Printout verification using a new image hash algorithm based on Radon, Wavelet, and Fast Fourier transforms

Certificate and document issuing processes are often in a form of printing document to a paper. Some might improve the trustworthiness of their paper via using secure paper or secure printing method such as watermark, hologram etc. The interesting problem is that this security mechanism is not enough because nowadays printer with secure printing module can be obtained easier than before, which means reproducing counterfeit document can be done uncomplicatedly. To solve this problem, a security framework using image hashing function to capture document fingerprint and perform a digital signature based on PKI-enabled infrastructures is proposed to guarantee this fingerprint integrity. The key component of this framework is using image hashing algorithm based on Radon, Wavelet and Fast Fourier transforms to capture perceptual appearance of printout and verify integrity of their document. Hash value from the proposed algorithm is invariant to some unintentional factors such as noise or some duplication process without any change of content while it is robust to some intentional attacks such as text and image editing. With the experimental results, our proposed scheme outperforms other existing image hashing algorithms in terms of both false acceptance rate and false rejection rate.