Statistical framework for effective retrieval of images based on content

Abstract

The new digital technology updates have resulted in huge image capture capabilities using multiple resolution techniques. However, this has led to a disadvantage with respect to storage and recovery efficiencies. To address this problem, content-based image retrieval (CBIR) has been coined and has become the core consideration for the effective recovery of massive data sets. With recovery competencies, CBIR has been used in many applications ranging from medical processing, video and audio recovery and recognition of old documents. In this article, we present a recovery model based on the Bivariable Gamma Blending Model with a perspective on the application of video retrievals from YouTube video conferences considered as a data source. The main advantage of this model is that the recovery of a relevant conference / video clip can be easily recovered, so that users can have their choice of reference within a very short duration. The efficiency of the model is tested using benchmark quality metrics such as the signal-to-noise ratio (SNR), the mean square error (MSE) and the structural similarity index (SSIR) ratio.