CN-116246253-B - Target association method, device, computer equipment and storage medium

Abstract

The present application relates to the field of object detection technologies, and in particular, to a method, an apparatus, a computer device, and a storage medium for object association. The method comprises the steps of extracting a first target object from first sensing data acquired by a first sensor and extracting a second target object from second sensing data acquired by a second sensor, adjusting a reference association gate according to a first distance between the first target object and the first sensor and a second distance between the second target object and the second sensor to obtain a dynamic association gate, and determining whether the first target object and the second target object belong to the same object or not based on the dynamic association gate. The method and the device can improve the accuracy of target association.

Inventors

• ZHANG ZHENLIN
• Cui Zhanshi
• CHEN YINZI

Assignees

• 中汽创智科技有限公司

Dates

Publication Date

20260508

Application Date

20230314

Claims (10)

1. 1. A method of target association, the method comprising: Extracting a first target object from first sensing data acquired by a first sensor, and extracting a second target object from second sensing data acquired by a second sensor; According to a first distance between the first target object and the first sensor and a second distance between the second target object and the second sensor, a reference correlation gate is adjusted to obtain a dynamic correlation gate, wherein the reference correlation gate is used for describing the maximum allowable deviation of the first target object and the second target object in a specified dimension; determining a deviation result between the object characteristic information of the first target object and the object characteristic information of the second target object; determining a target association degree between the first target object and the second target object under the condition that the deviation result falls into the dynamic association gate; And determining whether the first target object and the second target object belong to the same object according to whether the target association degree meets a preset association condition.
2. 2. The method of claim 1, wherein adjusting the reference correlation gate based on a first distance between the first target object and the first sensor and a second distance between the second target object and the second sensor results in a dynamic correlation gate, comprising: At least one of a position error correlation gate, a size error correlation gate and a speed error correlation gate is selected as a reference correlation gate; determining an observation distance according to a first distance between the first target object and the first sensor and a second distance between the second target object and the second sensor; And adjusting the reference correlation gate according to the observation distance and the reference distance corresponding to the reference correlation gate to obtain the dynamic correlation gate.
3. 3. The method according to claim 2, wherein the adjusting the reference correlation gate according to the observed distance and the reference distance corresponding to the reference correlation gate to obtain the dynamic correlation gate includes: Taking the ratio between the observed distance and the reference distance as an adjustment factor; And scaling the reference association gate based on the adjustment factor to obtain the dynamic association gate.
4. 4. The method according to claim 1, wherein determining whether the first target object and the second target object belong to the same object according to whether the target association degree satisfies a preset association condition comprises: If the target association degree meets a preset association condition, determining an error area based on the identification frame of the first target object and the identification frame of the second target object; and if no other objects except the first target object and the second target object exist in the error area, determining that the first target object and the second target object belong to the same object.
5. 5. The method according to claim 1, wherein determining whether the first target object and the second target object belong to the same object according to whether the target association degree satisfies a preset association condition comprises: if the target association degree meets a preset association condition, verifying whether the first target object and the second target object have the blocked object or not; And if the first target object and the second target object do not have the blocked object, determining that the first target object and the second target object belong to the same object.
6. 6. The method according to claim 4, wherein the method further comprises: Determining a first reference association degree between the first target object and a second reference object, wherein the second reference object is other objects extracted from the second sensing data except the second target object; Determining a second reference association degree between the second target object and a first reference object, wherein the first reference object is other objects extracted from the first sensing data except the first target object; and if the target association degrees are smaller than the first reference association degrees and the second reference association degrees, determining that the target association degrees meet a preset association condition.
7. 7. The method of claim 1, wherein extracting the first target object from the first sensed data acquired by the first sensor and extracting the second target object from the second sensed data acquired by the second sensor comprises: Extracting a first candidate object from the first sensing data and extracting a second candidate object from the second sensing data; and screening the first candidate object and the second candidate object based on a fixed association gate to obtain the first target object and the second target object.
8. 8. An object association apparatus, the apparatus comprising: the acquisition module is used for extracting a first target object from first sensing data acquired by the first sensor and extracting a second target object from second sensing data acquired by the second sensor; the adjustment module is used for adjusting the reference association gate according to a first distance between the first target object and the first sensor and a second distance between the second target object and the second sensor to obtain a dynamic association gate, wherein the reference association gate is used for describing the maximum allowable deviation of the first target object and the second target object in a specified dimension; The association module is used for determining a deviation result between the object characteristic information of the first target object and the object characteristic information of the second target object, determining a target association degree between the first target object and the second target object under the condition that the deviation result falls into the dynamic association door, and determining whether the first target object and the second target object belong to the same object according to whether the target association degree meets a preset association condition.
9. 9. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the steps of the method of any of claims 1 to 7 when the computer program is executed.
10. 10. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 7.

Description

Target association method, device, computer equipment and storage medium Technical Field The application relates to the technical field of artificial intelligence, in particular to the technical fields of automatic driving, intelligent traffic and target detection, and in particular relates to a target association method, a device, computer equipment and a storage medium. Background The automatic driving automobile covers modules such as perception, decision making, control and the like, wherein environmental perception relates to a plurality of different sensors, such as cameras, millimeter wave radars, laser radars and the like, and is the basis for performing target detection tasks in an automatic driving scene. In the case of target tracking based on targets detected by the respective sensors, a data correlation algorithm is generally adopted, and for the same target, measurements formed on the respective sensors must have some similar characteristic due to the same physical source, but due to noise interference and instability of the performance of the sensors themselves, the measured characteristics are not exactly the same, and the purpose of data correlation is to use the similar characteristic of the measurements to determine whether the measured data originate from the same target. One of the logical criteria for data correlation is that each measurement observation can only be correlated with one other measurement, but not two or more other measurements at the same time. However, the detection accuracy of the sensor is affected by the detection environment, and large fluctuation occurs, so that the algorithm is difficult to accurately correlate targets tracked by two sensors at the same time. Disclosure of Invention In view of the foregoing, it is desirable to provide a target association method, apparatus, computer device, and storage medium capable of target association accuracy. In a first aspect, the present application provides a target association method, the method comprising: Extracting a first target object from first sensing data acquired by a first sensor, and extracting a second target object from second sensing data acquired by a second sensor; according to a first distance between a first target object and a first sensor and a second distance between a second target object and a second sensor, adjusting a reference association door to obtain a dynamic association door; Based on the dynamic association gate, it is determined whether the first target object and the second target object belong to the same object. In one embodiment, adjusting the reference correlation gate according to a first distance between the first target object and the first sensor and a second distance between the second target object and the second sensor to obtain a dynamic correlation gate includes: At least one of a position error correlation gate, a size error correlation gate and a speed error correlation gate is selected as a reference correlation gate; Determining an observation distance according to a first distance between a first target object and a first sensor and a second distance between a second target object and a second sensor; And adjusting the reference correlation gate according to the observation distance and the reference distance corresponding to the reference correlation gate to obtain the dynamic correlation gate. In one embodiment, according to the observation distance and the reference distance corresponding to the reference correlation gate, the reference correlation gate is adjusted to obtain a dynamic correlation gate, which includes: Taking the ratio of the observation distance to the reference distance as an adjustment factor; and scaling the reference association gate based on the adjustment factor to obtain the dynamic association gate. In one embodiment, determining whether the first target and the second target object belong to the same object based on the dynamic association gate includes: Determining a target association degree between the first target object and the second target object based on the dynamic association gate; And determining whether the first target object and the second target object belong to the same object according to whether the target association degree meets a preset association condition. In one embodiment, determining whether the first target object and the second target object belong to the same object according to whether the target association degree meets a preset association condition includes: If the association degree meets the association condition corresponding to the dynamic association threshold, determining the target association degree between the first target object and the second target object; And determining whether the first target object and the second target object belong to the same object according to whether the target association degree meets a preset association condition. In one embodiment, the method further comprises: de