Real time image processing tasks as well as high-dimensional signal processing applications require efficient implementations beyond the scope of baseline realisations. The objective of this task is to exploit the parallel computing capacities of GPU (graphics processing units) processors and to realize GPU implementations of the Baseline Algorithms.
On top of these baseline algorithms some advanced image processing algorithms, which are needed for the proof of principle application Baseline Algorithms.
On top of these baseline algorithms some advanced image processing algorithms, which are needed for the proof of principle application 3D MALDI Imaging, will be implemented.
We have the following aims:
- GPU Implementation of baseline algorithms: The Baseline Algorithms will be realized using parallel computing on GPU processors.
- Advanced image processing algorithms: Building on the GPU realizations of the baseline algorithms some advanced image processing algorithms (segmentation, classification) will be implemented on a GPU processor.
- Evaluation: The efficiency in terms of accuracy, required computation time, speed up factors and storage requirements will be evaluated in comparison with other GPU implementations using standard function systems for coding/decoding.