Granted Patent B2

US 12,293,574 · App. 17/814,461 · Granted May 6, 2025

Inventors: Lei Zhang (Beijing, CN); Kai Yang (Beijing, CN); Qijuan Yin (Beijing, CN); Wuzhao Zhang (Beijing, CN); Xiaoyan Wang (Beijing, CN)

Assignee: APOLLO INTELLIGENT DRIVING TECHNOLOGY (BEIJING) CO., LTD.

G06V10/80G06V10/776G06V20/58

Abstract

Claims (51)

1. A method for detecting obstacle, comprising:

detecting an obstacle in a vehicle traveling process using plural obstacle detection manners to obtain plural initial detection results of the obstacle, the plural initial detection results corresponding to the plural obstacle detection manners;

fusing the plural initial detection results to obtain a fusion result; and

obtaining a final detection result of the obstacle based on the fusion result,

wherein each of the plural initial detection results is used to indicate whether an obstacle exists, and the fusing the plural initial detection results to obtain a fusion result comprises:

selecting an initial detection result used to indicate the existence of the obstacle from the plural initial detection results as a to-be-processed detection result;

obtaining fusion information of the to-be-processed detection result; and

obtaining the fusion result based on the fusion information of the to-be-processed detection result,

wherein there exist plural to-be-processed detection results, and the fusion information of the to-be-processed detection results comprises: accuracy information of each of the plural to-be-processed detection results and relevance information between every two of the plural to-be-processed detection results; and the obtaining the fusion result based on the fusion information of the to-be-processed detection result comprises:

sorting the accuracy information of all the to-be-processed detection results in a descending order to obtain sorted accuracy information; and

obtaining the fusion result using a preset functional relationship based on the sorted accuracy information and the relevance information between the every two to-be-processed detection results, the preset functional relationship being used to indicate a relationship between the fusion result and the sorted accuracy information and the relevance information, wherein the sorted accuracy information comprises two pieces of accuracy information, and in the preset functional relationship, the fusion result is in a direct proportion relationship with one of the two pieces of accuracy information and a product of the other of the two pieces of accuracy information and the relevance information,

wherein the obtaining a final detection result of the obstacle based on the fusion result comprises:

if the fusion information is greater than or equal to a preset threshold, determining that the final detection result is the existence of the obstacle.

2. The method according to claim 1 , wherein the obtaining fusion information of the to-be-processed detection result comprises:

detecting environmental information of the vehicle by a pre-established accuracy detection model to obtain the accuracy information of each to-be-processed detection result; and

obtaining relevance information between every two obstacle detection manners corresponding to the every two to-be-processed detection results from a pre-configured configuration file, and taking the obtained relevance information as the relevance information between the every two to-be-processed detection results.

3. An electronic device, comprising:

at least one processor; and

a memory communicatively connected with the at least one processor;

wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform a method for detecting obstacle, wherein the method comprises:

detecting an obstacle in a vehicle traveling process using plural obstacle detection manners to obtain plural initial detection results of the obstacle, the plural initial detection results corresponding to the plural obstacle detection manners;

fusing the plural initial detection results to obtain a fusion result; and

obtaining a final detection result of the obstacle based on the fusion result,

wherein each of the plural initial detection results is used to indicate whether an obstacle exists, and the fusing the plural initial detection results to obtain a fusion result comprises:

selecting an initial detection result used to indicate the existence of the obstacle from the plural initial detection results as a to-be-processed detection result;

obtaining fusion information of the to-be-processed detection result; and

obtaining the fusion result based on the fusion information of the to-be-processed detection result,

wherein there exist plural to-be-processed detection results, and the fusion information of the to-be-processed detection results comprises: accuracy information of each of the plural to-be-processed detection results and relevance information between every two of the plural to-be-processed detection results; and the obtaining the fusion result based on the fusion information of the to-be-processed detection result comprises:

sorting the accuracy information of all the to-be-processed detection results in a descending order to obtain sorted accuracy information; and

obtaining the fusion result using a preset functional relationship based on the sorted accuracy information and the relevance information between the every two to-be-processed detection results, the preset functional relationship being used to indicate a relationship between the fusion result and the sorted accuracy information and the relevance information, wherein the sorted accuracy information comprises two pieces of accuracy information, and in the preset functional relationship, the fusion result is in a direct proportion relationship with one of the two pieces of accuracy information and a product of the other of the two pieces of accuracy information and the relevance information,

wherein the obtaining a final detection result of the obstacle based on the fusion result comprises:

if the fusion information is greater than or equal to a preset threshold, determining that the final detection result is the existence of the obstacle.

4. The electronic device according to claim 3 , wherein the obtaining fusion information of the to-be-processed detection result comprises:

detecting environmental information of the vehicle by a pre-established accuracy detection model to obtain the accuracy information of each to-be-processed detection result; and

obtaining relevance information between every two obstacle detection manners corresponding to the every two to-be-processed detection results from a pre-configured configuration file, and take the obtained relevance information as the relevance information between the every two to-be-processed detection results.

5. A non-transitory computer readable storage medium with computer instructions stored thereon, wherein the computer instructions are used for causing a method for detecting obstacle, wherein the method comprises:

detecting an obstacle in a vehicle traveling process using plural obstacle detection manners to obtain plural initial detection results of the obstacle, the plural initial detection results corresponding to the plural obstacle detection manners;

fusing the plural initial detection results to obtain a fusion result; and

obtaining a final detection result of the obstacle based on the fusion result,

wherein each of the plural initial detection results is used to indicate whether an obstacle exists, and the fusing the plural initial detection results to obtain a fusion result comprises:

selecting an initial detection result used to indicate the existence of the obstacle from the plural initial detection results as a to-be-processed detection result;

obtaining fusion information of the to-be-processed detection result; and

obtaining the fusion result based on the fusion information of the to-be-processed detection result,

wherein there exist plural to-be-processed detection results, and the fusion information of the to-be-processed detection results comprises: accuracy information of each of the plural to-be-processed detection results and relevance information between every two of the plural to-be-processed detection results; and the obtaining the fusion result based on the fusion information of the to-be-processed detection result comprises:

sorting the accuracy information of all the to-be-processed detection results in a descending order to obtain sorted accuracy information; and

obtaining the fusion result using a preset functional relationship based on the sorted accuracy information and the relevance information between the every two to-be-processed detection results, the preset functional relationship being used to indicate a relationship between the fusion result and the sorted accuracy information and the relevance information, wherein the sorted accuracy information comprises two pieces of accuracy information, and in the preset functional relationship, the fusion result is in a direct proportion relationship with one of the two pieces of accuracy information and a product of the other of the two pieces of accuracy information and the relevance information,

wherein the obtaining a final detection result of the obstacle based on the fusion result comprises:

if the fusion information is greater than or equal to a preset threshold, determining that the final detection result is the existence of the obstacle.

6. The non-transitory computer readable storage medium according to claim 5 , the obtaining fusion information of the to-be-processed detection result comprises:

detecting environmental information of the vehicle by a pre-established accuracy detection model to obtain the accuracy information of each to-be-processed detection result; and

obtaining relevance information between every two obstacle detection manners corresponding to the every two to-be-processed detection results from a pre-configured configuration file, and taking the obtained relevance information as the relevance information between the every two to-be-processed detection results.

Assignments (1)

Priority Claims (1)

Continuity (1)

References Cited (64)