CHAPTER 7: CONCLUSION AND FUTURE ENHANCEMENT
The core intention of this research is to facilitate various multimedia documents, such as image and video, with security and copyright protection. Digital watermarking acts as an aid for achieving this ambition. Besides, the proposed research work fulfills the following objectives:
Development of generic architecture for image and video watermarking technique;
Development of image watermarking using DWT and fuzzy-neuro technique and compare their performance;
Development of video watermarking using DWT and fuzzy-neuro technique and compare their performance;
Development of watermark recognition using chain code and support vector machine to classify the watermark;
Different image processing threats are applied on watermarked data to verify the quality of watermarked data;
Various image and video watermarking techniques are observed and scrutinized, besides, their limitations being analyzed;
Study and analysis of how various artificial intelligence techniques are applicable in the domain of image and video watermarking;
Different feature extraction and pattern recognition techniques are studied to get highest matching ratio of extracted watermark;
The strengths and weaknesses of current watermarking techniques are studied and new application which tries to solve the limitations has been developed;
The working of the proposed research work and experimental system is presented; and
The above mentioned research work is published in different national and international journals.
Implementation of image and video watermarking techniques in frequency domain has harnessed the accomplishment of this research work. As concerns security, biometric fingerprint has been taken as watermark in order to provide more security to track authorized or prohibited copies and detect the instigator of illegitimately made copies. For checking the integrity of retrieved watermark, watermark recognition is designed and developed using chain code and support vector machine. Furthermore, in this research work, diverse existing digital watermarking methods are studied and efforts have been made to improve of their performance in terms of robustness, security features and imperceptibility. Proposed generic architecture for image and video watermarking, as well as sub architectures of particular techniques are also presented in this research work. Moreover, algorithms for DWT based alpha blending technique and fuzzy-nuero based watermarking, for both image and video, are developed. Thereafter, Obtained results are compared with existing methods. With the observation of results of existing and proposed watermarking techniques, the performances of proposed watermarking techniques are quantified by the following chart shown in figure 7.1.
Figure 7.1: Performance Analysis
From this chart, it is evident that among all the other watermarking techniques, fuzzy-neuro watermarking technique provides higher PSNR value. The watermarked data that is more robust and precise against various image processing attacks is generated using fuzzy-nuero watermarking technique. In this research, different image processing threats such as noise, adjust, rotate, cropping and scaling are applied on watermarked data, yet, the quality of image remains unaffected and intact.
Besides this, a very significant segment has been occupied by digital watermarking in various domains, such as medical, education, media, e-commerce and e-governance, crime scene analysis etc. In the arena of data security, watermarks may be used for certification, authentication, and conditional access. In this day and age, certification has become a solemn issue regarding medical images, commercial photographs and videos, sensitive documents such as patents, artwork, and computer generated images, identity cards, passports, driving license etc. Substantially, providing authentication digital watermarking is the finest approach. Watermarks are used to stain white paper with the motive to authenticate the inventor and verify the authenticity of the document content. Such applications are particularly employed for security of official documents such as contracts, vills, memorandums, agreements etc. Digital watermarks, in this digital world, have benefited many e-governance and e-commerce applications used for intellectual property protection and ownership of the digital content. Digital watermarking, also, holds great influence on content creators such as artists, authors, and movie studios, and content providers like photo stock archives, libraries, and professional photographers, electronic commerce and graphics software vendors and manufacturers of digital still images, video cameras and digital video discs (DVDs).While crime scene analysis, crime scene photography is eminent, as photographs can be used as evidence in the court of law. So, providing security and authenticity to the captured photographs is an acute requirement, which is fulfilled by image watermarking. Also, in education sector, images, videos, documents require authenticity of the owner, just like in e-learning digital data requires copyright protection. Thus, image and video watermarking is an apt solution for providing authenticated access.
Farther, the mentioned research work may be extended by using another biometric data like iris, face, palm prints. PCA or SVD based DWT watermarking techniques may be applied for text and audio. Furthermore, other computational intelligence techniques like genetic algorithm may also be applied along with digital watermarking.