Efficient automatic detection of 3D video artifacts
read more
Citations
Trends in S3D-Movie Quality Evaluated on 105 Films Using 10 Metrics
Efficient no-reference metric for sharpness mismatch artifact between stereoscopic views
References
Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography
Object recognition from local scale-invariant features
An Introduction To Probability Theory And Its Applications
A No-Reference Objective Image Sharpness Metric Based on the Notion of Just Noticeable Blur (JNB)
Related Papers (5)
High dynamic range stereo video using SIFT and simultaneous multi-exposure
Frequently Asked Questions (14)
Q2. What future works have the authors mentioned in the paper "Efficient automatic detection of 3d video artifacts" ?
Future work will focus on improving the performance of the proposed approaches and developing new approaches for further 3D defects.
Q3. What was used as a reference in the experiment?
Deviations (CPBDd) of sharpness scores, which were estimated by the sharpness metric CPBD [15], were used as a reference in the experiment.
Q4. How do the authors exclude objects that do not appear in front of the screen?
In order to exclude objects that do not appear in front of the screen, the authors set the threshold T1 to a suitable value and perform connected component analysis with an 8-point neighbourhood to extract objects (connected components) that are displayed significantly in front of the screen.
Q5. How is the stereoscopic window violation detected?
Stereoscopic window violation is detected, using connected component analysis, when objects hit the vertical frame boundaries whilebeing in front of the virtual screen.
Q6. What is the synchronization probability of a frame?
which is computed considering a confidence level α, is used as the threshold for the estimation of the synchronization probability.
Q7. What is the condition for detecting a left or right SWV?
When a left or right SWV of duration dSWV frames is detected, the condition dSWV > fps 2is checked, where fps is the video frame rate, to determine whether the violation is perceived as annoying or not.
Q8. What is the definition of a synchronization mismatch?
Synchronization mismatch is detected based on the motion inconsistencies of feature point pairs between the LR views within a short time frame.
Q9. What is the definition of a retinal rivalry?
Retinal rivalry occurs on the left or right frame edges, when object regions positioned close to the left image’s left or right border do not have correspondence (are not displayed) in the right frame and vice versa.
Q10. What is the definition of a stereoscopic window effect?
Bent window effect: Sometimes, objects appearing infront of the virtual screen in theatre space extend vertically across the entire frame and hit both the top and bottom frame boundaries.
Q11. What are the common causes of 3D artifacts?
• Depth jump cut: During editing, video cuts between twoshots with very different average depth cause a temporary loss of the viewer’s 3D perception.
Q12. How many pixels can be measured at a comfortable viewing distance?
The width of an activity region on a high-definition (HD) image can be approximately calculated as about 64 pixels at a comfortable viewing distance.
Q13. How is the probability of a sharpness mismatch calculated?
the Probability of Sharpness Mismatches PSM is estimated and smoothed with the correction coefficients k, which are calculated considering the number of valid disparity mappings, asPSM = 1− exp(− P 2sm + k 22 · σ̃), (4)where σ̃ denotes the standard deviation between Psm and k.
Q14. What was the procedure used for the experiments?
In order to evaluate the performance of the proposed approaches, several HD 3D sequences (cf. TABLE I) were used for the experiments.