The prevalence of cheap and powerful consumer level graphics accelerated hardware introduces a significant growth of 3D applications. In this paper, we have proposed a GPU-accelerated blind and robust watermarking approach to the 3D polygon meshes on the basis of the geometry image transform and image watermarking, which performs watermark embedding and detection on the basis of the geometry image derived from a spherical parametrisation of the input mesh with the help of massive-parallel processing power of the GPUs on the display card. The experimental results show that our approach is successful in at least two aspects. First, the watermark is robust, the embedded watermark survives from common geometric attacks, cropping, simplification, and re-meshing attacks. Second, with the help of parallel computations on the GPUs, the embedding and detection process is extremely fast.

GPU-accelerated blind and robust 3D mesh watermarking by geometry image

In this paper, we implemented both sequential and parallel version of fractal image compression algorithms using CUDA (Compute Unified Device Architecture) programming model for parallelizing the program in Graphics Processing Unit for medical images, as they are highly similar within the image itself. There are several improvement in the implementation of the algorithm as well. Fractal image compression is based on the self similarity of an image, meaning an image having similarity in majority of the regions. We take this opportunity to implement the compression algorithm and monitor the effect of it using both parallel and sequential implementation. Fractal compression has the property of high compression rate and the dimensionless scheme. Compression scheme for fractal image is of two kind, one is encoding and another is decoding. Encoding is very much computational expensive. On the other hand decoding is less computational. The application of fractal compression to medical images would allow obtaining much higher compression ratios. While the fractal magnification an inseparable feature of the fractal compression would be very useful in presenting the reconstructed image in a highly readable form. However, like all irreversible methods, the fractal compression is connected with the problem of information loss, which is especially troublesome in the medical imaging. A very time consuming encoding pro- cess, which can last even several hours, is another bothersome drawback of the fractal compression.

GPU Accelerated Fractal Image Compression for Medical Imaging in Parallel Computing Platform

The paper is focused on computing acceleration for hybrid multiprocessor environment. A considered algorithm of content authentication used for digital images and video sequences is implemented and tested in diverse scenarios. Particular goal is capability of the authentication system to process high-resolution digital images or FullHD video sequences in real time. Aim of this work was to explore and take advantage of mixed CPU and GPU processing approach and to investigate possibilities to develop optimal authentication algorithm. Chosen algorithm is based on robust hashes and semi-fragile digital watermarking. Parallelization was achieved using combined OpenCL and OpenMP. Results for time execution were measured for both images and videos. Based on collected results acceleration rates were calculated along with maximum frames per second values for video processing. It can be concluded that parallelism contributes significantly in reducing the computation time by making optimal use of resources. Depending on the test scenario, the rate of acceleration is even thirty times higher comparing with single-core solutions. Introduced modifications reduce execution time while maintaining detection effectiveness.

Implementation of Digital Watermarking Algorithms in Parallel Hardware Accelerators

Debido al gran volumen de informacion que fluye a traves de Internet, las marcas de agua se utilizan ampliamente para proteger la autenticidad e integridad de la informacion. La insercion y la extraccion de marcas de agua se pueden hacer en el dominio espacial o de otros dominios de frecuencia, como la Transformada Discreta del Coseno (DCT) y la Transformada Discreta Wavelet (DWT). La insercion y la extraccion en dominios como DCT tienen un gran costo computacional en comparacion con los metodos espaciales. Sin embargo, el proceso de marcas de agua en el dominio de la frecuencia tiene mejores resultados en calidad y robustez debido al uso de coeficientes no correlacionados. En este trabajo, se propone utilizar una unidad de procesamiento grafico (GPU) para reducir el costo computacional de la insercion y extraccion de los bits de la marca de agua en el dominio de DCT. Se propone, para tomar ventaja de los bloques generados despues de la DCT, asignar la misma configuracion de bloques en la GPU. Tambien se hace uso de los diferentes tipos de memoria, como la constante y compartida, para optimizar el uso de los recursos del GPU. Los experimentos evaluan el desempeno de la marca de agua en la GPU, y muestran que el algoritmo que se ejecuta en la GPU es hasta 6 veces mas rapido en comparacion con el ejecutado en el CPU, aun tomando en consideracion el tiempo que lleva transferir datos desde la memoria RAM a la memoria de la GPU.

/pdf/a-parallel-watermarking-application-on-a-g-22k3ofvrrq.pdf

A Parallel Watermarking application on a G

As the demand for research on Image/ Content authentication has significantly increased, many authentication schemes have been proposed so far. But most of them are time consuming. This research concentrates on decreasing the time needed by an Image authentication algorithm. In this paper, we have shown a CUDA-based implementation of content authentication algorithm with NVIDIA's GeForce 8400 GS GPU device. Compared to original implementation on CPU, CUDA-based implementation works has shown significant amount of improvement in terms of operational speed up.

Image Authentication Algorithm on GPU

There has been large amounts of research on image authentication method. Many of the schemes perform well in verification results; however, most of them are time- consuming in traditional serial manners. And improving the efficiency of authentication process has become one of the challenges in image authentication field today. In the future, it's a trend that authentication system with the properties of high performance, real-time, flexible and ease for development. In this paper, we present a CUDA-based implementation of an image authentication algorithm with NVIDIA's Tesla C1060 GPU devices. Comparing with the original implementation on CPU, our CUDA-based implementation works 20x-50x faster with single GPU device. And experiment shows that, by using two GPUs, the performance gains can be further improved around 1.2 times in contras to single GPU.

/pdf/a-high-performance-image-authentication-algorithm-on-gpu-2beo47rq8x.pdf

A High Performance Image Authentication Algorithm on GPU with CUDA

Image authentication is an important technology to protect images from being malicious tampered and have became an indispensable part of digital world. The main schemes used for image authentication are signature and watermarking in the last decade. However, in traditional serial manners, the operations of both methods are time-consuming, and limit the wide use of image authentication. In this paper, we proposed a CUDA-based implementation of image authentication algorithm with NVIDIA's Tesla C1060. Comparing with the original implementation in CPU, our CUDA-based implementation works 19-36 times faster.

A CUDA Based Implementation of an Image Authentication Algorithm

Most recent schema matching systems combine multiple components, each of which employs a particular matching technique with several knobs. The multi- component nature has brought a tuning problem, that is to determine which components to execute and how to adjust the knobs (e.g., thresholds, weights, etc.) of these components for domain users. In this paper, we present an approach to automatically tune schema matching systems using genetic algorithms. We match a given schema S against generated matching scenarios, for which the ground truth matches are known, and find a configuration that effectively improves the performance of matching S against real schemas. To search the huge space of configuration candidates efficiently, we adopt genetic algorithms (GAs) during the tuning process. To promote the performance of our approach, we implement parallel genetic algorithms on graphic processing units (GPUs) based on NVIDIA's Compute Unified Device Architecture (CUDA). Experiments over four real-world domains with two main matching systems demonstrate that our approach provides more qualified matches over different domains.

Tuning Schema Matching Systems using Parallel Genetic Algorithms on GPU

Transaction parallelism in database systems is an attractive way of improving transaction performance. There exists two levels of transaction parallelism, inter-transaction level and intra-transaction level. With the advent of multicore processors, new hopes of improving transaction parallelism appear on the scene. The greatest execution efficiency of concurrent transactions comes from the lowest dependencies of them. However, the dependencies of concurrent transactions stand in the way of exploiting parallelism. In this paper, we present Resource Snapshot Model (RSM) for resource modeling in both levels. We propose a non-restarting scheduling algorithm in the inter-transaction level and a processor assignment algorithm in the intra-transaction level in terms of multi-core processors. Through these algorithms, execution performance of transaction streams will be improved in a parallel system with multiple heterogeneous processors that have different number of cores.

Jiwen Yang

Papers

A High Performance Image Authentication Algorithm on GPU with CUDA

A CUDA Based Implementation of an Image Authentication Algorithm

Tuning Schema Matching Systems using Parallel Genetic Algorithms on GPU

Resources Snapshot Model for Concurrent Transactions in Multi-Core Processors