الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 92 |  |  |
| | | from | 92 | | |
Abstract
In modern technologies, digital image processing is an essential field with various applications. Scientists are looking for advanced processing tools such as embedded and special hardware systems for big data processing in real-time every day.Raspberry Pi is a very useful and promising tool for image processing applications that provide the advantages of portability, parallelism, lowcost, and lowerpower consumption. Raspberry Pi can be used in various image processing applications in different fields such as medicine & healthcare, agriculture, intelligent transportation, education, security, and smart homes.
Internet of Things (IoT) is the communication of everything with anything else, with the primary goal of data transfer over a network. Raspberry Piis a quad-core computer with parallel processing capabilities that may be used to speed up computations and processes that making it ideal for IoT applications. Raspberry Pi is employed in IoT applications designed to accomplish many of the same tasks as a normal desktop computer.
Image authentication techniques have recently received much attention for protecting images against unauthorized access. As a result of the wide use of the internet nowadays, the need to ensure data integrity and authenticationhas increased.Since digital images can be quickly changed and the alteration is difficult to identify, image security and authentication have become a more important topic.Sensitive images must be shielded from manipulation attempts.Digital image watermarking technology is used to secure and ensure digital images’ copyright by embedding hidden information that proves its copyright.
This thesis presents a watermarking algorithm for securing transmitted color images using Quaternion Legendre-Fourier Moments (QLFMs)on a multi-core Raspberry Pi. The Raspberry Pi sends the secured image to the receiver side tocheck the received image integrity to ensure the authentication and that no alteration is done on the image.A Parallel Robust watermarking algorithm for color images using Quaternion Legendre-Fourier Moment (QLFM) in polar coordinates is implemented on Raspberry Pi (RPi) platform with parallel computing and C++ programming language.We can combine many Raspberry Pi’s into a ‘cluster’ for high-performance computation. Message Passing Interface (MPI) and OpenMP for parallel programming are used to accelerate the execution time for the color image watermarking algorithm implemented on the Raspberry Pi cluster.
Also, in this thesis, a zero watermarking method provides copyright protection for the transmitted color images using multi-channel orthogonal Legendre Fourier moments of fractional orders (MFrLFMs) AES-CBC encryption algorithm on Raspberry Pi. The Raspberry Pi can be mounted on a drone. A linked camera captures photos, which the Raspberry Pi uses to apply a zero watermarking approach to produce an ownership verification key in real-time. Before sending the ownership verification key and the original image to the monitoring station, the Raspberry Pi encrypts them with AES. Next, the monitoring station verifies the received image’s integrity to confirm the authenticity and that the image does not tamper.
A new parallel Quaternion Legendre Fourier Moment (QLFM) watermarking algorithm for color images is implemented on Raspberry Pi cluster of 4 nodes. We evaluated the output of the watermarked image in terms of visual imperceptibility andwatermark robustness of the QLFM watermarking algorithm. The peak signal calculates theinvisibility of the watermark to noise ratio.Structural similarity image index is computed to measure the watermarked image quality and similarity.Bit error Rate is also computed to evaluate the watermark robustness of the watermarking algorithm. Watermark reconstruction accuracy is computed and also a comparison between QLFMs and other quaternion moments that reveal the efficiency of QLFMs. The improvement ratio of parallelizing the watermarking algorithm on 4 Raspberry Pi’s cluster using all cores is 93.21 % of moment computation step and 91.59 % for reconstruction of the watermarking step for image size of 512X512 and moment order 40.
A zero watermarking algorithm is also implemented on Raspberry Pi and Advanced Encryption Standard (AES) is also used. Several experiments were done to evaluate the efficiency of the zero watermarking algorithm with AES encryption as: execution time, information Entropy, histogram analysis, correlation analysis and key sensitive analysis.