Systems and methods for retrieval of point cloud data from cache in response to triggering events

Patent metadata

Field	Value
Publication number	US-12449547-B2
Application number	US-202318162264-A
Country	US
Kind code	B2
Filing date	Jan 31, 2023
Priority date	Jan 31, 2023
Publication date	Oct 21, 2025
Grant date	Oct 21, 2025

Abstract

A LIDAR system includes a light source projecting laser light toward a field of view, a sensor detecting reflections from objects, and a processor. The processor generates point-cloud representations of the environment, outputs navigational information to vehicle processors, and stores the point-cloud representations in a memory as a point-cloud archive. Upon detecting a point-cloud archive output triggering event, the processor collects two or more point-clouds from the archive that were generated within a predetermined period of time relative to the detected event, and outputs them from the memory.

First claim

What is claimed is:

1. A LIDAR system, comprising:
at least one light source configured to project laser light toward a field of view of the LIDAR system;
at least one sensor configured to detect laser light reflections from objects in the field of view of the LIDAR system; and
at least one processor configured to:
use the laser light reflections to generate point-cloud representations of an environment of the LIDAR system within the field of view of the LIDAR system,
output navigational information based on the generated point-cloud representations to one or more processors associated with a vehicle on which the LIDAR system is mounted,
store at least some of the generated point-cloud representations in a memory to provide a point-cloud archive,
detect occurrence of a point-cloud archive output triggering event,
in response to detection of the point-cloud archive output triggering event, collect from the memory two or more point-clouds from the point-cloud archive that were generated within a predetermined period of time relative to the detected point-cloud archive output triggering event, and
output the two or more point-clouds collected from the memory.

2. The LIDAR system of claim 1 , wherein the memory comprises at least one volatile memory unit.

3. The LIDAR system of claim 1 , wherein the at least one processor is further configured to output one or more post-event point-clouds generated after detection of the point-cloud archive output triggering event.

4. The LIDAR system of claim 3 , wherein the one or more post-event point-clouds are included within a predetermined time window after detection of the point-cloud archive output triggering event.

5. The LIDAR system of claim 1 , wherein the predetermined period of time includes a rolling time window.

6. The LIDAR system of claim 5 , wherein the generated point-cloud representations are stored on a first in, first out basis after obtaining an initial group of point-cloud representations corresponding to the rolling time window.

7. The LIDAR system of claim 1 , wherein the predetermined period of time includes a total time from a last power-on event associated with the LIDAR system.

8. The LIDAR system of claim 1 , wherein the navigational information includes one or more identifiers and one or more range values associated with one or more objects detected in the environment of the LIDAR system.

9. The LIDAR system of claim 8 , wherein the one or more identifiers include an object type classifier associated with the one or more objects detected in the environment of the LIDAR system.

10. The LIDAR system of claim 1 , wherein the navigational information includes a point-cloud representation.

11. The LIDAR system of claim 1 , wherein the generated point-cloud representations each include a plurality of points spatially distributed relative to the field of view, and wherein each of the plurality of points is associated with a calculated range value.

12. The LIDAR system of claim 1 , wherein the predetermined period of time is determined based on a type classifier associated with the detected point-cloud archive output triggering event.

13. The LIDAR system of claim 1 , wherein the point-cloud archive output triggering event includes an inconsistency between outputs of two or more sensors associated with the vehicle.

14. The LIDAR system of claim 13 , wherein the two or more sensors include at least a camera and the LIDAR system.

15. The LIDAR system of claim 1 , wherein the point-cloud archive output triggering event includes an operator override of one or more autonomous driving system (ADS) or advanced driver assist system (ADAS) feature associated with the vehicle.

16. The LIDAR system of claim 1 , wherein the point-cloud archive output triggering event includes occurrence of an accident involving the vehicle.

17. The LIDAR system of claim 1 , wherein the point-cloud archive output triggering event includes a request received from the one or more processors associated with the vehicle or a server system remotely located relative to the vehicle.

18. The LIDAR system of claim 1 , wherein the point-cloud archive output triggering event includes a determination that an incorrect classification of a detected object has occurred.

19. The LIDAR system of claim 1 , wherein the at least one processor is configured to output the two or more point-clouds collected from the memory to one or more processing devices associated with the vehicle.

20. The LIDAR system of claim 1 , wherein the at least one processor is configured to output the two or more point-clouds collected from the memory to a non-volatile memory device onboard the vehicle.

21. The LIDAR system of claim 1 , wherein the at least one processor is configured to output the two or more point-clouds collected from the memory to a non-volatile memory device associated with the LIDAR system.

22. The LIDAR system of claim 1 , wherein the at least one processor is configured to output the two or more point-clouds collected from the memory to a server remotely located relative to the vehicle.

23. The LIDAR system of claim 1 , wherein the at least one processor is further configured to implement a multi-stage LIDAR data system including two or more stages.

24. The LIDAR system of claim 23 , wherein a first stage of the multi-stage LIDAR data system includes a pixel generator that creates pixel data from trace data received from the at least one sensor.

25. The LIDAR system of claim 24 , wherein the pixel generator stores at least some of the trace data in a trace data archive in the memory.

26. The LIDAR system of claim 24 , wherein the pixel generator stores at least some of the pixel data in a pixel data archive in the memory.

27. The LIDAR system of claim 24 , wherein the at least one processor is configured to retrieve and output, in response to detection of a pixel output triggering event, at least one of trace data from a trace data archive or pixel data from a pixel data archive that were generated within a pre-determined period of time relative to the detected pixel output triggering event.

28. The LIDAR system of claim 24 , wherein a second stage of the multi-stage LIDAR data system includes a point-cloud generator that generates the point-cloud representations from pixels extracted by the pixel generator.

29. The LIDAR system of claim 28 , wherein the point-cloud generator is configured to store one or more of the point-cloud representations in a point-cloud archive in the memory.

30. The LIDAR system of claim 28 , wherein the at least one processor is configured to retrieve and output, in response to detection of a point-cloud generator triggering event, at least one of point-cloud representation from a point-cloud archive that were generated within a pre-determined period of time relative to the detected point-cloud generator triggering event.

31. The LIDAR system of claim 28 , wherein a third stage of the multi-stage LIDAR data system includes a scene perception unit configured to identify and classify objects based on the point-cloud representations received from the point-cloud generator.

32. The LIDAR system of claim 31 , wherein the scene perception unit is configured to store at least some scene data in a scene data archive in the memory.

33

Assignees

• Innoviz Tech Ltd

Inventors

• Keilaf Omer David
• Goren Nir

Classifications

• G01S17/89

  for mapping or imaging · CPC title

• B60W2420/408

  Radar; Laser, e.g. lidar · CPC title

• B60W60/001

  Planning or execution of driving tasks · CPC title

• B60W60/007

  Emergency override (Handing over between remote control and on-board control or handing over between remote control arrangements G05D1/227) · CPC title

• B60W2556/45

  External transmission of data to or from the vehicle · CPC title