Watermarking Massive Remote Sensor Images in Parallel

The watermarking of digital images, video and multimedia products in general has been proposed for resolving copyright and verifying originality of content. Although there have been many tools for batch adding watermarking, the volume of data is very small. When the data volume increase, the general tools or traditional technologies are inefficient or even no use. The parallel method of adding watermarking on massive remote sensor image on Hadoop is researched. We extend the file visit approach to make Hadoop regard the whole Tiff files as a unit by expanding the file format that Hadoop support, especially making Hadoop support GeoTiff. Experiments show that the method is effective and scalable and Hadoop has several deficiencies in dealing with image.