Optimize Visual Hull with Concave Based on Depth Map

Authors

Guojun Chen, College of Computer and Communication Engineering, China University of Petroleum, Qingdao 266580, China;
Wei Xin, College of Computer and Communication Engineering, China University of Petroleum, Qingdao 266580, China;

Abstract

Abstract: Traditional method based on color image can recover the boundary of an object, but it can't recover the concave part. Although reconstruction based on depth image can revert the concave part, but the result has bad boundary which means that the edge information lost seriously. The method that fused the depth and RGB image was proposed to solve the individual shortcomings. Based on space subdivision the origin visual hull was computed based RGB image and depth image individually. In the following step, the two type visual hulls were combined for analyzing concave area and then fused into one. A parallel scheme under the CUDA platform based on GPU was implemented to accelerate processing speed. The results indicate that the proposed method can recover the visual hull for an object with complicated concave area and performs well in the part of speed and quality.

Recommended Citation

Chen, Guojun and Xin, Wei (2020) "Optimize Visual Hull with Concave Based on Depth Map," Journal of System Simulation: Vol. 27: Iss. 10, Article 42.
Available at: https://dc-china-simulation.researchcommons.org/journal/vol27/iss10/42

First Page

2508

Revised Date

2015-07-31

DOI Link

https://doi.org/

Last Page

2513

CLC

TP391.9

Recommended Citation

Chen Guojun, Wei Xin. Optimize Visual Hull with Concave Based on Depth Map[J]. Journal of System Simulation, 2015, 27(10): 2508-2513.