Shot Change Detection via Local Keypoint Matching

Abstract

Abstract-Shot change detection is an essential step in video content analysis. However, automatic shot change detection often suffers from high false detection rates due to camera or object movements. To solve this problem, we propose an approach based on local keypoint matching of video frames. This approach aims to detect both abrupt and gradual transitions between shots without modeling different kinds of transitions. Our experiment results show that the proposed algorithm is effective for most kinds of shot changes.

Temporal segmentation tool for high-quality real-time video editing software

“…• On the other hand, we have also compared our strategy with two algorithms that analyze the motion along the sequences: the first one [32] using salient feature matching (SFM) and the second one [21] using exhaustive block matching (EBM). Since SAD and CH only detect abrupt transitions, we have compared them not only with the complete proposed strategy (CPS), but with a light version of it that only detects abrupt transitions (ATD).…”

Fuzzy color histogram-based video segmentation

“…Huang et al [2] propose an approach based on local keypoint matching of video frames to detect abrupt and gradual transitions. By matching the same objects and scenes using contrast context histogram (CCH) in two adjacent frames, the method decides that there is no shot change.…”

“…iv Local keypoint matching (KM): Recognizing the objects and scenes throughout the video is the basic idea of the keypoint matching-based shot-boundary detection methods. The algorithm proposed in [2] matches the objects between consecutive frames, and determines if there is a shot boundary. We use scale invariant feature transform [31] and a simple matching algorithm for this purpose.…”

“…Various shot-boundary detection algorithms have been proposed [1][2][3][4][5][6] and compared [7][8][9][10]; however, to the best of our knowledge, no shot-boundary detection algorithm specialized for CBCD is found in the literature. Our aim is to propose an automatic shot-boundary detection algorithm for the videos on which various transformations are applied.…”

Video scene detection by link-constrained affinity-propagation

“…Since a video scene is a collection of semantically related shots, we first divide a video into shots by using the CCH feature based method [14]. The shot change detection algorithm is summarized as follows: 1) Construct the CCH keypoints of each frame.…”