Video Shot Boundary Detection

Developed an Automatic Shot Boundary Detection System which segregates a given input video into its constituent shots by identifying boundary frames between them. * Proposed a Dual-Step method for identification of shot boundaries with adaptable thresholds tailored to different video genres. * Employed CNN and Canny Edge Detector to recognize Abrupt Transitions. * Fuzzy Entropy, followed by CNN, is employed for identifying Gradual Transitions.