Research Article

Rendering Distortion Estimation Model for 3D High Efficiency Depth Coding

Table 3

Bjontegaard delta results for proposed algorithm relative to 3D-HTM, MV-HEVC, and HEVC simulcast, comparing the decoded rendered virtual views from original uncompressed texture video and depth map.

SequenceBD-rate (%)BD-PSNR (dB)
3D-HEVCMV-HEVCHEVC simulcast3D-HEVCMV-HEVCHEVC simulcast

Kendo−11.955−22.423−35.1350.4020.8531.537
Balloons−13.231−24.406−31.8940.4280.8441.176
Lovebird1−8.468−19.346−24.3970.4440.8061.052
Newspaper−11.122−17.092−25.1500.7541.0971.646
Undo_Dancer−11.236−14.624−24.4990.7381.0281.818
GT_Fly−18.345−24.548−48.7070.3030.4470.932
Poznan_Street−15.651−22.292−28.2870.6560.9661.306
Poznan_Hall2−13.222−20.565−31.5620.4090.7301.231

Average−12.904−20.662−31.2040.5170.8461.337