상세정보
학술대회
Binocular suppression analysis for efficient compression of UHD stereo images
- 발행일
- 202207
- 출처
- International Technical Conference on Circuits/Systems, Computers and Communications (ITC-CSCC) 2022, pp.796-798
- 협약과제
- 22ZH1200, 초실감 입체공간 미디어·콘텐츠 원천기술연구, 이태진
- 초록
- The latest codecs such as H.264 and H.265 support encoding for stereo images. However, it still has too high bit rate to be applied to the current broadcasting system. As an alternative, methods such as 'Crossview-SHVC' and 'VEI' have been studied. But when encoded in these ways, there is a difference in image quality between the reference video and the additional video. So far, it is presumable that binocular suppression occurs when we see the stereo images of asymmetric qualities. This paper analyzed the phenomenon of binocular suppression through subjective image quality evaluation in UHD stereo environment. As the quality of the reference image was higher, it was confirmed that the binocular suppression phenomenon was maintained even with a larger image quality difference. Also, as a result of the experiment, contrary to conventional wisdom, the degree of image quality difference in which binocular suppression was maintained was not related to the SI of the stereo image. Also, it was not related to the three-dimensional effect of stereo images.
- KSP 제안 키워드
- Binocular suppression, Bit Rate, Broadcasting system, Reference Image, Three dimensional(3D), Three dimensional effects, image quality evaluation, stereo images, subjective image quality