RGBDPlaneDetection

RGBD plane detection and color-based plane refinement with MRF optimization. This code is an implementation of the RGB-D plane detection part in the 3DLite paper shown below, specifically, Section 4.1 and 5.1 in the paper.

Reference

• General pipeline: Huang, Jingwei, et al. 3DLite: Towards Commodity 3D Scanning for Content Creation. ACM Transactions on Graphics, 2017.
• Plane detection: Feng, Chen, Yuichi Taguchi, and Vineet R. Kamat. Fast plane extraction in organized point clouds using agglomerative hierarchical clustering. Robotics and Automation (ICRA), 2014 IEEE International Conference on. IEEE, 2014.
• MRF: See the readme file in MRF 2.2 link for details about reference papers.

Dependencies

• OpenCV
• Eigen 3
• MRF 2.2 (already included)
• PEAC (already included)

Usage

RGBDPlaneDetection <-o> color_image depth_image

• -o is running MRF optimization to refine planes.

Build

In Windows, use Visual Studio to open sln file and compile and build the code.

In Linux, you have to write MakeFile or CMakeLists file by yourself. On the other hand, you can convert .vcxproj to CMakeLists easily if you use cmakeconverter or other converter.

Output

• Plane segmentation image in PNG
• Plane label image in PNG: plane label each pixel belongs to (starting from 0 to number_of_points_on_the_plane - 1). If a pixel is not on any plane, then its label value is number_of_points_on_the_plane.
• Plane data file in TXT. Each line represents one plane with format like this:

#plane_index(starting from 0) number_of_points_on_the_plane plane_color_in_png_image(r,g,b between [0,255]) plane_normal(1x3) plane_center(1x3) sx sy sz sxx syy szz sxy syz sxz

Here (sx sy sz) are average of sum of all 3D points (x, y, z) on the plane, (sxx syy szz sxy syz sxz) are the average of sum of x*x, y*y, z*z, x*y, y*z, z*z of all 3D points on the plane, respectively.

Note

• Currently the code only works on BundleFusion or 3DLite RGBD data. If you want to use other kinds of RGBD data, you need to rewrite the part of reading color and depth images, and reset the camera intrinsic parameters in plane_detection.h.
• Note for the scale factor for depth images in plane_detection.h.
• Sometimes the MRF 2.2 source code crashes in Visual Studio due to some memory management bug, but it seems to work fine in Linux. If you meet to problem, just have a try in Linux, or try to implement the graph-cut/max-flow code by yourself. One suggestion is to use boykov_kolmogorov_max_flow in Boost library.