An image is processed to encode a digital watermark, with different regions thereof processed using different levels of watermark intensity. In an image comprised of elements of differing sizes (e.g., halftone shapes of different sizes, or lines of different width), the different regions can be defined by reference to the sizes of elements contained therein. Regions characterized by relatively small elements can be watermarked at a relatively low intensity. Regions characterized by relatively large elements can be watermarked at a relatively high intensity. A variety of other features are also discussed.