3D Blob Classification for 6DoF Video Point Clouds

Solution Overview

Problem

The rendering of six-degree of freedom (6DoF) video, represented using point clouds, is computationally expensive and requires large storage or transmission capacity due to the complexity of processing and storing point cloud data, especially when dealing with a large number of points at high frame rates.

Innovation Solution

The implementation of a system and method for efficient 3D blob classification and transmission, which projects point cloud data onto a floor or ceiling plane, identifies blobs using a classification algorithm like K-means, generates bounding boxes, and encodes these boxes with coordinates to simplify processing and reduce data volume.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If point cloud data is used to represent 6DoF video, then the viewer can change position through translational movements, but the rendering becomes computationally expensive and requires large storage capacity

Engineering Contradiction:

Improveviewer position change capabilityVSAvoidrendering computational burden

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent extracts only the essential visual information from dense point cloud data by identifying and representing 3D blobs with simplified geometric primitives (spheres, ellipsoids, cylinders, cones). This extraction process removes redundant points while preserving the essential shape and position information needed for 6DoF video rendering, thereby reducing computational complexity.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent transforms point cloud data from a high-dimensional representation (individual point coordinates) to a parameterized representation (geometric primitive parameters such as center coordinates, radii, orientation angles). This parameter change enables more efficient storage and processing while maintaining the ability to render 6DoF video with translational movement capability.

Inventive Principle:

Principle #35Parameter changes

2Adaptability or versatility

If point cloud data is used to represent 6DoF video, then the viewer can change position through translational movements, but the storage capacity required becomes large

Engineering Contradiction:

Improveviewer position change capabilityVSAvoiddata volume

Core Design Contradiction:

Adaptability or versatilityVSQuantity of substance

Solution Approach 1:

The patent extracts only the essential visual information from dense point cloud data by identifying and representing 3D blobs with simplified geometric primitives (spheres, ellipsoids, cylinders, cones). This extraction process removes redundant points while preserving the essential shape and position information needed for 6DoF video rendering, thereby reducing computational complexity.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent transforms point cloud data from a high-dimensional representation (individual point coordinates) to a parameterized representation (geometric primitive parameters such as center coordinates, radii, orientation angles). This parameter change enables more efficient storage and processing while maintaining the ability to render 6DoF video with translational movement capability.

Inventive Principle:

Principle #35Parameter changes

3Adaptability or versatility

If point cloud data is used to represent 6DoF video, then the viewer can change position through translational movements, but the transmission capacity required becomes large

Engineering Contradiction:

Improveviewer position change capabilityVSAvoiddata transmission volume

Core Design Contradiction:

Adaptability or versatilityVSQuantity of substance

Solution Approach 1:

The patent extracts only the essential visual information from dense point cloud data by identifying and representing 3D blobs with simplified geometric primitives (spheres, ellipsoids, cylinders, cones). This extraction process removes redundant points while preserving the essential shape and position information needed for 6DoF video rendering, thereby reducing computational complexity.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent transforms point cloud data from a high-dimensional representation (individual point coordinates) to a parameterized representation (geometric primitive parameters such as center coordinates, radii, orientation angles). This parameter change enables more efficient storage and processing while maintaining the ability to render 6DoF video with translational movement capability.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS11568182B2System and method for 3D blob classification and transmission

Publication Date: 2023.01.31 INTEL CORP

AI summary

Embodiments described herein provide an apparatus comprising a processor to project and accumulate three-dimensional (3D) point data from a blob onto a plane; construct a histogram of the 3D point data; identify a center of mass of the blob based on histogram data; surround peaks in coordinates for data in the blob with a shape defined by a diameter of the blob based on the center of mass; obtain height data for the 3D point data; and calculate dimensions for a bounding box to surround the blob based on the shape and the height data. Other embodiments may be described and claimed.