CN-121997101-A - Target classification method of vehicle and electronic equipment

CN121997101ACN 121997101 ACN121997101 ACN 121997101ACN-121997101-A

Abstract

The embodiment of the application provides a target classification method and electronic equipment of a vehicle, and the method comprises the steps of controlling a plurality of sensors on the vehicle to respectively sense driving conditions in front of the vehicle to obtain a plurality of sensor data, controlling the plurality of sensors to perform pre-classification processing on at least one target object in the plurality of sensor data to obtain pre-classification results respectively corresponding to the plurality of sensor data, determining scene condition information of the at least one target object based on the plurality of sensor data, wherein the scene condition information is used for representing scene conditions of the at least one target object, determining a target weight coefficient corresponding to the pre-classification result based on the scene condition information and/or the pre-classification result, and performing weighted fusion on the pre-classification result based on the target weight coefficient to obtain the target classification result of the at least one target object. The method and the device solve the technical problem that the accuracy of the target classification of the vehicle in the related art is low.

Inventors

YANG SIYI
FENG TAO
WANG WENLI
HUANG QIUSHENG

Assignees

奇瑞汽车股份有限公司

Dates

Publication Date: 20260508
Application Date: 20260130

Claims (10)

1. A method of classifying objects of a vehicle, comprising: in the running process of a vehicle, controlling a plurality of sensors on the vehicle to respectively sense the running condition in front of the vehicle to obtain a plurality of sensor data; Controlling the plurality of sensors, and performing pre-classification processing on at least one target object in the plurality of sensor data to obtain pre-classification results respectively corresponding to the plurality of sensor data; determining scene condition information of the at least one target object based on the plurality of sensor data, wherein the scene condition information is used for representing scene conditions of the at least one target object; Determining a target weight coefficient corresponding to the pre-classification result based on the scene condition information and/or the pre-classification result; And carrying out weighted fusion on the pre-classification result based on the target weight coefficient to obtain a target classification result of the at least one target object.
2. The method according to claim 1, wherein controlling the plurality of sensors, and performing a pre-classification process on at least one target object in the plurality of sensor data to obtain pre-classification results respectively corresponding to the plurality of sensor data, includes: And controlling the plurality of sensors, and performing pre-classification processing on the at least one target object in the plurality of sensor data based on a preset pre-training algorithm corresponding to the plurality of sensors respectively to obtain the pre-classification result, wherein the preset pre-training algorithm is determined in advance based on the sensor types of the plurality of sensors, and each pre-classification result comprises the pre-classification type of the at least one target object and the confidence of the pre-classification type.
3. The method according to claim 1, wherein determining a target weight coefficient corresponding to the pre-classification result based on the scene condition information and/or the pre-classification result includes: Based on the scene condition information and/or the pre-classification result, adjusting initial weight coefficients corresponding to the plurality of sensors to obtain the target weight coefficients, wherein the initial weight coefficients are determined in advance based on the sensor types of the plurality of sensors; preferably, the initial vision weight coefficient corresponding to the vision sensor in the plurality of sensors is larger than the initial laser radar weight coefficient corresponding to the laser radar in the plurality of sensors.
4. The method according to claim 3, wherein adjusting initial weight coefficients corresponding to the plurality of sensors based on the scene condition information to obtain the target weight coefficients comprises: And under the condition that the scene condition information indicates that the visual observability of the at least one target object is larger than a preset observability threshold value, improving the initial visual weight coefficient in the initial weight coefficient to obtain the target weight coefficient.
5. The method for classifying a vehicle according to claim 3, wherein adjusting initial weight coefficients corresponding to the plurality of sensors based on the pre-classification result to obtain the target weight coefficients comprises: And under the condition that the laser radar pre-classification result in the pre-classification result indicates that the at least one target object is a traffic participant, increasing the initial laser radar weight coefficient in the initial weight coefficient to obtain the target weight coefficient.
6. The method of classifying objects of a vehicle according to claim 1, characterized in that the method further comprises: Determining moving speed, altitude information and region information of the at least one target object based on laser radar data in the plurality of sensor data, wherein the region information is used for indicating whether the at least one target object is in a road region where the vehicle runs; and determining a laser radar pre-classification result in the pre-classification results based on the moving speed, the altitude information and the area information.
7. The method of object classification for a vehicle according to claim 6, wherein determining a lidar pre-classification result of the pre-classification results based on the movement speed, the altitude information, and the location information comprises: Determining an initial laser radar pre-classification result of at least one target object based on the moving speed, the height information and the region information; determining the number of clustered obstacles corresponding to the at least one target object in the laser radar data; based on the number of the clustered obstacles, adjusting the initial laser radar pre-classification result to obtain the laser radar pre-classification result; Preferably, the initial lidar pre-classification result is adjusted based on the number of clustered obstacles to obtain the lidar pre-classification result, and the method comprises the step of determining that the lidar pre-classification result is a non-traffic participant when the initial lidar pre-classification result is a traffic participant and the number of clustered obstacles is greater than a preset number.
8. The object classification method of a vehicle according to any one of claims 1 to 7, characterized in that the method further comprises: And in the case that the target classification result of the at least one target object is a traffic participant and the at least one target object is not in a road area where the vehicle runs, adjusting the target classification result of the at least one target object to be a non-traffic participant.
9. The object classification method of a vehicle according to any one of claims 1 to 7, characterized in that the method further comprises: And determining the target classification result of the at least one target object based on the historical target classification result of the at least one target object under the condition that the confidence corresponding to the target classification result of the at least one target object is smaller than the preset confidence.
10. An electronic device, comprising: A memory storing an executable program; A processor for executing the program, wherein the program executes the object classification method of the vehicle according to any one of claims 1 to 9 when executed.

Description

Target classification method of vehicle and electronic equipment Technical Field The embodiment of the application relates to the technical field of vehicles, in particular to a target classification method of a vehicle and electronic equipment. Background Along with the development of an intelligent driving system, target classification is carried out, and the intelligent driving system becomes an important technology for supporting the safe operation of a vehicle decision-making module. However, the conventional target classification technology faces multiple challenges and limitations, and mainly is characterized in that the problem of efficiency and cost conflict exists in multi-sensor fusion classification, and data-level or feature-level fusion can make up for a short plate of a single sensor, but the processing of a large amount of original data or high-order features causes the aggravation of calculation delay, and the dependence on a high-power chip increases the hardware cost, so that the technology is difficult to popularize and apply in economic vehicles. Therefore, the accuracy of classifying the targets by the vehicle in the related art is low. Disclosure of Invention The embodiment of the application provides a target classification method of a vehicle and electronic equipment, which are used for at least solving the technical problem that the accuracy of target classification of the vehicle in the related technology is low. According to one aspect of the embodiment of the application, a target classification method of a vehicle is provided, and the method comprises the steps of controlling a plurality of sensors on the vehicle to respectively sense driving conditions in front of the vehicle to obtain a plurality of sensor data in the driving process of the vehicle, controlling the plurality of sensors to perform pre-classification processing on at least one target object in the plurality of sensor data to obtain pre-classification results respectively corresponding to the plurality of sensor data, determining scene condition information of the at least one target object based on the plurality of sensor data, wherein the scene condition information is used for representing scene conditions of the at least one target object, determining a target weight coefficient corresponding to the pre-classification result based on the scene condition information and/or the pre-classification result, and performing weighted fusion on the pre-classification result based on the target weight coefficient to obtain the target classification result of the at least one target object. In the embodiment of the application, the method comprises the steps of controlling a plurality of sensors, and performing pre-classification processing on at least one target object in the plurality of sensor data to obtain pre-classification results respectively corresponding to the plurality of sensor data, wherein the method comprises the steps of controlling the plurality of sensors, performing pre-classification processing on at least one target object in the plurality of sensor data based on a preset pre-training algorithm respectively corresponding to the plurality of sensors to obtain pre-classification results, and the preset pre-training algorithm is determined in advance based on the sensor types of the plurality of sensors, and each pre-classification result comprises the pre-classification type and the confidence of the pre-classification type of at least one target object. In the embodiment of the application, the target weight coefficient corresponding to the pre-classification result is determined based on the scene condition information and/or the pre-classification result, which comprises the steps of adjusting initial weight coefficients corresponding to a plurality of sensors based on the scene condition information and/or the pre-classification result to obtain the target weight coefficient, wherein the initial weight coefficient is determined in advance based on the sensor types of the plurality of sensors, and preferably, the initial visual weight coefficient corresponding to the visual sensor in the plurality of sensors is larger than the initial laser radar weight coefficient corresponding to the laser radar in the plurality of sensors. In the embodiment of the application, the initial weight coefficients corresponding to the plurality of sensors are adjusted based on scene condition information to obtain target weight coefficients, wherein the method comprises the step of improving the initial visual weight coefficients in the initial weight coefficients to obtain the target weight coefficients when the scene condition information indicates that the visual observability of at least one target object is larger than a preset observability threshold. In the embodiment of the application, the initial weight coefficients corresponding to the plurality of sensors are adjusted based on the pre-classification resu