Spatial Domain Digital Watermarking is a technique for the insertion of watermarked information (side information defined by the owner) into the source (cover) image/video in the spatial domain.
The most common algorithm for spatial domain watermarking is Least Significant Bit Modification. This method changes the least significant bits (LSB) of chosen pixels in the image. It is possible to use more LSB bits of the container image in a similar way.
The watermark object may also be embedded many times within the container image. Even if most of the watermarks are lost due to attacks, a single surviving one is enough.
This method is comparatively simple. It can survive simple operations such as cropping and addition of noise. However lossy compression is going to defeat the watermark. Also, a simple attack that sets all the LSB bits to ‘1’ will defeat the watermark with negligible perceptual impact to the cover object.
To extract the watermark, the LSB plane is extracted from the watermarked image and an exclusive-or operation is done using the watermark template.
To prevent extraction of the hidden information directly from the chroma LSB, the cover image may be pre-processed. One preprocessing technique is to split the cover image by blocks. The embedded watermark is then encoded by modifying the relationship between neighboring blocks.