Digital video shot boundary detector investigation

Abstract

Many algorithms have been proposed and evaluated for the detection of shot boundaries in digital video. Although these techniques have been verified, there remains a lack of standardised data to classify which techniques are best suited for certain applications. The Jensen-Shannon divergence (JSD) is one such technique used for shot boundary detection. In this article the JSD technique was adapted to handle monochromatic and RGB videos. This adaptation made it possible for the JSD technique to be evaluated in the RGB and monochromatic (grayscale) color spaces as well as the effect of video resizing in terms of recall, precision and execution time