A novel method for embedding and detecting a chaotic watermark in the digital spatial image domain, based on segmenting the image and locating regions that are robust to several image manipulations, is presented in this paper. Each selected region is approximated by an ellipse. The watermark is embedded on its bounding rectangle. This representation proves robust under geometric attacks. The controlled lowpass nature of the chaotic watermark ensures its immunity to lowpass filtering and JPEG compression. Experimental results display the robustness of the method under several kinds of attacks, such as JPEG compression, mean and median filtering, scaling, cropping and rotation.