| Abstract [eng] |
Rapid development of digital technologies has changed the approach to the intellectual property. Because of this reason, the need for protection of intellectual property is increasing, especially in the field of multimedia. One of the most popular methods of video content protection is video watermarking. In order to ensure proper protection of intellectual property, it is important to do a research of existing methods of video watermarking and to find out which of them is more efficient. Moreover, a detailed analysis, based on difference between original watermark and extracted watermark after various image attacks and compression would reveal which compression algorithm is more accurate. Therefore, in this Master thesis watermark embedding methods are analyzed, such as discrete cosine transform and discrete wavelet transform. After that, both methods were applied in the experiment, where the resistance of watermark to image attacks were analyzed. Experiment results showed that more effective is discrete wavelet transformation and this method was applied in video watermarking. Then watermarked video was modified by various attacks and compressed using two different algorithms in order to find out level of watermark robustness. Overall, based on results of the experiment, extracted watermark is more robust and visually recognizable when watermarked video content is compressed using H.264 algorithm than compressed using H.265. |
