# US 11,908,210 B2 · Method and apparatus for detecting object in three-dimensional (3D) point cloud

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 11,908,210 B2
- Document ID: US11908210B2
- Status: Granted
- Kind: B2
- Application number: 17/193,558
- Filed: 2021-03-05
- Published: 2021-09-09
- Granted: 2024-02-20
- Earliest claimed: 2020-03-06
- Assignees: Meissa Inc.
- Applicants: Meissa Inc.
- Inventors: Suk Won Choi
- Web page: https://patentgrep.com/patent/US11908210B2

## CPC

- G06V 20/647: {by matching two-dimensional images to three-dimensional objects}
- G06T 7/62: of area, perimeter, diameter or volume
- G06T 7/75: {involving models}
- G06T 17/10: Constructive solid geometry [CSG] using solid primitives, e.g. cylinders, cubes
- G06V 20/13: Satellite images
- G06T 2207/10028: Range image; Depth image; 3D point clouds
- G06V 20/188: {Vegetation}
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL

## Abstract

A method and apparatus for detecting an object in a three-dimensional (3D) point cloud are disclosed. The apparatus includes a three-dimensional (3D) model reconstruction unit configured to reconstruct a three-dimensional (3D) model using three-dimensional (3D) point cloud data, a two-dimensional (2D) object detection unit configured to detect an object in at least one two-dimensional (2D) image captured to generate the three-dimensional (3D) point cloud data, a three-dimensional (3D) position estimation unit configured to form a cylinder surrounding the object in the three-dimensional (3D) model based on a two-dimensional (2D) position of the detected object so as to estimate a three-dimensional (3D) position of the detected object using the cylinder, and an object region extraction unit configured to acquire three-dimensional (3D) point cloud data contained in the cylinder so as to extract an object region from the acquired three-dimensional (3D) point cloud data.

## Claims

### Claim 1

A method for detecting an object in a three-dimensional (3D) point cloud comprising:
by a three-dimensional (3D) model reconstruction unit, reconstructing a three-dimensional (3D) model using three-dimensional (3D) point cloud data;
by a two-dimensional (2D) object detection unit, detecting an object in at least one two-dimensional (2D) image captured to generate the three-dimensional (3D) point cloud data;
by a three-dimensional (3D) position estimation unit, extracting an external appearance of the object from a top-view image obtained from the three-dimensional (3D) model, extracting planar coordinates corresponding to the external appearance of the object, forming a cylinder surrounding the object in the three-dimensional (3D) model based on a two-dimensional (2D) position of the detected object based on the planar coordinates, and estimating a three-dimensional (3D) position of the detected object using the cylinder; and
by an object region extraction unit, acquiring three-dimensional (3D) point cloud data contained in the cylinder, and extracting an object region from the acquired three-dimensional (3D) point cloud data,
wherein a bottom surface of the cylinder is set to a planar figure including the planar coordinates.

### Claim 2 (depends on 1)

The method according to claim 1, further comprising:
by the three-dimensional (3D) model reconstruction unit, estimating a position of each point in the three-dimensional (3D) model using at least one ground reference point included in the two-dimensional (2D) image.

### Claim 3 (depends on 1)

The method according to claim 1, wherein:
the cylinder is formed to have a predetermined height or less from a ground surface.

### Claim 4 (depends on 1)

The method according to claim 1, further comprising:
by the object region extraction unit, acquiring three-dimensional (3D) point cloud data contained in the cylinder using a clustering algorithm.

### Claim 5 (depends on 1)

The method according to claim 1, further comprising:
by the object region extraction unit, forming the cylinder that has a tilted shape depending on an angle where the object is photographed.

### Claim 6 (depends on 1)

The method according to claim 1, further comprising:
by an inventory calculation unit, analyzing a volume of the extracted object region and to calculate an inventory of the object based on the analyzed volume.

### Claim 7 (depends on 1)

The method according to claim 1, wherein:
external appearance pattern information of the object including at least one of iron ore, cement, and wood is preset in the two-dimensional (2D) object detection unit.