Object recognition for 3D models and 2D drawings
A first method is disclosed for recognizing 3D objects in 3D models created by 3D scanners, depth sensing cameras or created by a 3D modeling software application. A second method is disclosed for recognizing 2D objects in drawings. The 3D/2D objects can be individual objects that have simple forms or combined objects that are comprised of a plurality of individual objects that are attached to each other in a certain manner to form one entity. The first and second methods serve a variety of medical, engineering, industrial, gaming and augmented reality applications.
1. A method for recognizing an individual 3D object represented by a 3D model according to a vector graphics format wherein the method comprising:
slicing the individual 3D object into sections with a plurality of planes creating a number of 2D shapes;
determining the axis pattern that describes a path connecting between the centers of the 2D shapes that are located on successive planes;
determining the shapes pattern that describes the configuration of the 2D shapes;
determining the areas pattern that describes the area variations between the 2D shapes; and
checking the axis pattern, the shapes pattern, and the areas pattern against a database that associates each unique combination of an axis pattern, a shapes pattern, and an areas pattern with a unique ID identifying an individual 3D object.
2. The method of claim 1 wherein said plurality of planes are parallel to the xy-plane, xz-plane, or yz-plane.
3. The method of claim 1 wherein one or more of said 2D shapes are located on the same section.
4. The method of claim 1 wherein one or more of said axis pattern, said shape pattern, and said areas pattern are described by one or more words.
5. The method of claim 1 wherein one or more of said axis pattern, said shape pattern, and said areas pattern are described by one or more formulas or equations.
6. The method of claim 1 wherein said 3D object is 2D object, said 3D model is a 2D drawing, said planes are rays, said 2D shapes are lines, and said area variation is a length variation between said lines.
7. The method of claim 1 wherein said 3D model is originally represented according to a raster graphics format and converted into a 3D model represented according to a vector graphics format using an edge detection program.
8. The method of claim 1 wherein said 3D model is originally represented by a set of points using the point cloud technique and converted into a 3D model represented according to a vector graphics format.
9. The method of claim 6 wherein said rays are parallel to the x-axis, the y-axis, or the z-axis.
10. The method of claim 6 wherein one or more lines of said lines are located on one layer of said layers.
11. A method for recognizing a combined 3D object represented by a 3D model according to a vector graphics format wherein the combined 3D object is comprised of a plurality of individual 3D objects attached to each other and the method comprising:
for each individual object;
slicing the individual 3D object into sections with a plurality of planes creating a number of 2D shapes;
determining the axis pattern that describes a path connecting between the centers of the 2D shapes that are located on successive planes;
determining the shapes pattern that describes the configuration of the 2D shapes;
determining the areas pattern that describes the area variations between the 2D shapes; and
checking the axis pattern, the shapes pattern, and the areas pattern against a database that associates each unique combination of an axis pattern, a shapes pattern, and an areas pattern with a unique ID identifying an individual 3D object;
listing the IDs of the individual 3D objects;
describing the attachment relationship between the individual 3D object; and
checking the IDs of the individual 3D objects and the attachment relationship with a database that associates each unique combination of IDs of individual 3D objects and an attachment relationship of individual 3D objects with a unique ID identifying a combined 3D object.
12. The method of claim 11 wherein said attachment relationship is represented by a graphical representation illustrating the individual 3D objects that are attached to each other.
13. The method of claim 11 wherein said attachment relationship is represented by a table indicating the list of the individual 3D objects that are attached to each other.
14. The method of claim 11 wherein said attachment relationship is represented by a text indicating the list of the individual 3D objects that are attached to each other.
15. The method of claim 11 wherein said 3D model is originally represented according to a raster graphics format and converted into a 3D model represented according to a vector graphics format using an edge detection program.
16. The method of claim 11 wherein said 3D model is originally represented by a set of points using the point cloud technique and converted into a 3D model represented according to a vector graphics format.
17. The method of claim 11 wherein said combined 3D object is a combined 2D object,
said individual 3D objects are individual 2D objects, said 3D model is a 2D drawing,
said planes are rays, said 2D shapes are lines, and said area variation is a length variation between said lines.
18. The method of claim 17 wherein said rays are parallel to the x-axis, the y-axis, or the z-axis.
19. The method of claim 17 wherein one or more lines of said lines are located on one layer of said layers.