Content Based Video Retrieval in Transformed Domain using Fractional Coefficients

摘要

With the development of multimedia and growing database there is huge demand of video retrievalsystems. Due to this, there is a shift from text based retrieval systems to content based retrievalsystems. Selection of extracted features play an important role in content based video retrieval. Goodfeatures selection also allows the time and space costs of the retrieval process to be reduced.Different methods[1,2,3] have been proposed to develop video retrievals systems to achieve betterperformance in terms of accuracy.The proposed technique uses transforms to extract the features. The used transforms are DiscreteCosine, Walsh, Haar, Kekre, Discrete Sine, Slant and Discrete Hartley transforms. The benefit ofenergy compaction of transforms in higher coefficients is taken to reduce the feature vector size bytaking fractional coefficients[5] of transformed frames of video. Smaller feature vector size results inless time for comparison of feature vectors resulting in faster retrieval of images. The feature vectorsare extracted and coefficients sets are considered as feature vectors (100%, 6.25%, 3.125%,1.5625%, 0.7813%, 0.39%, 0.195%, 0.097%, 0.048%, 0.024%, 0.012%, 0.006% and 0.003% ofcomplete transformed coefficients). The database consists of 500 videos spread across 10categories.