Abstract: Exemplary embodiments are described in which is performed not only a shot detection (continuous recording with a camera) and an association of several key-frames to the shots, it then being possible for a subsequent scene recognition to be based on the grouping of shots into scenes. Rather, it is observed that a scene only relates to one event in a setting. Since both can change within a shot, not every scene boundary is at the same time also a shot boundary. In addition, not every shot is short enough, so that a reliable retrieval of different picture contents is not guaranteed. Therefore, exemplary embodiments are shown which are capable of defining sub-shots so that in principle, scene and shot boundaries are also sub-shot boundaries at the same time. Sub-shots furthermore include only video pictures with a small change in picture content.

Abstract: A method and a device are for an automated comparison of at least two sets of measuring values. The measuring values of the two sets are assigned respectively to one class from a finite number of classes defined by indices so that a frequency distribution is defined respectively for each of the two sets, which frequency distribution indicates for each class a frequency of the measuring values assigned to this class. A distance measure reflecting a similarity or dissimilarity between the two sets of measuring values between these frequency distributions is calculated as a function of a final value of a first auxiliary value termed here match by way of example. The first auxiliary value match is calculated by an algorithm using two sets of variables in that, with a given maximum distance dmax?1 for all integral distances d with 0?d?dmax, beginning with d=0 and continuing to larger distances d, respectively for all indices i and j at a distance from each other by the distance d.

Abstract: The automated detection of scene boundaries supports the user in browsing a video. Establishing the similarity between two respective frame sequences is made by choosing and comparing respective key-frames from each frame sequence, which are temporally as far apart from each other as possible but lie outside gradual transitions. Additionally or alternatively, film grammar- and probability-based taking into account of all types of transitions with respect to their separating and/or merging effect on the frame sequences surrounding the transitions and a differently weighted incorporation of the effects of the different types of transitions into uniting frame sequences k into scenes is provided. Thus, all types of transitions are incorporated in determining frame sequence boundaries as scene boundaries with a corresponding weighting, both their separating and merging effect is taken into account, and this knowledge is introduced on a probability basis into the decision in existing scene detection methods.

Abstract: A method and a device are for an automated comparison of at least two sets of measuring values. The measuring values of the two sets are assigned respectively to one class from a finite number of classes defined by indices so that a frequency distribution is defined respectively for each of the two sets, which frequency distribution indicates for each class a frequency of the measuring values assigned to this class. A distance measure reflecting a similarity or dissimilarity between the two sets of measuring values between these frequency distributions is calculated as a function of a final value of a first auxiliary value termed here match by way of example. The first auxiliary value match is calculated by an algorithm using two sets of variables in that, with a given maximum distance dmax?1 for all integral distances d with 0?d?dmax, beginning with d=0 and continuing to larger distances d, respectively for all indices i and j at a distance from each other by the distance d.

Abstract: The automated detection of scene boundaries supports the user in browsing a video. Establishing the similarity between two respective frame sequences is made by choosing and comparing respective key-frames from each frame sequence, which are temporally as far apart from each other as possible but lie outside gradual transitions. Additionally or alternatively, film grammar- and probability-based taking into account of all types of transitions with respect to their separating and/or merging effect on the frame sequences surrounding the transitions and a differently weighted incorporation of the effects of the different types of transitions into uniting frame sequences k into scenes is provided. Thus, all types of transitions are incorporated in determining frame sequence boundaries as scene boundaries with a corresponding weighting, both their separating and merging effect is taken into account, and this knowledge is introduced on a probability basis into the decision in existing scene detection methods.

Abstract: Exemplary embodiments are described in which is performed not only a shot detection (continuous recording with a camera) and an association of several key-frames to the shots, it then being possible for a subsequent scene recognition to be based on the grouping of shots into scenes. Rather, it is observed that a scene only relates to one event in a setting. Since both can change within a shot, not every scene boundary is at the same time also a shot boundary. In addition, not every shot is short enough, so that a reliable retrieval of different picture contents is not guaranteed. Therefore, exemplary embodiments are shown which are capable of defining sub-shots so that in principle, scene and shot boundaries are also sub-shot boundaries at the same time. Sub-shots furthermore include only video pictures with a small change in picture content.