Method and system for automatic characterization of a three-dimensional (3D) point cloud

A system comprising at least one processor and at least one memory comprising executable instructions, which, when executed by the at least one processor, cause the system to:

• access a 3D point cloud, the 3D point cloud being a set of data points representative of the object;
• determine, based on the 3D point cloud, a 3D reconstructed object;
• iteratively smooth and contract the 3D reconstructed object to generate an average line of the 3D reconstructed object, wherein the average line is a meshed structure, and wherein the contracting is iteratively performed until the average line has an internal volume that is lower than a pre-determined volume;
• generate, based on the average line, a digital framework of the 3D point cloud, the digital framework being a ramified 3D tree structure, the digital framework being representative of a base structure of the object, wherein ramifications of the ramified 3D tree structure are determined by identifying points in the average line that have more than one neighboring point;
• morph a 3D reference model of the object onto the 3D reconstructed object, the morphing being based on the digital framework, the 3D reference model comprising one or more landmarks such that, upon morphing the 3D reference model of the object onto the 3D reconstructed object, the one or more landmarks provide indication of a corresponding one or more areas of interest of the 3D point cloud;
• determine, based on the morphed 3D reference model and the 3D reconstructed object, characteristics of the object, the characteristics of the object being determined in the one or more areas of interest;
• and
• refine the one or more areas of interest, the refining comprising:
• intersecting the 3D point cloud on a projection plane;
• determining a contour of the projection of the 3D point cloud;
• determining a convex hull of the projection, thereby determining a plurality of convex hull feature points, consecutive convex hull feature points being interconnected by a segment of the convex hull;
• and
• determining, based on relative distances between consecutive convex hull feature points, sub-areas of interest.

A non-transitory computer-readable medium comprising computer-executable instructions that cause a system to:

• access a 3D point cloud, the 3D point cloud being a set of data points representative of the object;
• determine, based on the 3D point cloud, a 3D reconstructed object;
• iteratively smooth and contract the 3D reconstructed object to generate an average line of the 3D reconstructed object, wherein the average line is a meshed structure, and wherein the contracting is iteratively performed until the average line has an internal volume that is lower than a pre-determined volume;
• generate, based on the average line, a digital framework of the 3D point cloud, the digital framework being a ramified 3D tree structure, the digital framework being representative of a base structure of the object, wherein ramifications of the ramified 3D tree structure are determined by identifying points in the average line that have more than one neighboring point;
• morph a 3D reference model of the object onto the 3D reconstructed object, the morphing being based on the digital framework, the 3D reference model comprising one or more landmarks such that, upon morphing the 3D reference model of the object onto the 3D reconstructed object, the one or more landmarks provide indication of a corresponding one or more areas of interest of the 3D point cloud;
• determine, based on the morphed 3D reference model and the 3D reconstructed object, characteristics of the object, the characteristics of the object being determined in the one or more areas of interest;
• and
• refine the one or more areas of interest, the refining comprising:
• intersecting the 3D point cloud on a projection plane;
• determining a contour of the projection of the 3D point cloud;
• determining a convex hull of the projection, thereby determining a plurality of convex hull feature points, consecutive convex hull feature points being interconnected by a segment of the convex hull;
• and
• determining, based on relative distances between consecutive convex hull feature points, sub-areas of interest.