Evaluation of depth perception based on binocular stereo vision
| Autor: | Hui Ren, Zhi-bin Su, Ling-feng Chen, Dong-rui Li |
|---|---|
| Rok vydání: | 2017 |
| Předmět: |
Computer science
business.industry media_common.quotation_subject 020208 electrical & electronic engineering 020206 networking & telecommunications Stereoscopy 02 engineering and technology Luminance Visualization law.invention law Feature (computer vision) Perception 0202 electrical engineering electronic engineering information engineering Contrast (vision) Computer vision Artificial intelligence Parallax Depth perception business media_common |
| Zdroj: | ICNC-FSKD |
| DOI: | 10.1109/fskd.2017.8393240 |
| Popis: | With the development of image and video technology based on binocular stereo vision, more and more researches have focused on perception effects and influence factors of stereoscopic 3D (S3D) images. Up to now, investigation of factor analysis and evaluation for the depth perception of S3D image is still rare. Therefore, in this paper, we designed three subjective experiments with the content-based factors of luminance, contrast and subject size for depth perception assessment of S3D image sequences generated by computer, under different combination conditions of parallax and shape. Experiments results demonstrated that the luminance and size of main object exhibited remarkable impact on space perception and volume perception, while the content feature of contrast only has clear influence on the volume perception. Our studies have practical significance and reference value to improve viewing experience and relative objective models for the S3D images and videos. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|