Optical Cryptosystems. Naveen K. Nishchal

14]. Since 1995, a large number of research articles have appeared with so many different techniques. These topics are discussed in detail in the following chapters.

References

[1] A Al Falou 2018 Advanced Secure Image Processing for Communications (Bristol: IOP Publishing)

      [2] S Ramakrishnan 2018 Cryptographic and Information Security Approaches for Images and Videos (Boca Raton, FL: CRC Press)

[3] J R Vacca 2017 Computer and Information Security Handbook 3rd edn (Amsterdam: Elsevier)

[4] Pfleeger C P, Pfleeger S L and Margulies J 2018 Security in Computing 5th edn (Noida: Pearson India)

[5] Beckett B 1988 Introduction to Cryptology (Oxford: Blackwell)

[6] Stallings W 2000 Cryptography and Network Security; Principles and Practice 2nd edn (Hoboken, NJ: Prentice Hall)

[7] van Tilborg H C A 2000 Fundamentals of Cryptology (Boston, MA: Kluwer)

[8] T J Naughton and J Sheridan 2005 Optics in information systems SPIE Int. Tech. Gr. Newsletter 16 1–12

      [9] B Javidi 2005 Optical and Digital Techniques for Information Security (Berlin: Springer)

      [10] B Javidi 2006 Optical Imaging Sensors and Systems for Homeland Security Applications (New York: Springer)

      [11] Alfalou A and Brosseau C 2009 Optical image compression and encryption methods Adv. Opt. Photon 1 589–636

      [12] Chen W, Javidi B and Chen X 2014 Advances in optical security systems Adv. Opt. Photon 6 120–55

[13] Javidi B et al 2016 Roadmap on optical security J. Opt. 18 083001

[14] Muniraj I and Sheridan J T 2019 Optical Encryption and Decryption (Bellingham, WA: SPIE Press)

      IOP Publishing

      Optical Cryptosystems

      Naveen K Nishchal

      Chapter 2

      Optical techniques of image encryption: symmetric cryptosystems

      2.1 Introduction

      In an encryption system, an input data/image is encoded in such a fashion that only the application of the correct key would reveal the original information. The security of digital methods is being enhanced by using more powerful algorithms. Longer key lengths are chosen such that even advanced computers would require an unreasonable amount of time to break the key. Therefore, digital techniques are falling short of expectation due to the fact that when the encryption key length becomes longer, the processing speed goes down. It is primarily because digital techniques process data serially and in one dimension. Optical processing is inherently two-dimensional (2D) and does parallel processing. Every pixel of the 2D image can be both relayed and processed at the same time. So, when a large volume of data/information is to be processed, parallel processing offers enormous advantages. In addition to the fast speed, optical technology offers several advantages such as high space-bandwidth product, the possibility of including biometrics (face, fingerprint, iris, etc), and suitability for 2D data/images. Using the concept of holography, a three-dimensional (3D) object/scene can also be secured.

      For encoding data securely, optics offers several degrees of freedom such as amplitude, phase, wavelength, polarization, spatial frequency, and optical angular momentum to encode data securely. An intensity sensing device, such as a charge-coupled device (CCD) camera cannot record any phase information. It is possible to tuck away an optically based message in only one small section of a 2D array, a trick that forces unauthorized users to find the message’s position before they can begin to decode it. Therefore, it is believed that optical encryption techniques would provide a more complex environment and would be more resistant to attacks than purely electronic systems are.

      With the pioneering work on double random phase encoding (DRPE), optical techniques for information security have triggered much interest [1–3]. In the DRPE scheme, an input image to be encrypted is bonded with a random phase mask (RPM) and the product function is Fourier transformed. In the Fourier plane, another RPM is placed. Thus, the spectrum is again multiplied with the second RPM and its Fourier transformation is carried out, which gives a noisy image. This noisy image is called the encrypted image, which is a stationary white noise. Both the RPMs used in input and the Fourier plane are statistically independent and their values lie in the range [0,2π]. For decryption, the process is reversed and the conjugate of the RPMs at respective locations are used.

      An optical information system consists of light source, lenses, mirrors, beam splitters, detectors, display devices such as a spatial light modulator (SLM), and a CCD camera. These components can be arranged in various configurations to suit the type of desired optical information processing setup.

      Information in the form of a light wave passes through a converging lens that introduces delay or phase shift to the incident wavefront by an amount proportional to the thickness of lens, refractive index of the lens, and the wavelength of light. The light is distributed at the back focal plane of the lens, according to the spatial frequencies that are present in the original information. This spatial distribution in the back focal plane can be described mathematically as the Fourier transform of the input information. The Fourier transform capability of the converging lens is a crucial property of the optical information processors because it allows further manipulation of the optical information in the spatial frequency domain [4].

      Holograms have been used in credit cards, identity cards, monetary bills, and many other important documents for security purposes. With the rapid technological advancement in the computers, CCD technology, image processing hardware and software, printers, scanners and copiers, it is increasingly becoming possible to reproduce complex pictures, symbols, logos, etc. Therefore, it has now become possible to duplicate a holographic pattern. The publication of pioneering work on DRPE has broadened the research area of information security like encryption, authentication, watermarking, and hiding. The optical encryption techniques have been realized and have stimulated research in the information security areas [5–8].

      Pattern recognition is a science that concerns the description or classification of measurements. Pattern recognition techniques are often important components of intelligent systems and are used for both data processing and in decision-making. Opto-electronic techniques of pattern recognition for secure verification purposes are growing rapidly because of unique advantages offered by optical technologies [9]. Of late, the idea of authentication of credit cards, passports, driving license, and other personal identities has attracted much attention. The schemes use complex phase patterns that cannot be seen/copied by an intensity sensing device. Some security-enhanced optical security verification schemes have also been reported.

      Amongst the optical security techniques, encryption is an effective approach to ensure the protection of information (text, image, and video) from unauthorized use or access. To secure the stored information, it is required to encrypt the data. An unauthorized user cannot reveal the original data without knowledge of the exact random key