aries-robot/polydepth

Interactive Penetration Depth (PD) calculation library for rigid models

PolyDepth

PolyDepth is an interactive penetration depth calculation library written in C++ for rigid polygon soup models. For two intersected polygonal models, PolyDepth can compute a minimum distance to separate them by translation:

• It can compute a single distance value to separate the intersected model
• It can also report a set of distance values for each intersected region
• It has different options to choose for initialization

The code is written using Visual C++ 9.0 in MS Windows and is not tested on other platforms yet. To run the code, VC++ 9.0 or higher is needed.

Dependencies

To run PolyDepth, two libraries are needed, PQP and C2A.

1. Please download PQP at here.

2. Copy all contents of PQP to a new folder named PQP in the root path.

3. Please download C2A at here.

4. Copy only the C2A folder into the root path.

5. Copy the files PQP.sln and PQP.vcproj into PQP folder.

About the project

In PolyDepth.sln, two projects exist other than C2A and PQP:

• PolyDepthDemo: the Demo function for showing how to use PolyDepth.lib

• PolyDepth: interactive translational penetration depth calculation, and form PolyDepth.lib

The main function:

PolyDepthReturnValue 
PolyDepth(Transform*				object_1_pose,			// input pose for the object_1
          C2A_Model*				object_1,			// input model of object_1
	  Transform*				object_2_pose,			// input pose for the object_2
	  C2A_Model *				object_2,			// input model of object_2
	  Tri*                                  last_triangle1,
	  Tri*                                  last_triangle2,
	  const std::vector<Coord3D>&           object_1_vertices,		// vertices of object_1 
	  const std::vector<Coord3D>&           object_2_vertices,		// vertices of object_2
	  Transform&				output_object_1_pose,		// output: final pose of object_1 when it is seperated from the collided object_2
	  Transform&				output_object_2_pose,		// output: no use
	  std::vector<Coord3D>&		        local_penetration_depth,	// output: local penetration depths
	  std::vector<Coord3D>&		        local_penetration_features1,	// output: local penetration features from object_1
	  std::vector<Coord3D>&		        local_penetration_features2,	// output: local penetration features from object_2
          Coord3D&			        global_penetration_depth,	// output: global penetration depth
          int&                                  number_of_iteration,

		  // clear configuration could be following: 
		  //  1. maximally clear configuration computed via FindMaximallyClearConfigure
		  //  2. previous object1's pose (translation and rotation) that may be free of contact 
		  //  3. any pose that user thinks free of contact, in this case use the constructor
		  //     ClearConfiguration(Coord3D).
	   const				ClearConfigurations& clear_conf,
	   bool				        use_motion_coherance = false,
	   bool				        use_max_clear_conf = false,

		  // NOTE: call the object distance "d". if 0.0 < d < absolute_contact_configuration, 
		  // we claim that two objects are in contact configuration with each other. So, this number 
		  // is ultimately important for many applications. You can simply give a minus value,
		  // that Polydepth can pick a proper number for you. In case of minus value 
		  // so Polydepth computes this value to its purpose for you,
		  // i.e. 0.0012 * bbox size of object 2. In those simulation environments where
		  // simulation precision is important, you may specify this value by yourself.
		  //
	    PQP_REAL				absolute_contact_configuration = -1.0,
		  // provide grid resolution of the feature space; smaller value yields smaller number of features
	    PQP_REAL				feature_clustering_resolution = 5);

The result of PolyDepth is saved in a result folder under the root path.

Demo

To run the demo, you need GLUT & OpenGL. Please make sure the Startup project is setup correctly to PolyDepthDemo.

Citation

PolyDepth was presented in SIGGRAPH 2012. Please visit our website for more deatils. To cite PolyDepth in your academic research, please use the following bibtex lines:

@article{10.1145/2077341.2077346,
author = {Je, Changsoo and Tang, Min and Lee, Youngeun and Lee, Minkyoung and Kim, Young J.},
title = {PolyDepth: Real-Time Penetration Depth Computation Using Iterative Contact-Space Projection},
year = {2012},
issue_date = {January 2012},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
volume = {31},
number = {1},
issn = {0730-0301},
url = {https://doi.org/10.1145/2077341.2077346},
doi = {10.1145/2077341.2077346},
abstract = {We present a real-time algorithm that finds the Penetration Depth (PD) between general polygonal models based on iterative and local optimization techniques. Given an in-collision configuration of an object in configuration space, we find an initial collision-free configuration using several methods such as centroid difference, maximally clear configuration, motion coherence, random configuration, and sampling-based search. We project this configuration on to a local contact space using a variant of continuous collision detection algorithm and construct a linear convex cone around the projected configuration. We then formulate a new projection of the in-collision configuration onto the convex cone as a Linear Complementarity Problem (LCP), which we solve using a type of Gauss-Seidel iterative algorithm. We repeat this procedure until a locally optimal PD is obtained. Our algorithm can process complicated models consisting of tens of thousands triangles at interactive rates.},
journal = {ACM Trans. Graph.},
month = {feb},
articleno = {5},
numpages = {14},
keywords = {collision detection, polygon-soups, penetration depth, Animation, dynamics}
}

Contact

If you have any questions or find some bugs, please feel free to connect with Min Tang by e-mail: [email protected] or [email protected].