A shot clustering method based on spectral segmentation theory

Abstract

The invention relates to a shot clustering method based on the spectrum segmentation theory, which comprises the following steps: utilizing, the spectrum segmentation theory for shot clustering; extracting feature vectors of each unspecified shot; calculating similarity between each two categories according to the extracted feature vectors; then constituting each shot cluster as a weighted undirected graph; segmenting each shot category into two shot categories by a using spectrum according to the similarity between each two categories; using Bayesian information criteria to judge whether the segmentation is effective or not, the effectively segmented shot sub-categories are iteratively segmented, the ineffectively segmented shot categories are terminals; finally syncretizing the classification results after the segmentation to get the optimal shot classification number and the classification result. The invention solves the difficult problem that the optimized classification number is difficult to estimate in the clustering algorithm, and improves the recall ratio and the pertinency ratio of the clustering result by utilizing the precise classification spectrum segmentation; the proposed overall fusion operation has a function of correcting the classification errors, thereby effectively avoiding the problem of local optimum relation.

Description

technical field [0001] The invention belongs to the field of video content analysis and retrieval, and in particular relates to a method for clustering shots. Background technique [0002] A video shot refers to a piece of semantically uninterrupted video content, which is the basic structure and semantic unit of video information retrieval. Clustering these units representing video semantics is the basis of video semantic analysis. Current clustering algorithms can be broadly classified into supervised and unsupervised. Supervised clustering trains the classifier through a given sample set, and the classification is accurate, but the sample set needs to be manually labeled. The unsupervised clustering algorithm has the function of self-learning and does not need training samples, but it is difficult to determine the optimal number of classifications, and the classification results are sensitive to the initial division. [0003] In recent years, there has been a lot of res...

AI Technical Summary

Problems solved by technology

[0005] The technical problem to be solved by the present invention is to overcome the deficiencies of the prior art and provide a shot clustering method based on spectral segmentation theory, which can estimate the number of optimal classifications that are difficult to estimate in the clustering algorithm under the condition of low complexity, Spectral segmentation using accurate binary classification improves the recall and precision of clustering results; the proposed global fusion operation has the function of error correction for classification errors, avoiding the problem of local optimal solutions

Images