Fig. 14From: Reliable 3D video streaming considering region of interestPerformance comparison of PV and CV. Left figures show the identified regions I–III using PV, and right figures show the regions using CV for the first frame of video “Interview”. As can be seen for the region III, using PV is more accurate rather than using CV, while for the region III, CV outperforms compared to PV. a Original video frame. b Extracted region I by PV. c Extracted region I by CV. d Extracted region II by PV. e Extracted region II by CV. f Extracted region III by PV. g Extracted region III by CVBack to article page