Video Quality Analysis for Concert Video Mashup Generation - Advanced Concepts for Intelligent Vision Systems - page 11

7 7

6 6

5 5

4 4

3 3

2 2

1 1

C1 C2 C3 All C1 C2 C3 All

(a) (b)

Fig. 7. Mean scores for (a)camera stability and (b) image quality. The methods are
represented as: random (

). The horizontal axis
represents the mashups corresponding to the three concerts C1-C3. “All” represents
the mean across all the concerts. Error bars show confidence intervals of 95% of the
mean value.

), manual ( ♦ ), and quality-based (

∗

are comparable. However, if the concert recordings contain low image quality
the mashup quality cannot be improved.

4Con lu on

This paper describes a method for evaluating signal quality of concert video
recordings captured by the audience. We measure the blockiness, blurriness,
brightness and shakiness of the video frames and then combine them into an
overall multiplicative measure for the quality of a video segment. The method is
applied to automatically select high-quality segments in a mashup from multiple-
camera recordings. We compare the quality of these mashups against the mashups
created by a random selection of segments and by a professional video-editor.
The subjective evaluation shows that the perceived quality of our mashups is
comparable to the mashups created by the professional video-editor and signif-
icantly higher than the mashups generated randomly. Further analysis shows
that the mashup quality depends not only on the methods used on generating
them but also on the recording quality of the concert videos.

References

1. Shrestha, P., Weda, H., Barbieri, M., Sekulovski, D.: Synchronization of multiple
camera videos using audio-visual features. IEEE Trans. on Multimedia 12(1), 79-92
(2010)
2. Yan, W., Kankanhalli, M.S.: Detection and removal of lighting & shaking artifacts
in home videos. In: Proc. of the 10 th ACM Int. Conf. on Multimedia, pp. 107-116
(2002)
3. Li, X.: Blind measurement of blocking artifacts in images. In: Int. Conf. on Image
Processing, vol. 1, pp. 449-452 (2002)
4. Wang, Z., Sheikh, H.R., Bovik, A.C.: No-reference perceptual quality assessment
of JPEG compressed images. In: Proc. of Int. Conf. on Image Processing, vol. 1,
pp. 477-480 (2002)