Object registration kinect with ICP and orientation - object

I am trying to register a small symmetric object (Ex: a cube) using point clouds from kinect. I am rotating the object around its axis and getting point clouds of the object at orientations at 0, 90, 180, 270 degrees w.r.t. a global frame.
Now, I would like to combine all these four individual point clouds into one point cloud that represents/models the object. For this I would like to use ICP and the orientation information. Is there any approach that takes advantage of the orientation information to be used with Iterative Closest Point (ICP) approach ?
Or is there any other simpler way to do this ?
Thank you

Two possibilities:
Use the known transform to initialise the algorithm.
Incorporate distance from the known alignment into
the error cost used in the alignment phase.

Related

Most efficient and effective way to create a surface from 3d points

Say I had a point cloud with n number of points in 3d space(relatively densely packed together). What is the most efficient way to create a surface that goes contains every single point in it and lets me calculate values such as the normal and curvature at some point on the surface that was created? I also need to be able to create this surface as fast as possible(a few milliseconds hopefully working with python) and it can be assumed that n < 1000.
There is no "most efficient and effective" way (this is true of any problem in any domain).
In the first place, the surface you have in mind is not mathematically defined uniquely.
A possible approach is by means of the so-called Alpha-shapes, implemented either from a Delaunay tetrahedrization, or by the ball-pivoting method. For other methods, lookup "mesh reconstruction" or "surface reconstruction".
On another hand, normals and curvature can be computed locally, from neighbors configurations, without reconstructing a surface (though there is an ambiguity on the orientation of the normals).
I could suggest Nina Amenta's Power Crust algorithm (link to code), or also meshlab suite, which can compute the curvatures too.

Consistent normal calculation of a point cloud

Is there a library in python or c++ that is capable of estimating normals of point clouds in a consistent way?
In a consistent way I mean that the orientation of the normals is globally preserved over the surface.
For example, when I use python open3d package:
downpcd.estimate_normals(search_param=o3d.geometry.KDTreeSearchParamHybrid(
radius=4, max_nn=300))
I get an inconsistent results, where some of the normals point inside while the rest point outside.
many thanks
UPDATE: GOOD NEWS!
The tangent plane algorithm is now implemented in Open3D!
The source code and the documentation.
You can just call pcd.orient_normals_consistent_tangent_plane(k=15).
And k is the knn graph parameter.
Original answer:
Like Mark said, if your point cloud comes from multiple depth images, then you can call open3d.geometry.orient_normals_towards_camera_location(pcd, camera_loc) before concatenating them together (assuming you're using python version of Open3D).
However, if you don't have that information, you can use the tangent plane algorithm:
Build knn-graph for your point cloud.
The graph nodes are the points. Two points are connected if one is the other's k-nearest-neighbor.
Assign weights to the edges in the graph.
The weight associated with edge (i, j) is computed as 1 - |ni ⋅ nj|
Generate the minimal spanning tree of the resulting graph.
Rooting the tree at an initial node,
traverse the tree in depth-first order, assigning each node an
orientation that is consistent with that of its parent.
Actually the above algorithm comes from Section 3.3 of Hoppe's 1992
SIGGRAPH paper Surface Reconstruction from Unorganized Points. The algorithm is also open sourced.
AFAIK the algorithm does not guarantee a perfect orientation, but it should be good enough.
If you know the viewpoint from where each point was captured, it can be used to orient the normals.
I assume that this not the case - so given your situation, which seems rather watertight and uniformly sampled, mesh reconstruction is promising.
PCL library offers many alternatives in the surface module. For the sake of normal estimation, I would start with either:
ConcaveHull
Greedy projection triangulation
Although simple, they should be enough to produce a single coherent mesh.
Once you have a mesh, each triangle defines a normal (the cross product). It is important to note that a mesh isn't just a collection of independent faces. The faces are connected and this connectivity enforces a coherent orientation across the mesh.
pcl::PolygonMesh is an "half edge data structure". This means that every triangle face is defined by an ordered set of vertices, which defines the orientation:
order of vertices => order of cross product => well defined unambiguous normals
You can either use the normals from the mesh (nearest neighbor), or calculate a low resolution mesh and just use it to orient the cloud.

Triangulate camera position and orientation in regards to known objects

I made an object tracker that calculates the position of an object recorded in a live camera feed using stereoscopic cameras. The math was simple, once you know the camera distance and orientation. However, now I thought it would be nice to allow me to quickly extract all these parameters, so when I change my setup or cameras I will be able to quickly calibrate it again.
To calculate the object position I made some simplifications/assumptions, which made the math easier: the cameras are in the same YZ plane, so there is only a distance in x between them. Their tilt is also just in the XY plane.
To reverse the triangulation I thought a test pattern (square) of 4 points of which I know the distances to each other would suffice. Ideally I would like to get the cameras' positions (distances to test pattern and each other), their rotation in X (and maybe Y and Z if applicable/possible), as well as their view angle (to translate pixel position to real world distances - that should be a camera constant, but in case I change cameras, it is quite a bit to define accurately)
I started with the same trigonometric calculations, but always miss parameters. I am wondering if there is an existing solution or a solid approach. If I need to add parameter (like distances, they are easy enough to measure), it's no problem (my calculations didn't give me any simple equations with that possibility though).
I also read about Homography in opencv, but it seems it applies to 2D space only, or not?
Any help is appreciated!

Point Cloud - Principal Axes - Use of Inertia

I have got point clouds of different primitive objects (cone, plane, torus, cylinder, sphere, ellipsoid). The all vary in orientation, position and scaling. Furthermore all of them are initialized with a unique set of parameters (e.g. height, radius, etc.) so that their shape can be quiet different (some cones are tall, others are small and fat).
Now to my question:
I am trying to find the objects "principal components". Using PCA doesn't lead to good results, since rotated primitives can have their main variation in any direction (which doesn't have to be necessarily along the length of the objects).
The only chance that I see is to use somehow the symmetry of my primitives. Isn't there a method based on inertia? Maybe some way to find the main symmetry axis and two others perpendicular to it?
Can you give me some advice or point me to papers or implementations (maybe even python)?
Thanks a lot, Merlin.
PS: This is what I get if I only apply a PCA. Especially for cones this doesn't really work. Only cones that are almost identical in shape share the same orientation, but I need them all to point in one direction (e.g. up).
So you got cones and just need to rotate them all in the same direction?
If so you can fit a triangle to them and point the peak (e.g with the perpendicular bisectors of the sides) to your main axis.
You have an interesting problem. Normally used shape descriptors (VFH) that are invariant to shape but not pose (which is what you would want, really) would not be invariant to stretching in the shape.
I think to succeed at this you need to be clearer about the invariants that you are trying to maintain when a shape changes. Is it a topological invariant? If so, then here is a good starting point: https://www.google.com.tr/search?q=topologically+invariant+shape+descriptor
I decided to just stick to simple PCA since it's the only method that is totally generic and doesn't depend on prior (expert) knowledge about the data.

how to convert depth in Z-cordinate

I'm making a project in which i need to draw user's feet in a rectangle of 640*480
and i'm mapping the skeleton joints coordinate into depth coordinates to fit it in the box,
but the DepthImagePoint gives x-y coordinate and depth(in mm) but i want x-z coordinate.
How can I get the z coordinate to fit in the resolution of 640*480?
or Can i somehow convert skeleton joints coordinate to proper resolution to fit the box?
I'm using Microsoft Kinect SDK with C#.
Thanks in Advance.
There are several functions in the CoordinateMapper that will map points between the different data streams. MapSkeletonPontToDepthPoint being the first to come to mind. Others in the mapper may also assist in finding the proper depth mapping you are looking for.
http://msdn.microsoft.com/en-us/library/jj883690.aspx

Resources