Patent No. | 10,768,303 |
---|---|

Issue Date | September 08, 2020 |

Title | Method For Identifying Individual Trees In Airborne Lidar Data And Corresponding Computer Program Product |

Inventorship | Biao Xiong, Enschede (NL) Dong Yang, Haarle (NL) |

Assignee | YaDo Holding B.V., Eindhoven (NL) Biao Xiong, Eindhoven (NL) |

1. A method for identifying individual trees in airborne lidar data, comprising the steps of:a. obtaining lidar data of a group of one or more trees to be separated into individual trees, the lidar data comprising a plurality of lidar data points;

b. define voxels in a regular 3D grid on the basis of the lidar data points;

c. applying an image segmentation algorithm to obtain at least one segment comprising a subset of the 3D voxels;

wherein the following steps are performed if at least two segments are obtained in step c: d. for each of a first segment and a second neighbouring segment of said at least two segments:

I. find the root voxel of said segment, the root voxel being a voxel having the lowest height of the voxels of said segment;

II. find the branch voxels of said segment, wherein a branch voxel is a voxel connected directly or indirectly to the root voxel;

e. merging the first segment and the neighbouring second segment if:

I. the distance between the root voxel of the first segment and the root voxel of the second segment is less than a first threshold; and

II. the distance between the root voxel of the first segment and the closest branch voxel of the second segment is less than a second threshold; and

III. the distance between each of the branch voxels of the first segment and the corresponding closest branch voxels of the second segment is less than a third threshold.

b. define voxels in a regular 3D grid on the basis of the lidar data points;

c. applying an image segmentation algorithm to obtain at least one segment comprising a subset of the 3D voxels;

wherein the following steps are performed if at least two segments are obtained in step c: d. for each of a first segment and a second neighbouring segment of said at least two segments:

I. find the root voxel of said segment, the root voxel being a voxel having the lowest height of the voxels of said segment;

II. find the branch voxels of said segment, wherein a branch voxel is a voxel connected directly or indirectly to the root voxel;

e. merging the first segment and the neighbouring second segment if:

I. the distance between the root voxel of the first segment and the root voxel of the second segment is less than a first threshold; and

II. the distance between the root voxel of the first segment and the closest branch voxel of the second segment is less than a second threshold; and

III. the distance between each of the branch voxels of the first segment and the corresponding closest branch voxels of the second segment is less than a third threshold.