Granted Patent B2

US 12,539,805 · App. 18/307,615 · Granted Feb 3, 2026

Inventor: Wenyao Su (Shanghai, CN)

Assignee: Huawei Technologies Co., LTD.

B60Q1/1423G06V10/22G06V10/56G06V10/60G06V10/764G06V10/774G06V20/56H05B47/11B60Q2300/314

Abstract

Claims (72)

1 . An intelligent light switching method, wherein the method comprises:

obtaining an image, wherein the image comprises lamp source information;

calculating an ambient light brightness value corresponding to the image;

classifying, based on the lamp source information, a lamp source comprised in the image to obtain a classification result, wherein the classification result is obtained by determining a vanishing point horizontal line (VP) line in the image, wherein the classifying the lamp source comprised in the image to obtain the classification result comprises:

inputting the image into a lamp source detection model; and

obtaining a lamp source category in the image based on the lamp source detection model, wherein the obtaining the lamp source category in the image based on the lamp source detection model comprises:

selecting, from the image, a bright spot whose brightness value is greater than a preset threshold, and setting a lamp box for the bright spot;

performing pairing on the lamp box to obtain a plurality of lamp box pairs;

determining the VP line based on the plurality of lamp box pairs, wherein the VP line is used to distinguish between a first region and a second region of the image; and

classifying the bright spot based on a position relationship between the bright spot and the VP line and a color feature of the bright spot, to obtain different lamp source categories; and

performing light switching based on the ambient light brightness value corresponding to the image and the classification result.

2 . The method according to claim 1 , wherein the calculating an ambient light brightness value corresponding to the image comprises:

selecting at least one region from the image, and calculating a brightness value of the at least one region; and

calculating, based on the brightness value of the at least one region, the ambient light brightness value corresponding to the image.

3 . The method according to claim 1 , wherein

the lamp source detection model is trained by using a plurality of sample images, wherein each of the plurality of sample images comprises a lamp source and annotation information of the lamp source.

4 . The method according to claim 1 , wherein before the performing pairing on the lamp box, the method further comprises:

performing overlap removal on the lamp box, to ensure that there is no overlap or tangent between lamp boxes.

5 . The method according to claim 1 , wherein the determining a vanishing point horizontal line VP line based on the plurality of lamp box pairs comprises:

mapping lamp box centers of the lamp box pairs onto rows in the image, to obtain an in-row quantity distribution map of the lamp box pairs;

selecting a VP line based on the in-row quantity distribution map;

correcting the VP line based on a preset correction value, to obtain a corrected VP line; and

adjusting the corrected VP line by using a reference value, to obtain a VP line used to classify the bright spot, wherein the reference value is used to describe a change of a pitch angle of a video camera that captured the image.

6 . An apparatus, comprising:

at least one processor; and

one or more memories coupled to the at least one processor and storing programming instructions for execution by the at least one processor to:

obtain an image, wherein the image comprises lamp source information;

calculate an ambient light brightness value corresponding to the image;

classify based on the lamp source information a lamp source comprised in the image to obtain a classification result, wherein the classification result is obtained by determining a vanishing point horizontal line (VP) line in the image, wherein the classify the lamp source comprised in the image to obtain the classification result comprises:

input the image into a lamp source detection model; and

obtain a lamp source category in the image based on the lamp source detection model, wherein the obtain the lamp source category in the image based on the lamp source detection model comprises:

select, from the image, a bright spot whose brightness value is greater than a preset threshold, and setting a lamp box for the bright spot;

perform pairing on the lamp box to obtain a plurality of lamp box pairs;

determine the VP line based on the plurality of lamp box pairs, wherein the VP line is used to distinguish between a first region and a second region of the image; and

classify the bright spot based on a position relationship between the bright spot and the VP line and a color feature of the bright spot, to obtain different lamp source categories; and

perform light switching based on the ambient light brightness value corresponding to the image and the classification result.

7 . The apparatus according to claim 6 , wherein the calculate an ambient light brightness value corresponding to the image comprises:

select at least one region from the image, and calculating a brightness value of the at least one region; and

calculate based on the brightness value of the at least one region the ambient light brightness value corresponding to the image.

8 . The apparatus according to claim 6 , wherein the lamp source detection model is trained by using a plurality of sample images, wherein each of the plurality of sample images comprises a lamp source and annotation information of the lamp source.

9 . The apparatus according to claim 6 , wherein before the perform pairing on the lamp box, the instructions further cause the apparatus to:

perform overlap removal on the lamp box, to ensure that there is no overlap or tangent between lamp boxes.

10 . The apparatus according to claim 6 , wherein the determine a vanishing point horizontal line VP line based on the plurality of lamp box pairs comprises:

map lamp box centers of the lamp box pairs onto rows in the image, to obtain an in-row quantity distribution map of the lamp box pairs;

select a VP line based on the in-row quantity distribution map;

correct the VP line based on a preset correction value, to obtain a corrected VP line; and

adjust the corrected VP line by using a reference value, to obtain a VP line used to classify the bright spot, wherein the reference value is used to describe a change of a pitch angle of a video camera that captured the image.

11 . The method according to claim 1 , wherein the image is captured by a video camera disposed at a fixed position of a vehicle.

12 . The apparatus according to claim 6 , wherein the image is captured by a video camera disposed at a fixed position of a vehicle.

13 . A non-transitory computer-readable medium storing computer instructions, that when executed by one or more processors, cause the one or more processors to perform operations comprising:

obtaining an image, wherein the image comprises lamp source information;

calculating an ambient light brightness value corresponding to the image;

classifying, based on the lamp source information, a lamp source comprised in the image to obtain a classification result, wherein the classification result is obtained by determining a vanishing point horizontal line (VP) line in the image, wherein the classifying the lamp source comprised in the image to obtain the classification result comprises:

inputting the image into a lamp source detection model; and

obtaining a lamp source category in the image based on the lamp source detection model, wherein the obtaining the lamp source category in the image based on the lamp source detection model comprises:

selecting, from the image, a bright spot whose brightness value is greater than a preset threshold, and setting a lamp box for the bright spot;

performing pairing on the lamp box to obtain a plurality of lamp box pairs;

determining the VP line based on the plurality of lamp box pairs, wherein the VP line is used to distinguish between a first region and a second region of the image; and

classifying the bright spot based on a position relationship between the bright spot and the VP line and a color feature of the bright spot, to obtain different lamp source categories; and

performing light switching based on the ambient light brightness value corresponding to the image and the classification result.

14 . The non-transitory computer-readable medium according to claim 13 , wherein the calculating an ambient light brightness value corresponding to the image comprises:

selecting at least one region from the image, and calculating a brightness value of the at least one region; and

calculating, based on the brightness value of the at least one region, the ambient light brightness value corresponding to the image.

15 . The non-transitory computer-readable medium according to claim 13 , wherein the lamp source detection model is trained by using a plurality of sample images, wherein each of the plurality of sample images comprises a lamp source and annotation information of the lamp source.

16 . The non-transitory computer-readable medium according to claim 13 , wherein the operations comprise:

before the performing pairing on the lamp box, performing overlap removal on the lamp box, to ensure that there is no overlap or tangent between lamp boxes.

17 . The non-transitory computer-readable medium according to claim 13 , wherein the determining a vanishing point horizontal line VP line based on the plurality of lamp box pairs comprises:

mapping lamp box centers of the lamp box pairs onto rows in the image, to obtain an in-row quantity distribution map of the lamp box pairs;

selecting a VP line based on the in-row quantity distribution map;

correcting the VP line based on a preset correction value, to obtain a corrected VP line; and

adjusting the corrected VP line by using a reference value, to obtain a VP line used to classify the bright spot, wherein the reference value is used to describe a change of a pitch angle of a video camera that captured the image.

18 . The non-transitory computer-readable medium according to claim 13 , wherein the image is captured by a video camera disposed at a fixed position of a vehicle.

Assignments (3)

Priority Claims (1)

Continuity (2)

References Cited (19)