses, furthermore, inter symbol interference can be avoided at the correlation receiver. In this paper, as the first step for estimating the azimuth and the elevation information, we propose a novel PHY symbol structure for IR-UWB system based on IEEE 802.15.4a standardization and orthogonal pulses to enhance the data rates without the bit-error-rate (BER) performance degradation. In the proposed scheme, the pulse used at the possible burst position is the modified hermite polynomial (MHP) pulse of order n equal to 3 and the pulse used at guard interval position is the MHP pulse of order n equal to 4. The performance of the proposed structure is evaluated by computer simulations over the IEEE 802.15.4a channel model. Simulation results show that the proposed symbol structure with orthogonal pulses can double the data rates of the systems without BER performance degradation.

A high resolution micro computed tomography (micro-CT) system for live small animal imaging has been developed. The system consists of an x-ray source with micro focus spot and high brightness, rotating gantry with a x-ray tube and flat panel detector pair and a stationary and a horizontally positioned small animal bed to achieve a conebeam mode scan. The system is optimized for in vivo small animal imaging and the capability of administering respiratory anesthesia during scanning. The Feldkamp algorithm was adopted in image reconstruction with graphic processing unit (GPU). We evaluated the spatial resolution, image contrast, and uniformity of system using phantom. As the result, the spatial resolution of the system was the 56lp/mm at 10% of the MTF curve, and the radiation dose to the sample was 98mGy. The minimal resolving contrast was found to be less than 46 CT numbers on low-contrast phantom. We present the image test results of the bone and lung, and heart of the live mice.

We apply multiple-wavelength digital holography in combination with epi-fluorescence microscopy in order to generate quantitative phase and fluorescence information from cell samples. While digital holography provides high precision morphological information, the addition of fluorescence supplies the specificity needed to identify cellular constituents. By the application of multiple-wavelength digital holography it is possible to obtain the complete wavefront data deterministically and in real-time. We demonstrate this dual-mode imaging capability through the investigation of living cells.

Terrain rendering has been intensively investigated in the last decade, and many methods have been proposed. However, most methods assume a single, static terrain data set that does not change over time, so that elaborate offline preprocessing methods can be used to reduce online rendering time. Such methods cannot be used in applications that work on dynamic terrain data, e.g. interactive terrain modeling, simulation, or remote sensing data analysis applications. In this paper, we present a new data structure that allows GPU-based bottom up and top down mesh refinement. The resulting adaptive triangle meshes can be used for continuous level of detail rendering of dynamic terrain data, while guaranteeing an upper bound on the screen space error. We evaluate our techniques by applying them to dynamic terrain data produced by interactive fusion of multimodal remote sensing data sets.

We present a common framework for accurate and automatic registration of two geometrically complex 3D range scans by using linear or planar features. The linear features of a range scan are extracted with an efficient split-and-merge line-fitting algorithm, which refines 2D edges extracted from the associated reflectance image considering the corresponding 3D depth information. The planar features are extracted employing a robust planar segmentation method, which partitions a range image into a set of planar patches. We propose an efficient probability-based RANSAC algorithm to automatically register two overlapping range scans. Our algorithm searches for matching pairs of linear (planar) features in the two range scans leading to good alignments. Line orientation (plane normal) angles and line (plane) distances formed by pairs of linear (planar) features are invariant with respect to the rigid transformation and are utilized to find candidate matches. To efficiently seek for candidate pairs and groups of matched features we build a fast search codebook. Given two sets of matched features, the rigid transformation between two scans is computed by using iterative linear optimization algorithms. The efficiency and accuracy of our registration algorithm were evaluated on several challenging range data sets.