ABSTRACT
Due to digitalization, the information which was in structured form and handled by only technical people has now shifted in a larger way to unstructured form which can be used by all. This increased the use of multimedia elements like images and videos in number of applications. Adversely this gave rise to image and video forgery. The image forgery problem was found to be well addressed as a large number of research activities have being found in the literature. Currently video usage has been greatly enhanced in various fields due to multimedia systems. Unfortunately sufficient work was not found in the literature for video forgery detection. Thus this field remained unexposed. Working on these lines, this research work mainly focus on Video forgery detection with an enhanced feature of its elimination. Machine learning is used as the working technology alone with image processing to explore the motion parameters and to extract the required features. This research work also goes one step further to identify the forge object so as to enable its removal. This work is put forward as system design approach in which the methods implemented are evaluated qualitatively and quantitatively so as to provide for a graphical comparative analysis.
