Optical Encryption Based On Computational Ghost Imaging And Watermarking

Ghost imaging is a new imaging method to obtain object information based on the spatial intensity correlation. With the development of computational ghost imaging scheme,its application has enlarged to other fields, such as optical encryption, watermarking, which not only promises the advantages of optical encrytion and watermarking,but also has the feature of nonlocality. The studies on the optical encryption and watermarking based on computational ghost imaging are important and referenced.Firstly, based on the principle of computational ghost imaging and the advantages of QR-coded, an optical encryption scheme based on QR-coded computational ghost imaging is proposed in this thesis. N random phase screens are generated by Alice as a secret key and then shared with Bob. An image is firstly encoded by Alice with QR code, and then the QR-coded image is further encrypted with the aid of computational GI optical system. Measurement results from the bucket detector in GI system, corresponding to N random phase screens, are transmitted to Bob. With the random phase screens and the measurement results, Bob firstly obtains a reconstructed QR-coded image by computational GI technique, and then decrypts the QR-coded image using QR decoding APP at a remote party. The experimental and numerical results show that the authorized users(Alice and Bob) can recover completely the original image, but the eavesdroppers can not acquire any information about the image even the eavesdropping ratio(ER) is up to 60%.Secondly, an optical encryption based on QR-coded compressive ghost imaging scheme is further proposed with compressive sensing technique and is verified by numerical simulation and experimental. The experimental results match those results from theoretical analysis and numerical simulation. The results prove that our schemes are practical at the real environment, can reduce the amount of data and recovery time without decreasing the security and the quality of image.Finally, with the application of watermarking in copyright protection, information encryption and secure communication field, the security of the watermarking is more and more crucial. A watermarking scheme based on computational ghost imaging is proposed in this thesis. In the scheme, the watermark image is firstly encrypted by utilizing the no-locality of computational ghost imaging. N random phase screens are generated as a secret key. N measurement results from the bucket detector are coded into two value matrix and then embedding the host image. Simulation results show that authorized users can extract watermark image by secret key and correlation recovery algorithm. In addition, the scheme has high security performances, because the unauthorized user can’t extract clear watermarks from the recovered image even eavesdrop the partial secret key.