Shot-boundary detection: unraveled and resolved?
read more
Citations
Automatic soccer video analysis and summarization
A Survey on Visual Content-Based Video Indexing and Retrieval
A Formal Study of Shot Boundary Detection
Video shot boundary detection: Seven years of TRECVid activity
Information theory-based shot cut/fade detection and video summarization
References
Random variables and stochastic processes
Film Art: An Introduction
Automatic partitioning of full-motion video
Rapid scene analysis on compressed video
Related Papers (5)
Frequently Asked Questions (13)
Q2. Why did the authors select the dimensions of the blocks used in the block-matching procedure as 4?
Due to small frame dimensions in the resulting DC sequence, the authors selected the dimensions of the blocks used in the block-matching procedure as 4 4 pixels.
Q3. How can the authors compute the discontinuity value of a histogram?
If histograms are used as features, the discontinuity value can be obtained by bin-wise computing the difference between frame histograms.
Q4. How many differences were discarded to reduce the influence of motion and noise?
When computing the discontinuity as a sum of region-histogram differences, eight largest differences were discarded to efficiently reduce the influence of motion and noise.
Q5. What is the basis of detecting shot boundaries in video sequences?
The basis of detecting shot boundaries in video sequences is the fact that frames surrounding a boundary generally display a significant change in their visual contents.
Q6. What is the way to reduce the influence of extreme factors on the detection performance?
An effective way to reduce the influence of extreme factors on the detection performance is to embed additional information in the shot-boundary detector.
Q7. What are the advantages of the detector presented in this paper?
The facts that the detection method presented in this paper can operate on a wide range of video sequences without human supervision, and keep the constant high detection quality for each of them, are the major advantages the proposed detector has over the methods from recent literature.
Q8. How do the authors compute the discontinuity values for a shot?
For this purpose the authors compute the discontinuity values by compensating the motion between video frames using a blockmatching procedure similar to the one proposed in [23]
Q9. What is the probability function for the distribution in Fig. 8(a)?
The shape of the distribution in Fig. 8(a) indicates that a good analytic estimate for this distribution and so for the likelihood function can be found in the family of functions given as(19)Using the similar principle of global shape matching, the distribution in Fig. 8(b) and so the likelihood function can best be modeled using a Gaussian function(20)The most suitable parameter combinations and are then found experimentally, such that the rate of detection mistakes for the training sequences is minimized.
Q10. What is the simplest way of measuring the visual content discontinuity between two frames?
The simplest way of measuring the visual-content discontinuity between two frames is to compute the mean absolute change of intensity between the frames and for all frame pixels, i.e., for and , where andare the frame dimensions [12].
Q11. How can the authors support the decision about the presence of a dissolve?
the decision about the presence of a dissolve can be supported by investigating the behavior of the intensity variance in the “suspected” series of frames (e.g., those where pattern matching from Fig. 2 shows good results) and by checking how well this behavior fits the downwards-parabolic pattern.
Q12. How can the authors reduce the influence of a priori probability on the detection performance?
by properly modeling a priori probability and by securing its convergence to 0.5, the influence of this probability on the detection performance should be minimized as soon as a reasonable shot length is reached.
Q13. What can be done to reduce the influence of motion on discontinuity values?
while features and metrics can be found such that the influence of motion on discontinuity values is strongly reduced, the influence of strong and abrupt lighting changes on discontinuity values and thus also on the detection performance cannot be reduced that easily.