Please use this identifier to cite or link to this item:
|Title:||Image Retrieval Using Dither Block Truncation Coding with Advanced Similarity Measurement Algorithm|
|Keywords:||Engineering and Technology,Computer Science,Imaging Science and Photographic Technology|
|Abstract:||Content based image and video (Sequence of images) retrieval has been progressively newlinemore used to describe the way toward retrieving the required images or pictures and videos newlinefrom a huge collection of dataset based on features that are separated from the sequence of newlineimages. Several methods has been developed in which image retrieval using different newlinefeatures, there are also different methods by which we can retrieve videos also. Videos are newlinenothing but just collection of images but in a sequential way or we can say that if we want to newlineconvert the videos into the image form then by using some of the methods we can convert newlinevideo into the set of sequence of images, now after getting the images it will be easy for us to newlinedo work on it. As we have seen that day by day the data and the information is increasing newlinevery rapidly, so due to this the image retrieval and the video retrieval methods becoming very newlinetime consuming. So for this we need some mechanism by which we can reduce the newlinecomputational time. This can be done in two phases first of all we need to compress the data newlineand after compressing we can retrieve the data. newlineTherefore, this thesis uses a technique for content based video retrieval based on newlineimproved version of block truncation coding, i.e. ordered dither block truncation coding. As newlinewe know that in the previous method the computational time was high, complexity was also newlinehigh and the quality was not good these were the problems of block truncation method. newlineInitially CBVR contains three steps: key frame extraction, feature extraction and video newlineretrieval based on the similarity measurement. In this thesis, two image features are extracted newlinefrom the video key frames namely block YCbCr color feature, and block bit feature from newlineODBTC. These two highlights are utilized to register the comparability measures of video newlinekey frames for content based video recovery applications. This method is not only used as a newlinecompression method but also used as a retrieval method. The execution of the proposed newlinetechnique has been assessed th|
|Appears in Departments:||Faculty of Engineering and Technology|
Files in This Item:
Items in Shodhganga are protected by copyright, with all rights reserved, unless otherwise indicated.