search

tatfook / SummerOfCode

1 stars 2 forks source link

从图片生成3D模型：3d reconstruction from images #3

Open LiXizhi opened 7 years ago

LiXizhi commented 7 years ago

张颂泽

mentors:

language: NPL / lua

BarryZSZ commented 7 years ago

3D Reconstuction（3维重建）

  1. Time of Flight，ToF（飞行时间发）

  1. Structured Light（结构光法）

  1. Triangle measuring method（三角测距法）

  1. Shape From Shading, SFS（阴影恢复形状法）

  2. Multi-View Stereo, MVS（立体视觉法）

BarryZSZ commented 7 years ago

Method of 3D Restoration

Step 1

Find the informations of each pixels' features by SIFT(Scale-invariant feature transform) algorithm.

Link of the codes： SIFT

Paper: D. G. Lowe, "Object recognition from local scale-invariant features," Proceedings of the Seventh IEEE International Conference on Computer Vision, Kerkyra, 1999, pp. 1150-1157 vol.2.

Step 2

Structure from Motion: Using the feature informations from Step 1, match the pixels of different images. Then, estimate the camera parameters and get Sparse reconstruction. Here, we use the Bundler algorithm.

Link of the codes: Bundler

Paper: Goesele, M., Curless, B., & Seitz, S. M. (2006). Multi-View Stereo Revisited. IEEE Computer Society Conference on Computer Vision and Pattern Recognition (Vol.2, pp.2402-2409). IEEE Computer Society.

Step 3

Multi-view Stereo Reconstruction: Based on the results from Step 2, get the Dense Reconstruction by PMVS(patch-based multi-view stereo) algorithm.

Link of the codes: CMVS-PMVS

Paper: Y. Furukawa and J. Ponce, "Accurate, Dense, and Robust Multiview Stereopsis," in IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 32, no. 8, pp. 1362-1376, Aug. 2010. doi: 10.1109/TPAMI.2009.161

Step 4

Remove the noise by Poisson Surface Reconstruction algorithm.

Link of the codes: PoissionSR

Paper: Kazhdan, M., Bolitho, M., & Hoppe, H. (2006). Poisson surface reconstruction. Eurographics Symposium on Geometry Processing (Vol.32, pp.61-70). Eurographics Association.