Digital watermarks have been embedded invisibly in digital media to protect copy rights of legal owners. The embedded watermarks can be extracted to indicate ownership originals. In this paper a robust Full-Band Image Watermark method is investigated by using various attacks on the watermarked image. Image attacks have evolved so much that they are not just destructive, that is, degrading the visual quality of the image, but somehow, constructive, that is, altering the original work in creative ways to make it look as if a new piece of creation. We launched many destructive and constructive attacks on images embedded with our watermarking method to test its security. The experimental results prove that it is very robust against image attacks.