Granted Patent B2

US 12,480,768 · App. 18/117,185 · Granted Nov 25, 2025

Inventors: Wei Long (Shanghai, CN); Xuehe Zheng (Shanghai, CN); Congmin Bai (Shanghai, CN)

Assignee: Telenav, Inc.

G01C21/32G01C21/3811G01C21/3844

Abstract

Claims (56)

1 . A navigation system comprising:

a control circuit configured to:

capture a semantic frame from a sensor data stream for a region of interest,

calculate a score (EQ1) for semantic points in the semantic frame,

align the semantic points includes a pose error corrected,

generate a semantic grid map of the semantic points including a high-level grid and a low-level grid,

calculate grid map statistics by counting the semantic points in the low-level grid within the high-level grid,

filter the semantic points in the low-level grid based on the grid map statistics of the high-level grid, and

generate a semantic map from the low-level grid filtered and the high-level grid; and

a communication circuit, coupled to the control circuit, configured to: process the semantic map through a network for displaying on and controlling a device, based on the semantic map, to an area of an unmapped roadway.

2 . The system as claimed in claim 1 wherein the control circuit is configured to calculate a score (EQ1) for semantic points includes the control circuit is configured to:

determine a frame center for the semantic frame;

calculate a distance between the frame center and the semantic point; and

calculate a change of pose between two successive ones of the semantic frame.

3 . The system as claimed in claim 1 wherein the control circuit is configured to align the semantic points includes:

receive a GPS location for the frame center from a location circuit; and

calculate a GPS location for the semantic point by counting pixels between the frame center and the semantic point for fusing the sensor data stream into the semantic frame.

4 . The system as claimed in claim 1 wherein the control circuit is configured to generate a semantic grid map of the semantic points includes generating the high-level grid with dimensions of H×H meters and the low-level grid with dimensions of L×L centimeters within the high-level grid.

5 . The system as claimed in claim 1 wherein the control circuit is configured to capture the semantic frame as an overhead view of the region of interest surrounding the device includes identifying a frame center as a geolocation of the device.

6 . The system as claimed in claim 1 wherein the control circuit is configured to capture a semantic frame from a sensor data stream includes a front optical sensor, a rear optical sensor, a left side optical sensor, and a right optical sensor provide the sensor data stream.

7 . The system as claimed in claim 1 wherein the control circuit is configured to parse the semantic map through the high-level grid to the low-level grid provides centimeter accuracy on a roadway.

8 . A method of operation for a navigation system comprising:

capturing a semantic frame from a sensor data stream for a region of interest;

calculating a score (EQ1) for semantic points in the semantic frame;

aligning the semantic points including correcting a pose error;

generating a semantic grid map of the semantic points including a high-level grid and a low-level grid;

calculating grid map statistics by counting the semantic points in the low-level grid within the high-level grid;

filtering the semantic points in the low-level grid based on the grid map statistics of the high-level grid; and

generating a semantic map from the filtering of the low-level grid and the high-level grid for displaying on and controlling a device, based on the semantic map, to an area of an unmapped roadway.

9 . The method as claimed in claim 8 wherein calculating a score (EQ1) for semantic points includes:

determining a frame center for the semantic frame;

calculating a distance between the frame center and the semantic point

calculating a change of pose between two successive ones of the semantic frame.

10 . The method as claimed in claim 8 wherein aligning the semantic points includes:

receiving a GPS location for the frame center from a location circuit; and

calculating a GPS location for the semantic point by counting pixels between the frame center and the semantic point includes rotating the semantic points for fusing the sensor data stream into the semantic frame.

11 . The method as claimed in claim 8 wherein generating a semantic grid map of the semantic points includes generating the high-level grid with dimensions of H×H meters and the low-level grid with dimensions of L×L centimeters within the high-level grid.

12 . The method as claimed in claim 8 wherein capturing the semantic frame as an overhead view of the region of interest surrounding the device includes identifying a frame center as a geolocation of the device.

13 . The method as claimed in claim 8 wherein capturing a semantic frame from a sensor data stream includes a front optical sensor, a rear optical sensor, a left side optical sensor, and a right optical sensor providing the sensor data stream.

14 . The method as claimed in claim 8 further comprising parsing the semantic map through the high-level grid to the low-level grid provides centimeter accuracy on a roadway.

15 . A non-transitory computer readable medium including instructions for a navigation system comprising:

capturing a semantic frame from a sensor data stream for a region of interest;

calculating a score (EQ1) for semantic points in the semantic frame;

aligning the semantic points including correcting a pose error;

generating a semantic grid map of the semantic points including a high-level grid and a low-level grid;

calculating grid map statistics by counting the semantic points in the low-level grid within the high-level grid;

filtering the semantic points in the low-level grid based on the grid map statistics of the high-level grid; and

generating a semantic map from the filtering of the low-level grid and the high-level grid for displaying on and controlling a device, based on the semantic map, to an area of an unmapped roadway.

16 . The non-transitory computer readable medium including the instructions as claimed in claim 15 wherein calculating a score (EQ1) for semantic points includes:

determining a frame center for the semantic frame;

calculating a distance between the frame center and the semantic point; and

calculating a change of pose between two successive ones of the semantic frame.

17 . The non-transitory computer readable medium including the instructions as claimed in claim 15 wherein generating a semantic grid map of the semantic points includes generating the high-level grid with dimensions of H×H meters and the low-level grid with dimensions of L×L centimeters within the high-level grid.

18 . The non-transitory computer readable medium including the instructions as claimed in claim 15 wherein capturing the semantic frame as an overhead view of the region of interest surrounding the device includes identifying a frame center as a geolocation of the device.

19 . The non-transitory computer readable medium including the instructions as claimed in claim 15 wherein capturing a semantic frame from a sensor data stream includes a front optical sensor, a rear optical sensor, a left side optical sensor, and a right optical sensor providing the sensor data stream.

20 . The non-transitory computer readable medium including the instructions as claimed in claim 15 further comprising parsing the semantic map through the high-level grid to the low-level grid provides centimeter accuracy on a roadway.

Assignments (1)

Continuity (1)

References Cited (16)