VSCAN: An Enhanced Video Summarization Using Density-Based Spatial Clustering

Abstract

Abstract. In this paper, we present VSCAN, a novel approach for generating static video summaries. This approach is based on a modified DBSCAN clustering algorithm to summarize the video content utilizing both color and texture features of the video frames. The paper also introduces an enhanced evaluation method that depends on color and texture features. Video Summaries generated by VSCAN are compared with summaries generated by other approaches found in the literature and those created by users. Experimental results indicate that the video summaries generated by VSCAN have a higher quality than those generated by other approaches.

Tensor-Based Shot Boundary Detection in Video Streams

“…For the qualitative evaluation of performance of the method, we incorporated the methodology, as well as ground truth user summaries, in the VSUMM method, proposed by de Avila et al [2,22], as well as the VSCAN method presented in the work of Mahmoud et al [23,35]. In the former approach, the Comparison of User Summaries (CUS) measure is evaluated.…”

“…However, as presented by Mahmoud et al, more informative parameters are precision P and recall R, since they also convey information on the keyframes present in AS and not present in US or vice versa. As a trade-off of the two, the so-called F measure is used [20,35,45]. Thus, in our experiments, we used the measures defined as follows:…”

“…The other video summarization method, called VSCAN, was proposed by Mahmoud et al [35]. Their method is based on a modified density-based spatial clustering of applications with noise-clustering method (DBSCAN), which is used to video summarization from color and texture features of each frame.…”

“…Nevertheless, if there is additional knowledge of a signal type, then methods that compute specific features can be more discriminative but only for this specific type of a signal. This happens with color histograms, frequently used in video shot detection [2,7,35]. Such methods will not work with monochrome or hyperspectral images, though.…”

“…These methods, in turn, rely on efficient algorithms of boundary detection in the visual streams. Basically, the summarization methods are divided into static and dynamic ones [33,35]. In the former, a set of consecutive frames with sufficiently coherent contents is represented by a single representative frame, called a key frame.…”