How to measure the relevance of a retargeting approach?
ECCV'10 Proceedings of the 11th European conference on Trends and Topics in Computer Vision - Volume Part II
| Hi-index | 754.84 |
Coding methods for television signals have been proposed which would rely on the frame-to-frame correlation. But it was to be expected that the instrumentation and efficiency of such methods would be severely impeded by the large influx of new samples caused by scene changes. If, however, the spatial resolution of the new scene could be temporarily reduced, the scene change problem could be largely overcome. In this paper, psychophysical experiments and their results are described which were designed to test the hypothesis that the human observer would not perceive a temporary reduction of spatial detail after scene changes. For the tests, the bandwidth of standard television signals was temporarily reduced after each scene change by means of a transient controlled low-pass filter, the minimum bandwidth and recovery time of the filter being the main experimental variables. An average recovery time of780msec was found permissible by the most critical observers, when the minimum bandwidth of the new scene was250kc, i.e., one twentieth of the system bandwidth, reached at the end of the recovery transient.