Video segmentation is the primary step used to analyse the video and extract the meaningful data out of it, making it easier to index based on the content of the video rather than just indexing the video based on the title and manual description. Shot boundary detection is one of the key step involved in the process of analysing the video and segment it into different parts. Videos acts the long form data item with a lot of data incorporated. Analysis of these long form data items can be done with a series of different techniques such as the frame similarity, object based video segmentation, shot boundary detection, content based video indexing. CNN is used to enhance the existing methods of this process as it gives the best way to extract the features of the each frame which accumulates in tailoring the frames to find the meaning. Frame similarity is calculated based on the features of each frame compared with its nearby frames by the extracting the features using CNN. This similarity calculated is used to make detailed summary of the video and helps in implementing the content based indexing.