Jaap Smit

Design of an on-chip reflectance map

A reflectance map design is described which uses a minimal amount of memory for the table, in order to be applicable as an on-chip shader. The shader is designed for use with the volumetric super resolution hardware, which performs shading at supersampled locations. However, the design may be used as well to support surface visualization applications. Despite the small table size, the image quality obtained is excellent, even on smooth surfaces.

Mapping of DSP Algorithms on Field Programmable Function Arrays

This position paper discusses reconfigurability issues in low-power hand-held multimedia systems. A reconfigurable systems-architecture is introduced, with a focus on a Field Programmable Function Array (FPFA). Application domain specific algorithms determine the granularity of FPFA processor tiles. Several algorithms are discussed and mapped onto a FPFA processor tile.

Chameleon - Reconfigurability in Hand-Held Multimedia Computers

In this paper a reconfigurable systems-architecture in combination with a QoS driven operating system is introduced that can deal with the inherent dynamics of future mobile systems. We claim that a radical new approach has to be taken in order to fulfill the requirements - in terms of processing power and energy consumption - of future mobile applications.

Image Quality Improvements in Volume Rendering

This paper presents some methods to improve the image quality obtained with volume rendering. By computing the opacity, color and gradient of each sample point directly at the sample position, the image quality has improved over methods which compute these values at the voxel positions. A new method for calculating the gradient is presented. These improvements result in small details becoming clearly visible. It also allows high zoom rates without generating blurry images. The opacity is corrected for the sample rate, allowing a consistent translucency setting.

Graphics algorithms on field programmable function arrays

The amount of energy consumed in basic CMOS building blocks, like external RAM, external bus-structures, multipliers, local (cache) memory and on chip bus-structures, is analyzed thoroughly to find ways for substantial improvement of the power consumption of high speed graphics algorithms. A Field Programmable Function Array capable of low-power execution of a wide range of algorithms is introduced. Aspects of the compilation of the volume rendering algorithm to this architecture are discussed.

On the energy complexity of algorithms realized in CMOS, a graphics example

A theory about the energy consumption of algorithms realized in CMOS, presented in related work, makes it possible to calculate the minimal amount of energy dissipated for the execution of an algorithm. The rendering of a dense dataset with three variants of the Volume Rendering algorithm with be considered as an example of the methodology. The absolute lower bound of the energy consumption is calculated for the rendering ofa dense 256^3 dataset using implementations of the algorithms in an 1µm CMOS process. Predictions of the energy consumption in future CMOS generations are given as well.

Improving image quality of volume-rendered three-dimensional medical data

Volume visualization is the technique of displaying two dimensional projections of three dimensional data. The data is acquired from a medical scanner, like MRI, CT, SPECT, or US scanners. Visualizing a given three dimensional medical dataset can be done by surface rendering algorithms or by direct volume rendering algorithms. Surface rendering algorithms require an intermediate geometric representation and are therefore less attractive. In our approach volume rendering is used. To improve image quality of such projections of the volume data, special care should be taken to (a) the interpolation step, (b) the estimation of the local gradient and (c) the assignment of opacity values at sample positions. These aspects are addressed in this paper.

Design of a parallel VLSI engine for real-time visualization of 3D medical images

Three dimensional medical scanners are widely available in todays hospitals to acquire a dataset of the human body without the need for surgery. The usefulness of this diagnostic information is limited by the lack of techniques to visualize the datasets. With the increasing computer power of todays workstations it is possible to make a transparent view of the 3D dataset. An interactive mode is necessary, however, to fully explore the 3D dataset. If both a high resolution and a high interactive speed is required, the necessary computational power is enormous. Therefore it is necessary to map the algorithms for volume visualization in a rather specific way onto (dedicated) chips to overcome the performance gap. This paper discusses a high-performance special-purpose low-power system, the Real-Time Volume Rendering Engine (RT-VRE), capable of rendering a 3D dataset of 2563 voxels onto a display of 7502 pixels with an interaction rate of 25 images per second. The RT-VRE allows biomedical engineers to interactively visualize and investigate their data.

MOD/R : A knowledge assisted approach towards top-down only CMOS VLSI design

MOD/R models all views on the design space in relations. This is achieved by eliminating the package constraints, as are apparent in PCB oriented hardware description languages. Assisted by knowledge engineering it allows for a top-down, mostly hierarchical decomposition, virtually eliminating the need for bottom-up assembly.

Accurate measurements in volume data

An algorithm for very accurate visualization of an iso- surface in a 3D medical dataset has been developed in the past few years. This technique is extended in this paper to several kinds of measurements in which exact geometric information of a selected iso-surface is used to derive volume, length, curvature, connectivity and similar geometric information from an object of interest. The actual measurement tool described in this paper is fully interactive. The highly accurate iso-surface volume- rendering algorithm is used to describe the actual measurement that should be performed. For instance, objects for which volumes should be calculated, or paths from which the length should be calculated can be selected at sub-voxel resolution. Ratios of these quantities can be used to automatically detect anomalies in the human body with a high degree of confidence. The actual measurement tool uses a polygon-based algorithm that can distinguish object connectivity at sub-voxel resolution, in exactly the same manner as the iso-surface algorithm. Segmentation based on iso-surfaces geometrical topology can be done at this point. The combination of the iso-surface volume-rendering algorithm and the polygon-based algorithm makes it possible to achieve both visual interaction with the dataset and highly accurate measurements. We believe that the proposed method contributes to the integration of visual and geometric information and is helpful in clinical diagnosis.