Granted Patent B2

US 12,586,312 · App. 18/560,159 · Granted Mar 24, 2026

Inventors: Norman Haag (Stuttgart, DE); Jan Sparbert (Rutesheim, DE); Stefan Spiessberger (Weinstadt, DE)

Assignee: ROBERT BOSCH GMBH

G06T17/00G01S17/89G06T7/62G06T7/70G06V10/26G06V2201/07

Abstract

Claims (33)

1 . A method for determining concealed objects in a 3D point cloud representing an environment, the method comprising the follow steps:

producing a 3D point cloud using an active sensor, wherein each point of the 3D point cloud represents a distance measurement by the active sensor;

determining a background region within the 3D point cloud;

determining shadows within the background region of the 3D point cloud, wherein the shadows are regions within the background region at which there are substantially no points of the 3D point cloud; and

determining, in an environment of the active sensor, an object which is concealed by crosstalk in the active sensor, by identifying, in the background region of the 3D point cloud, at least one shadow which cannot be attributed to an object in a foreground of the 3D point cloud,

wherein a type and/or a size and/or a position of the concealed object is determined based on a characteristic of the shadow produced by the object,

wherein the information about the shadow, determined in the background region, of the object concealed in the foreground is used to add points representing the concealed object to the 3D point cloud in a region of the foreground in which the concealed object was determined, and

wherein a dead time of a detector of the active sensor is taken into account when determining a depth extent of the region in which points representing the concealed object are added in the 3D point cloud.

2 . The method according to claim 1 , wherein the background region is identified within the 3D point cloud by:

identifying, for each solid angle of the 3D point cloud, a last echo resulting from a scanning signal emitted by the active sensor; and/or

determining contiguous areas within the 3D point cloud that:

have a predefined minimum distance from the active sensor, and/or

are oriented within a predefined angle range in relation to a main axis of the active sensor, and/or

have a predefined minimum size.

3 . The method according to claim 1 , wherein, using a further sensor of a different type than the active sensor and/or map material related to a current environment, a plausibility check is made of:

(i) the determined background region, and/or

(ii) a presence of the concealed object, and/or

(iii) a type and/or a size and/or a position of the concealed object, and/or

(iv) existing crosstalk.

4 . The method according to claim 1 , wherein the determination of the shadows in the background region of the 3D point cloud takes place using a machine learning method.

5 . The method according to claim 1 , wherein information about the concealed object and/or the 3D point cloud to which the concealed object is added is taken into account during environment recognition in an environment recognition system.

6 . The method according to claim 1 , wherein the active sensor is a lidar sensor of an environment sensing system of a transportation arrangement.

7 . A device for determining concealed objects in a 3D point cloud representing an environment, comprising:

an active sensor; and

an evaluation unit;

wherein the evaluation unit is configured to determine concealed objects in a 3D point cloud representing an environment, the evaluation unit configured to:

produce a 3D point cloud using the active sensor, wherein each point of the 3D point cloud represents a distance measurement by the active sensor,

determine a background region within the 3D point cloud,

determine shadows within the background region of the 3D point cloud, wherein the shadows are regions within the background region at which there are substantially no points of the 3D point cloud, and

determine, in an environment of the active sensor, an object which is concealed by crosstalk in the active sensor, by identifying, in the background region of the 3D point cloud, at least one shadow which cannot be attributed to an object in a foreground of the 3D point cloud,

wherein a type and/or a size and/or a position of the concealed object is determined based on a characteristic of the shadow produced by the object,

wherein the information about the shadow, determined in the background region, of the object concealed in the foreground is used to add points representing the concealed object to the 3D point cloud in a region of the foreground in which the concealed object was determined, and

wherein a dead time of a detector of the active sensor is taken into account when determining a depth extent of the region in which points representing the concealed object are added in the 3D point cloud.

Assignments (1)

Priority Claims (1)

Continuity (1)

References Cited (16)