![Open Access](https://i-scholar.in/lib/pkp/templates/images/icons/fulltextgreen.png)
![Restricted Access](https://i-scholar.in/lib/pkp/templates/images/icons/fulltextred.png)
![Open Access](https://i-scholar.in/lib/pkp/templates/images/icons/fulltextgreen.png)
![Open Access](https://i-scholar.in/lib/pkp/templates/images/icons/fulltext_open_medium.gif)
![Restricted Access](https://i-scholar.in/lib/pkp/templates/images/icons/fulltextred.png)
![Restricted Access](https://i-scholar.in/lib/pkp/templates/images/icons/fulltext_restricted_medium.gif)
Parallel Hybrid PSO-Based Fragile Image Watermarking
Subscribe/Renew Journal
Particle Swarm Optimization (PSO) algorithm is used in the literature to solve different types of optimization problems. Most PSO algorithms suffer from the problem of the long processing time and the fact that PSO gets trapped easily in a local minimum. This paper argues the need to address both problems simultaneously to enhance the performance of PSO unlike other research studies in the literature that typically address only one of these problems. To emphasize this, the paper proposes to combine the basic PSO algorithm with both evolutionary operators and parallel processing to solve the rounding error problem of a DCT-based fragile image watermarking algorithm. Evolutionary operators help PSO jump out of local minima, while parallel processing helps speed up processing. In such watermarking algorithms, the watermark bits are typically embedded by modifying the least significant bits of the frequency coefficients of the DCT-transformed host image. The host image is then converted into the spatial domain using inverse DCT and the real numbers of the inverse-DCT coefficients are rounded to integers. This enables detecting the slightest modification of the host image, but the rounding process results in a significant difference between the extracted watermark and the embedded one and reduces the watermarked host image quality. The proposed algorithm is compared to three cases: a) using PSO alone, b) using PSO with evolutionary operators without parallel processing, and c) using PSO with parallel processing and without evolutionary operators. The experimental results show the superiority of our approach.
Keywords
![](https://i-scholar.in/public/site/images/abstractview.png)
Abstract Views: 183
![](https://i-scholar.in/public/site/images/pdfview.png)
PDF Views: 1