A novel image hash algorithm resistant to print–scan

With the rapid development of multimedia technology, how to establish the integrity of images has become an unavoidable problem. The transform between a digital image and its hard-copy makes the problem more complex. To simplify the content based authentication, we propose a print–scan resistant image hashing algorithm based on the Radon and wavelet transform. The Radon transforms an image to its luminance distribution, before the wavelet extracts the relationship of the different areas from the luminance distribution. Experimental results show that our algorithm is not only robust to print–scan and other common content-preserving processing, but also discriminable to content changes.