System and method for incremental data processing

Patent metadata

Field	Value
Publication number	US-10887433-B2
Application number	US-201815944486-A
Country	US
Kind code	B2
Filing date	Apr 3, 2018
Priority date	Apr 3, 2018
Publication date	Jan 5, 2021
Grant date	Jan 5, 2021

Abstract

Systems and methods are provided for segmenting a data frame to be acquired into a number of incremental data of equal data length. A first incremental data of the data frame can be acquired from one or more sensors. The first incremental data of the data frame can be processed while a next incremental data of the data frame is being acquired from the one or more sensors. The acquiring and processing of incremental data of the data frame can continue until a last incremental data of the data frame is acquired and processed. Processed incremental data can be outputted as a processed data frame.

First claim

The invention claimed is:

1. A method comprising:
acquiring, by one or more sensors on a vehicle, a data frame, wherein: the one or more sensors includes a Lidar; and the Lidar rotates about an axis of the vehicle;
segmenting, by one or more computing processors, the data frame to be acquired into a number of incremental data of equal data length, wherein each of the incremental data of the data frame corresponds to an arc of a 360 degree scan, the arc being proportional to a ratio of the incremental data to the data frame;
processing, by the one or more computing processors, a first incremental data of the data frame while a next incremental data of the data frame is being acquired from the one or more sensors, the acquiring and the processing of incremental data of the data frame continuing until a last incremental data of the data frame is acquired and processed;
identifying an object in the first incremental data of the data frame;
identifying the object in a first incremental data of a second data frame;
determining a relative movement of the object from the data frame to the second data frame;
processing a first incremental data of a third data frame based on the relative movement of the object to reduce a data processing time and improve latency;
reconstructing the third data frame based on the processing of the first incremental data of the third data frame to reduce a file size of the third data frame; and
outputting, by the one or more computing processors, processed incremental data of the reconstructed third data frame as a processed data frame.

2. The method of claim 1 , wherein segmenting the data frame to be acquired into the number of the incremental data is user selectable.

3. The method of claim 1 , wherein the processed data frame is a three dimensional map of an environment, the three dimensional map comprises relative distances between the object and a second object in the environment.

4. The method of claim 1 , further comprising:
identifying, using one or more machine learning techniques, the object in the incremental data of the data frame;
detecting the amount of movement in the object from the first incremental data to the next incremental data using the one or more machine learning techniques; and
analyzing, using the one or more machine learning techniques, the object with a corresponding object from a previous incremental data of a previous data frame to determine relative motion of the object.

5. The method of claim 4 , wherein the analyzing the object with a corresponding object from a previous incremental data further comprises:
identifying the corresponding object from the previous incremental data of the previous data frame;
determining, using one or more indicators and using one or more machine learning techniques, relative movements between the corresponding object to the object between data frames; and
identifying, based on the relative movements, one or more incremental data to be minimally processed in a next data frame to be acquired.

6. The method of claim 4 , wherein analyzing the one or more objects with the previously identified objects from the previous incremental data of the previous data frame to determine relative motion of the one or more objects further comprises:
identifying the corresponding object from the previous incremental data of the previous data frame;
determining, using one or more indicators and using one or more machine learning techniques, the object has not moved between data frames; and
identifying one or more incremental data to bypass processing in a next data frame to be acquired.

7. The method of claim 1 , wherein outputting processed incremental data as a processed data frame further comprises:
reassembling one or more processed incremental data of the data frame into the processed data frame.

8. The method of claim 1 , wherein the processing of the first incremental data of the third data frame is based on whether the object is approaching or becoming more distant from the data frame to the second data frame.

9. The method of claim 1 , wherein the processing the first incremental data of the third data frame further comprises diverting, to at least one other system of the vehicle, a processing power that is determined to no longer be needed to process the first incremental data of the third data frame based on the relative movement of the object.

10. A system comprising:
one or more sensors including a Lidar on a vehicle, wherein the Lidar rotates about an axis of the vehicle;
one or more processors; and
a memory storing instructions that, when executed by the one or more processors, cause the system to perform:
acquiring, by the one or more sensors, a data frame;
segmenting, by the one or more processors, the data frame to be acquired into a number of incremental data of equal data length, wherein each of the incremental data of the data frame corresponds to an arc of a 360 degree scan, the arc being proportional to a ratio of the incremental data to the data frame;
processing, by the one or more processors, a first incremental data of the data frame while a next incremental data of the data frame is being acquired from the one or more sensors, the acquiring and the processing of incremental data of the data frame continuing until a last incremental data of the data frame is acquired and processed;
identifying an object in the first incremental data of the data frame;
identifying the object in a first incremental data of a second data frame;
determining a relative movement of the object from the data frame to the second data frame;
processing a first incremental data of a third data frame based on the relative movement of the object to reduce a data processing time and improve latency;
reconstructing the third data frame based on the processing of the first incremental data of the third data frame to reduce a file size of the third data frame; and
outputting, by the one or more computing processors, processed incremental data of the reconstructed third data frame as a processed data frame.

11. The system of claim 10 , wherein segmenting the data frame to be acquired into the number of the incremental data is user selectable.

12. The system of claim 10 , wherein the instructions further cause the system to perform:
identifying, using one or more machine learning techniques, the object in the incremental data of the data frame; and
analyzing, using the one or more machine learning techniques, the object with a corresponding object from a previous incremental data of a previous data frame to determine relative motion of the object.

13. The system of claim 12 , wherein the analyzing the object with a corresponding object from a previous incremental data further comprises:
identifying the corresponding object from the previous incremental data of the previous data frame;
determining, using one or more indicators and using one or more machine learning techniques, relative movements between the corresponding object to the object between data frames; and
identifying, based on the relative movements, one or more incremental data to be minimally processed in a next data frame to be acquired.

14. A non-transitory computer readable medium comprising instructions that, when executed, cause one or more processors to perform:
acquiring, by one or more sensors on a vehicle, a data frame, wherein: the one or more sensors includes a Lidar; and the Lidar rotates about an axis of the vehicle;
segmenting the data frame to be acquired into a number of incremental data of equal data length, wherein each of the incremental data of the da

Assignees

• Pony Ai Inc

Inventors

• Fu Haoying
• Yang Mengda
• Yu Xiang
• Lou tiancheng
• Peng Jun
• Xiao Sinan
• Li Tianyi
• Song Hao

Classifications

• G01S17/89 Primary

  for mapping or imaging · CPC title

• G01S17/00

  Systems using the reflection or reradiation of electromagnetic waves other than radio waves, e.g. lidar systems · CPC title

• G01S15/50

  Systems of measurement, based on relative movement of the target · CPC title

• G01S7/4802

  using analysis of echo signal for target characterisation; Target signature; Target cross-section · CPC title

• G06N20/00

  Machine learning · CPC title