Content-based access to video objects: Temporal Segmentation, visual summarization, and feature extraction

Günsel, Bilge; Tekalp, A. Murat; Beek, P.J.L. van

doi:10.1016/s0165-1684(98)00010-3

Cited by 47 publications

(28 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Another commonly used method of object delineation is based on the intensity gradient [3], in which the points with the highest gradient in a local neighborhood are selected as the feature points corresponding to edges of the object. Generally, both the GWT based and intensity gradient based selection methods can be modeled as finding the local maximums of a significance measure function.…”

Section: Local Feature Point Selectionmentioning

confidence: 99%

“…Each node is shifted randomly to find its best matching position. In [3], [4], every feature point is examined independently and no global placement information of features points is used. Consequently the object's structural information may be lost.…”

Section: Searchingmentioning

confidence: 99%

“…In content-based video compression, in order to save bits representing the shape and motion vectors of an object, two-dimensional (2-D) or three-dimensional (3-D) mesh based algorithms have been used [1], [3], [4], [6]. The idea is to represent an object by a set of nodes and connecting line segments which usually form triangles.…”

mentioning

confidence: 99%

“…The amplitudes of the GWT coefficients instead of the local intensity [1], [3], [4] are used to represent local features. A nonuniform 2-D mesh instead of a uniform grid in [22], [23] is used to represent the global structure.…”

mentioning

confidence: 99%

See 3 more Smart Citations

Object tracking using the Gabor wavelet transform and the golden section algorithm

Zheng

Ahalt

2002

IEEE Trans. Multimedia

View full text Add to dashboard Cite

Abstract-This paper presents an object tracking method for object-based video processing which uses a two-dimensional (2-D) Gabor wavelet transform (GWT) and a 2-D golden section algorithm. An object in the current frame is modeled by local features from a number of the selected feature points, and the global placement of these feature points. The feature points are stochastically selected based on the energy of their GWT coefficients. Points with higher energy have a higher probability of being selected since they are visually more important. The amplitudes of the GWT coefficients of a feature point are then used as the local feature. This takes advantage of the characteristics of Gabor wavelets which are highly localized in both the time and the frequency domains. The global placement of the feature points is determined by a 2-D mesh whose feature is the area of the triangles formed by the feature points. In this way, a local feature is represented by a GWT coefficient amplitude vector, and a global feature is represented by a triangle area vector. One advantage of the 2-D mesh is that the direction of its triangle area vector is invariant to affine transform. Consequently, the similarity between two local features or two global features can be defined as a function of the angle and the length ratio between two vectors, and the overall similarity between two objects is a weighted sum of the local and global similarities. In order to find the corresponding object in the next frame, the 2-D golden section algorithm is employed, and this can be shown to be the fastest algorithm to find the maximum of a unimodal function. Our results show that the method is robust to object deformation and supports object tracking in noisy video sequences.Index Terms-Content-based video, feature points, Gabor wavelets, golden section, mesh, object-based video, object tracking.

show abstract

Section: Local Feature Point Selectionmentioning

confidence: 99%

Section: Searchingmentioning

confidence: 99%

mentioning

confidence: 99%

mentioning

confidence: 99%

See 2 more Smart Citations

Object tracking using the Gabor wavelet transform and the golden section algorithm

Zheng

Ahalt

2002

IEEE Trans. Multimedia

View full text Add to dashboard Cite

show abstract

“…Günsel et al [6] developed an objectbased indexing of video filmed by a single camera, dealing with the motion and shape properties of the viewed objects and considering the camera motion and the object trajectories and interactions. A work by Hervieu et al [7,8] proposed a HMM-based shot classification method and rare event detection using the mobile object trajectories that may be used after a shot segmentation and a tracking processing.…”

Section: Introductionmentioning

confidence: 99%

Activity-based temporal segmentation for videos of interacting objects using invariant trajectory features

Hervieu

Bouthémy

Cadre

2008

2008 15th IEEE International Conference on Image Processing

View full text Add to dashboard Cite

This paper presents a content-based approach for temporal segmentation of videos. Tracked objects are characterized by their 2D trajectories which are used in a meaningful way to model visual semantics, i.e., the observed single video object activities and their interactions. To this end, hierarchical Semi-Markov Chains (SMCs) are computed in order to take into account the temporal causalities of object motions. Object movements are characterized using local invariant features computed from the curvature and velocity values while interactions are represented by the temporal evolution of the distance between objects. We have evaluated our method on squash video sequences, and have favorably compared with other methods including Hidden Markov Models (HMMs).

show abstract

Video summarization based on the psychological unfolding of a drama

Moriyama

Sakauchi

2002

Systems & Computers in Japan

View full text Add to dashboard Cite

SUMMARYThis paper proposes a method of composing digests of TV videos based on their psychological unfolding. Most of the studies of video summarization to date have been based on the video structure, such as cuts or objects in the video. It is generally considered, however, that the observer of a TV drama emphasizes instead the psychological unfolding of the video content (such as the climax). However, no method of summarizing a drama based on the psychological aspect of the video has been proposed. Track structures [such as cutting, lines spoken by the actors, BGM (background music), and sound effects] constitute physical components, and the dramatist skillfully manipulates the time structure based on his empirical knowledge in order to present a particular psychological impression. This paper proposes a method in which the psychologically important part is detected on the basis of empirical knowledge as a corresponding time-series pattern of the track structure, and a summarized video is generated. As an experiment, the proposed method was applied to a TV drama, and the time length was reduced to approximately 1/14. Subjective evaluation experiments show that the proposed method preserves the content of the main plot, with naturalness of extraction, thus indicating the usefulness of the proposed method.

show abstract

Content-based access to video objects: Temporal Segmentation, visual summarization, and feature extraction

Cited by 47 publications

References 30 publications

Object tracking using the Gabor wavelet transform and the golden section algorithm

Object tracking using the Gabor wavelet transform and the golden section algorithm

Activity-based temporal segmentation for videos of interacting objects using invariant trajectory features

Video summarization based on the psychological unfolding of a drama

Contact Info

Product

Resources

About