CN-121979287-A - Unmanned aerial vehicle automatic identification tracking system and method

CN121979287ACN 121979287 ACN121979287 ACN 121979287ACN-121979287-A

Abstract

The invention discloses an unmanned aerial vehicle automatic identification tracking system and method, which belong to the technical field of photoelectric automatic zooming, and comprise the steps of detecting a target object by adopting a detection module to obtain an azimuth angle, a pitch angle and a distance of the target object, adjusting the azimuth angle and the pitch angle of the identification module through a turntable to enable the azimuth angle and the pitch angle of the identification module and the target object to be within a set deviation range, automatically zooming and automatically focusing a coarse television of the identification module to enable the target object to be in a view field of the coarse television, roughly tracking the target object, and finely tracking the target object. According to the unmanned aerial vehicle automatic tracking and identifying method, the preliminary detection of the target object, the coarse tracking and the fine tracking of the target object are connected in series, and each step realizes automatic processing, so that the operation flow and the control time of the automatic tracking of the target object are greatly simplified, the manual intervention is avoided, and the identifying efficiency and the accuracy of the target object are greatly improved.

Inventors

JIAO YULIANG
LIU YIPENG
LIU XU
WANG QI
PI LI
HUANG ZONGJUN
QI SHUXIAN

Assignees

湖北华中长江光电科技有限公司

Dates

Publication Date: 20260505
Application Date: 20251231

Claims (10)

1. The unmanned aerial vehicle automatic identification tracking method is characterized by adopting an unmanned aerial vehicle automatic identification tracking system to carry out tracking identification, wherein the unmanned aerial vehicle automatic identification tracking system comprises a detection module and an identification module, and the identification module adjusts the azimuth and the angle through a turntable, and the unmanned aerial vehicle automatic identification tracking method is characterized by comprising the following steps: S1, detecting a target object by adopting a detection module to obtain an azimuth angle, a pitch angle and a distance of the target object relative to an unmanned aerial vehicle automatic identification tracking system; S2, adjusting the azimuth angle and the pitch angle of the identification module through the turntable, enabling the azimuth angle and the pitch angle of the identification module and the target object to be in a set deviation range, and automatically zooming and automatically focusing the rough television of the identification module, so that the target object is in the view field of the rough television; s3, roughly tracking the target object; the method comprises the steps of identifying a target object by a coarse image plate of a coarse television, obtaining the pixel deviation of the target object in the center of an image cross of the coarse image plate, so that the pixel deviation is effective; S4, carrying out fine tracking on the target object; the method comprises the steps of calibrating a linear module value of an identification module based on the distance and the external temperature of a target object, automatically focusing a precision television by adopting the linear module, obtaining the pixel deviation amount of the target object at the center of an image cross of a precision image plate, enabling the pixel deviation amount to be effective, and enabling the identification module to perform precision tracking by adjusting a quick reflector, so that the precision tracking reaches a stable state.
2. The unmanned aerial vehicle automatic identification tracking method according to claim 1, wherein the coarse television of the identification module in step S2 performs automatic zooming and automatic focusing based on a zooming model and a focusing model, and the zooming model is: (equation 1) Wherein the said For identifying the linear distance between the module and the target object A k preset distance interval; Is the linear scaling factor of the k-th interval, The linear fit pitch for the k-th interval.
3. The unmanned aerial vehicle automatic identification tracking method of claim 2, wherein the focusing model is: (equation 2) Wherein the said Is the kth temperature interval, the A focusing proportion coefficient of the kth interval, the focusing proportion coefficient is the k-th interval The pitch is fitted for the focus of the kth interval.
4. The unmanned aerial vehicle automatic tracking and recognition method according to claim 1, wherein in the step S4, the straight line module adopts a focusing model to perform automatic focusing, and the focusing model of the straight line module is: (equation 3) Wherein the said The value is a linear module value, the a, the b and the c are parameters to be solved, the t is a temperature value, and the The stability correction value is obtained according to the temperature change.
5. The unmanned aerial vehicle automatic tracking and identifying method according to claim 1, wherein the pixel deviation amounts in the step S3 and the step S4 of coarse tracking and fine tracking effectively refer to that the mean square error of the pixel deviation of the horizontal axis and the vertical axis of the target object is within a set threshold, and the pixel deviation mean square error of the horizontal axis and the vertical axis of the target object is calculated by the following steps: (equation 4) Wherein the said The mean square error of pixel deviation of the horizontal axis and the vertical axis of the target object is obtained, N is the number of current pixel deviation groups counted in real time, and N is the number of current pixel deviation groups counted in real time Is a horizontal axis pixel deviation value Or vertical axis pixel offset value The said Mean value of N groups of pixel deviations 。
6. The method according to claim 1, wherein in the step S3, whether the unmanned aerial vehicle automatic recognition tracking system can stably track the target object is determined by simultaneously determining in real time whether the azimuth angle and the pitch angle of the current target object are within a set interval; The judging mode of whether the target azimuth angle and the pitch angle are in a set interval is as follows: (equation 5) (Equation 6) Wherein the said Representing the orientation of the object reported by the servo turntable, said Representing the pitch angle of the target object reported by the servo turntable Representing the orientation of the object reported by the detection module, the Representing the pitch angle of the target object reported by the detection module; the said Representing the set angle deviation.
7. The method for automatically identifying and tracking the unmanned aerial vehicle according to claim 1, wherein in the step S2, the target object is placed in an optimal field of view of the identification module, and the optimal field of view of the identification module is obtained by: (equation 7) Wherein, the For the best view angle of recognition, For the minimum number of identified pixels, For the size of the picture element, For a fixed focal length, the minimum identification pixel is set Obtaining the pixel size of the identification module And after a fixed focal length The optimal recognition visual angle can be calculated.
8. The unmanned aerial vehicle automatic tracking and recognition method according to claim 1, wherein in the step S2, the adjusting the azimuth angle and the pitch angle of the recognition module by the turntable further comprises: when no object exists in the field of view of the identification module, the identification module is adopted to perform fan scanning until the object falls into the field of view of the identification module.
9. The unmanned aerial vehicle automatic tracking and identification method of claim 1, further comprising: S5, continuously tracking the target object, marking the target object by adopting a marking module, and closing the continuous tracking of the target object.
10. An unmanned aerial vehicle automatic tracking recognition system for executing the unmanned aerial vehicle automatic tracking recognition method according to any one of claims 1 to 9, comprising: the control module is respectively in communication connection with the detection module, the identification module and the marking module; the detection module is used for searching the target object to obtain the azimuth angle, the pitch angle and the distance of the target object; the identification module adjusts the azimuth through the servo turntable and is used for carrying out rough tracking and fine tracking on the target object; The marking module is used for marking the target object.

Description

Unmanned aerial vehicle automatic identification tracking system and method Technical Field The invention belongs to the technical field of photoelectric automatic zooming, and particularly relates to an unmanned aerial vehicle automatic identification tracking system and method. Background With the continuous progress of technology, the demand of the photoelectric automatic zooming technology in various applications is increasing. The photoelectric automatic zooming technology is mainly used for capturing and tracking a moving target object in real time so as to ensure that the target object is stably tracked in the center of an image cross and kept clear, and is widely applied to scenes such as security and protection turntables, unmanned aerial vehicle pods, field investigation and the like. In a conventional photoelectric system, image recognition and tracking generally require a plurality of instructions to be sent manually, and the purpose of tracking a target can be achieved by continuously controlling the photoelectric device. In a complex photoelectric device, the tracking of the photoelectric system is only one function, and when many functions of a large system are controlled by manual instructions, the problems of complexity, long time consumption and the like of manual instruction input may affect the normal operation of the photoelectric system, so that the predetermined function cannot be completed. Moreover, based on the integration of the existing photoelectric system, more and more functions are integrated on the photoelectric system, and the normal use of the photoelectric system cannot be adapted only by adopting a manual control mode to complete image identification and tracking. Disclosure of Invention Aiming at one or more of the defects or improvement demands in the prior art, the invention provides an unmanned aerial vehicle automatic identification tracking method, which is used for solving the problems that the existing photoelectric identification tracking depends on a manual command, and the photoelectric identification tracking operation is complicated and cannot normally run due to the complexity and long time consumption of the manual command. In order to achieve the above purpose, the invention provides an unmanned aerial vehicle automatic identification tracking method, which adopts an unmanned aerial vehicle automatic identification tracking system to carry out tracking identification, wherein the unmanned aerial vehicle automatic identification tracking system comprises a detection module and an identification module, the identification module adjusts the azimuth and the angle through a turntable, and the method comprises the following steps: S1, detecting a target object by adopting a detection module to obtain an azimuth angle, a pitch angle and a distance of the target object relative to an unmanned aerial vehicle automatic identification tracking system; S2, adjusting the azimuth angle and the pitch angle of the identification module through the turntable, enabling the azimuth angle and the pitch angle of the identification module and the target object to be in a set deviation range, and automatically zooming and automatically focusing the rough television of the identification module, so that the target object is in the view field of the rough television; s3, roughly tracking the target object; the method comprises the steps of identifying a target object by a coarse image plate of a coarse television, obtaining the pixel deviation of the target object in the center of an image cross of the coarse image plate, so that the pixel deviation is effective; S4, carrying out fine tracking on the target object; the method comprises the steps of calibrating a linear module value of an identification module based on the distance and the external temperature of a target object, automatically focusing a precision television by adopting the linear module, obtaining the pixel deviation amount of the target object at the center of an image cross of a precision image plate, enabling the pixel deviation amount to be effective, and enabling the identification module to perform precision tracking by adjusting a quick reflector, so that the precision tracking reaches a stable state. As a further improvement of the present invention, the coarse television of the identification module in step S2 performs automatic zooming and focusing based on a zooming model and a focusing model, where the zooming model is: (equation 1) Wherein the saidFor identifying the linear distance between the module and the target objectA k preset distance interval; Is the linear scaling factor of the k-th interval, The linear fit pitch for the k-th interval. As a further improvement of the present invention, the focusing model is: (equation 2) Wherein the saidIs the kth temperature interval, theA focusing proportion coefficient of the kth interval, the focusing proportion coefficient is the k-th intervalThe pitch