An Innovative Technique for Adaptive Video Summarization
05 Aug 2011-pp 592-600
TL;DR: This paper presents an innovative video summarization technique based on inter-frame information variation which makes the algorithm adaptive in respect to the information content of the source video.
Abstract: Video summarization is a procedure to reduce the size of the original video without affecting vital information presented by the video. This paper presents an innovative video summarization technique based on inter-frame information variation. Similar group of frames are identified based on inter-frame information similarity. Key frames of a group are selected using disturbance ratio (DR), which is derived by measuring the ratio of information changes between consecutive frames of a group. The frames in the summarized video are selected by considering continuation in understanding the message carried out by the video. Higher priority is given to the frames which have higher information changes, and no-repetition to reduce the redundant areas in the summarized video. The higher information changes in the video frames are detected based on the DR measure of the group and this makes our algorithm adaptive in respect to the information content of the source video. The results show the effectiveness of the proposed technique compared to the related research works.
Citations
More filters
04 Jun 2012
TL;DR: This proposed research work minimizes the time required for processing each of the video frames firstly, by reducing their effective size, and then it is followed by an efficient technique for generating the summarized video.
Abstract: Video summarization plays a very significant role in navigating a video, to understand its information or to search the required event information. Our proposed research work minimizes the time required for processing each of the video frames firstly, by reducing their effective size, and then it is followed by an efficient technique for generating the summarized video. The information contained in a frame is extracted using object and motion based features where the object based feature helps to evaluate the importance of the given frame compared to its neighboring frames and the motion based feature helps to estimate the dynamism of the frame. Disturbance Ratio [DR] based measurement is used in the next step to select the shot
Cites background or methods or result from "An Innovative Technique for Adaptiv..."
...This is because, our proposed technique works on the motion and the object based features that provide reasonable better information of dynamism and presence of objects in the frame instead of entropy based information which we had proposed earlier [5]....
[...]
...PM PM[5] PM PM[5] PM PM[5] Vid_1 7 7 18 28 90 88 Vid_2 5 6 15 16 47 38 Some essential properties which are needed to be included in the video summarization to prove its efficiency are shown in Table 3 and a comparison with the related research works [3] is also presented....
[...]
...Table 2 includes a comparative study between our proposed approach and previous approach in [5]....
[...]
...Video Summarization is divided into two broad categories (1) Domain Specific [2, 11, 16, 17] and (2) Non Domain Specific [1, 2, 3, 5, 4, 12, 13, 14, 15]....
[...]
...In this work we have extended our previous work [5] in terms of reduction in the processing time for generating the summarized video....
[...]
References
More filters
TL;DR: It is shown that recognition of geometrical patterns and alphabetical characters independently of position, size and orientation can be accomplished and it is indicated that generalization is possible to include invariance with parallel projection.
Abstract: In this paper a theory of two-dimensional moment invariants for planar geometric figures is presented. A fundamental theorem is established to relate such moment invariants to the well-known algebraic invariants. Complete systems of moment invariants under translation, similitude and orthogonal transformations are derived. Some moment invariants under general two-dimensional linear transformations are also included. Both theoretical formulation and practical models of visual pattern recognition based upon these moment invariants are discussed. A simple simulation program together with its performance are also presented. It is shown that recognition of geometrical patterns and alphabetical characters independently of position, size and orientation can be accomplished. It is also indicated that generalization is possible to include invariance with parallel projection.
7,963Â citations
TL;DR: The proposed framework includes some novel low-level processing algorithms, such as dominant color region detection, robust shot boundary detection, and shot classification, as well as some higher-level algorithms for goal detection, referee detection,and penalty-box detection.
Abstract: We propose a fully automatic and computationally efficient framework for analysis and summarization of soccer videos using cinematic and object-based features. The proposed framework includes some novel low-level processing algorithms, such as dominant color region detection, robust shot boundary detection, and shot classification, as well as some higher-level algorithms for goal detection, referee detection, and penalty-box detection. The system can output three types of summaries: i) all slow-motion segments in a game; ii) all goals in a game; iii) slow-motion segments classified according to object-based features. The first two types of summaries are based on cinematic features only for speedy processing, while the summaries of the last type contain higher-level semantics. The proposed framework is efficient, effective, and robust. It is efficient in the sense that there is no need to compute object-based features when cinematic features are sufficient for the detection of certain events, e.g., goals in soccer. It is effective in the sense that the framework can also employ object-based features when needed to increase accuracy (at the expense of more computation). The efficiency, effectiveness, and robustness of the proposed framework are demonstrated over a large data set, consisting of more than 13 hours of soccer video, captured in different countries and under different conditions.
943Â citations
TL;DR: It is argued that video summarisation would benefit from greater incorporation of external information, particularly user based information that is unobtrusively sourced, in order to overcome longstanding challenges such as the semantic gap and providing video summaries that have greater relevance to individual users.
Abstract: Video summaries provide condensed and succinct representations of the content of a video stream through a combination of still images, video segments, graphical representations and textual descriptors. This paper presents a conceptual framework for video summarisation derived from the research literature and used as a means for surveying the research literature. The framework distinguishes between video summarisation techniques (the methods used to process content from a source video stream to achieve a summarisation of that stream) and video summaries (outputs of video summarisation techniques). Video summarisation techniques are considered within three broad categories: internal (analyse information sourced directly from the video stream), external (analyse information not sourced directly from the video stream) and hybrid (analyse a combination of internal and external information). Video summaries are considered as a function of the type of content they are derived from (object, event, perception or feature based) and the functionality offered to the user for their consumption (interactive or static, personalised or generic). It is argued that video summarisation would benefit from greater incorporation of external information, particularly user based information that is unobtrusively sourced, in order to overcome longstanding challenges such as the semantic gap and providing video summaries that have greater relevance to individual users.
468Â citations
TL;DR: It is demonstrated that the method detects both fades and abrupt cuts with high accuracy and it is shown that it captures satisfactorily the visual content of the shot.
Abstract: New methods for detecting shot boundaries in video sequences and for extracting key frames using metrics based on information theory are proposed. The method for shot boundary detection relies on the mutual information (MI) and the joint entropy (JE) between the frames. It can detect cuts, fade-ins and fade-outs. The detection technique was tested on the TRECVID2003 video test set having different types of shots and containing significant object and camera motion inside the shots. It is demonstrated that the method detects both fades and abrupt cuts with high accuracy. The information theory measure provides us with better results because it exploits the inter-frame information in a more compact way than frame subtraction. It was also successfully compared to other methods published in literature. The method for key frame extraction uses MI as well. We show that it captures satisfactorily the visual content of the shot.
311Â citations
TL;DR: The algorithm, which escapes the complexity of existing methods based, for example, on clustering or optimization strategies, dynamically and rapidly selects a variable number of key frames within each sequence by analyzing the differences between two consecutive frames of a video sequence.
Abstract: Video summarization, aimed at reducing the amount of data that must be examined in order to retrieve the information desired from information in a video, is an essential task in video analysis and indexing applications. We propose an innovative approach for the selection of representative (key) frames of a video sequence for video summarization. By analyzing the differences between two consecutive frames of a video sequence, the algorithm determines the complexity of the sequence in terms of changes in the visual content expressed by different frame descriptors. The algorithm, which escapes the complexity of existing methods based, for example, on clustering or optimization strategies, dynamically and rapidly selects a variable number of key frames within each sequence. The key frames are extracted by detecting curvature points within the curve of the cumulative frame differences. Another advantage is that it can extract the key frames on the fly: curvature points can be determined while computing the frame differences and the key frames can be extracted as soon as a second high curvature point has been detected. We compare the performance of this algorithm with that of other key frame extraction algorithms based on different approaches. The summaries obtained have been objectively evaluated by three quality measures: the Fidelity measure, the Shot Reconstruction Degree measure and the Compression Ratio measure.
175Â citations