Object tracking method and device and electronic equipment

Abstract

The invention relates to an object tracking method and device and electronic equipment, and the method comprises the steps: recognizing the attribute information of an object under the condition that the object is detected through image collection equipment; determining an initial filtering frame of the object according to the attribute information; determining moving information in the moving process of the object; adjusting the initial filtering frame according to the movement information to obtain a target filtering frame; and tracking the object based on the target filtering frame. Therefore, the pan-tilt camera can timely and accurately track the detected object, and the user experience is improved.

Description

Technical Field

[0001] The present application relates to the field of object tracking technology, and in particular to an object tracking method, device and electronic device. Background Art

[0002] With the development of cameras, pan-tilt cameras have become widely used due to their ability to continuously track detected objects. A pan-tilt camera pre-determines a rectangular frame in the center of the entire camera image as a filter frame. When the object is within the filter frame, the pan-tilt camera remains stationary. When the object leaves the filter frame, the pan-tilt camera rotates to ensure that the object remains within the filter frame, thus achieving continuous tracking of the detected object.

[0003] Currently, during AI detection, gimbal cameras are prone to jitter in the object frame coordinates calculated by the image algorithm due to changes in image lighting and the current algorithm's use of low-bit inference (limited by chip computing power, a low-bit inference algorithm is required to ensure the algorithm model's responsiveness). Furthermore, because the filter frame of existing gimbal cameras is a fixed width and height area, jitter in the filter frame causes the gimbal to jitter along with the object frame coordinates in order to keep the object within the fixed filter frame. This results in the gimbal also jittering along with the object frame coordinates, causing the gimbal camera to fail to track the detected object in a timely manner or even lose the object, severely impacting the user experience. Summary of the Invention

[0004] The present application provides an object tracking method, device and electronic device to solve the technical problem in the prior art that, because the filter frame of the gimbal camera in the prior art is an area of ​​fixed width and height, in order to keep the object within the fixed filter frame, the filter frame jitters, causing the gimbal to jitter along with the object frame coordinates, thereby causing the gimbal camera to fail to track the detected object in a timely manner or even lose the object, seriously affecting the user experience.

[0005] In a first aspect, the present application provides an object tracking method, the method comprising:

[0006] When an object is detected by an image acquisition device, identifying attribute information of the object;

[0007] determining an initial filtering frame of the object according to the attribute information;

[0008] Determining movement information of the object during movement;

[0009] Adjusting the initial filter frame according to the movement information to obtain a target filter frame;

[0010] The object is tracked based on the target filter frame.

[0011] As a possible implementation manner, the attribute information includes an object type and an object size of the object, and determining an initial filtering frame of the object based on the attribute information includes:

[0012] Determining a type coefficient corresponding to the object according to the object type;

[0013] Multiplying the object width value included in the object size by the type coefficient and the preset size coefficient to obtain an initial width value of the initial filter frame;

[0014] Multiplying the object height value included in the object size by the type coefficient and the size coefficient to obtain an initial height value of the initial filter frame;

[0015] An initial filter frame of the object is determined according to the initial width value and the initial height value.

[0016] As a possible implementation manner, determining the initial filter frame of the object according to the initial width value and the initial height value includes:

[0017] Taking the center point of the shooting picture of the image acquisition device as the center point of the initial filter frame, and determining the center point coordinates of the center point of the initial filter frame in the shooting picture;

[0018] Determining initial vertex coordinates of the filter frame according to the center point coordinates, and the initial width value and initial height value of the initial filter frame;

[0019] A filter frame corresponding to the initial vertex coordinates, the initial width value, and the initial height value is determined as the initial filter frame.

[0020] As a possible implementation manner, determining the movement information of the object during its movement includes:

[0021] Acquire historical object frame coordinates and historical time points of the object detected within a preset historical time period, wherein the historical object frame coordinates include vertex coordinates, height value, and width value of the historical object frame;

[0022] Determining, based on the historical object frame coordinates and the historical time points, a historical speed and a historical acceleration of the object corresponding to each historical time point, to obtain a plurality of historical speeds and a plurality of historical accelerations;

[0023] determining an average speed of the plurality of historical speeds, and an average acceleration of the plurality of historical accelerations;

[0024] determining the average speed as the speed of the object during the movement, and determining the average acceleration as the acceleration of the object during the movement;

[0025] The velocity and the acceleration are determined as movement information of the object during its movement.

[0026] As a possible implementation, adjusting the initial filter frame according to the movement information to obtain a target filter frame includes:

[0027] determining a velocity variance of the object based on a plurality of historical velocities and the velocity;

[0028] determining an acceleration variance of the object based on a plurality of historical accelerations and the acceleration;

[0029] Determining a velocity weight of the object according to the velocity, the velocity variance, the acceleration, and the acceleration variance; wherein the velocity, the velocity variance, the acceleration, and the acceleration variance are negatively correlated with the velocity weight;

[0030] The initial filter frame is adjusted according to the speed weight to obtain a target filter frame.

[0031] As a possible implementation manner, determining the velocity weight of the object according to the velocity, the velocity variance, the acceleration, and the acceleration variance includes:

[0032] Matching the speed and the speed variance with a preset speed interval to obtain a target speed interval to which the speed and the speed variance belong;

[0033] Determining a first speed weight corresponding to the target speed interval from a correspondence between preset speed intervals and speed weights;

[0034] Matching the acceleration and the acceleration variance with a preset acceleration interval to obtain a target acceleration interval to which the acceleration and the acceleration variance belong;

[0035] Determining a second speed weight corresponding to the target acceleration interval from a preset correspondence between acceleration intervals and speed weights;

[0036] A weighted sum calculation is performed on the first speed weight and the second speed weight to obtain a speed weight of the object.

[0037] As a possible implementation, adjusting the initial filter frame according to the speed weight to obtain a target filter frame includes:

[0038] Multiplying the initial width value and the initial height value corresponding to the initial filter frame by the speed weight respectively to obtain a target width value and a target height value;

[0039] Taking the center point of the initial filter frame as the center point of the target filter frame, and determining the target center coordinates of the center point of the target filter frame in the shooting picture of the image acquisition device;

[0040] Determining target vertex coordinates of the target filter frame according to the target center coordinates, and the target width value and the target height value of the target filter frame;

[0041] A filter frame corresponding to the target vertex coordinates, the target width value, and the target height value is determined as a target filter frame.

[0042] As a possible implementation, the image acquisition device is a pan-tilt camera, and tracking the object based on the target filter frame includes:

[0043] In the process of tracking the object, obtaining the object frame coordinates of the object, the object frame coordinates including the vertex coordinates, height value, and width value of the object frame;

[0044] Determining whether the object is completely within the target filter frame according to the object frame coordinates;

[0045] When it is determined that the object is not completely within the target filtering frame, the pan-tilt camera is controlled to move in the direction of the object so that the object is completely within the target filtering frame.

[0046] In a second aspect, an embodiment of the present application provides an object tracking device, the device comprising:

[0047] An identification module, configured to identify attribute information of an object when the object is detected by the image acquisition device;

[0048] A first determining module, configured to determine an initial filtering frame of the object according to the attribute information;

[0049] A second determining module is used to determine movement information of the object during movement;

[0050] An adjustment module, configured to adjust the initial filter frame according to the movement information to obtain a target filter frame;

[0051] A tracking module is used to track the object based on the target filtering frame.

[0052] As a possible implementation manner, the attribute information includes the object type and object size of the object, and the first determining module includes:

[0053] A first determining submodule, configured to determine a type coefficient corresponding to the object according to the object type;

[0054] A second determining submodule is configured to multiply the object width value included in the object size by the type coefficient and a preset size coefficient to obtain an initial width value of the initial filtering frame;

[0055] A third determining submodule is configured to multiply the object height value included in the object size by the type coefficient and the size coefficient to obtain an initial height value of the initial filtering frame;

[0056] The fourth determining submodule is configured to determine an initial filtering frame of the object according to the initial width value and the initial height value.

[0057] As a possible implementation, the third determining submodule is specifically configured to:

[0058] Taking the center point of the shooting picture of the image acquisition device as the center point of the initial filter frame, and determining the center point coordinates of the center point of the initial filter frame in the shooting picture;

[0059] Determining initial vertex coordinates of the filter frame according to the center point coordinates, and the initial width value and initial height value of the initial filter frame;

[0060] A filter frame corresponding to the initial vertex coordinates, the initial width value, and the initial height value is determined as the initial filter frame.

[0061] As a possible implementation, the second determining module is specifically configured to:

[0062] Acquire historical object frame coordinates and historical time points of the object detected within a preset historical time period, wherein the historical object frame coordinates include vertex coordinates, height value, and width value of the historical object frame;

[0063] Determining, based on the historical object frame coordinates and the historical time points, a historical speed and a historical acceleration of the object corresponding to each historical time point, to obtain a plurality of historical speeds and a plurality of historical accelerations;

[0064] determining an average speed of the plurality of historical speeds, and an average acceleration of the plurality of historical accelerations;

[0065] determining the average speed as the speed of the object during the movement, and determining the average acceleration as the acceleration of the object during the movement;

[0066] The velocity and the acceleration are determined as movement information of the object during its movement.

[0067] As a possible implementation, the adjustment module includes:

[0068] a fifth determining submodule, configured to determine a velocity variance of the object based on a plurality of historical velocities and the velocity;

[0069] a sixth determining submodule, configured to determine an acceleration variance of the object based on a plurality of historical accelerations and the acceleration;

[0070] a seventh determination submodule, configured to determine a velocity weight of the object based on the velocity, the velocity variance, the acceleration, and the acceleration variance; wherein the velocity, the velocity variance, the acceleration, and the acceleration variance are negatively correlated with the velocity weight;

[0071] The adjustment submodule is configured to adjust the initial filter frame according to the speed weight to obtain a target filter frame.

[0072] As a possible implementation, the seventh determining submodule is specifically configured to:

[0073] Matching the speed and the speed variance with a preset speed interval to obtain a target speed interval to which the speed and the speed variance belong;

[0074] Determining a first speed weight corresponding to the target speed interval from a correspondence between preset speed intervals and speed weights;

[0075] Matching the acceleration and the acceleration variance with a preset acceleration interval to obtain a target acceleration interval to which the acceleration and the acceleration variance belong;

[0076] Determining a second speed weight corresponding to the target acceleration interval from a preset correspondence between acceleration intervals and speed weights;

[0077] A weighted sum calculation is performed on the first speed weight and the second speed weight to obtain a speed weight of the object.

[0078] As a possible implementation, the adjustment submodule is specifically configured to:

[0079] Multiplying the initial width value and the initial height value corresponding to the initial filter frame by the speed weight respectively to obtain a target width value and a target height value;

[0080] Taking the center point of the initial filter frame as the center point of the target filter frame, and determining the target center coordinates of the center point of the target filter frame in the shooting picture of the image acquisition device;

[0081] Determining target vertex coordinates of the target filter frame according to the target center coordinates, and the target width value and the target height value of the target filter frame;

[0082] A filter frame corresponding to the target vertex coordinates, the target width value, and the target height value is determined as a target filter frame.

[0083] As a possible implementation, the tracking module is specifically configured to:

[0084] In the process of tracking the object, obtaining the object frame coordinates of the object, the object frame coordinates including the vertex coordinates, height value, and width value of the object frame;

[0085] Determining whether the object is completely within the target filter frame according to the object frame coordinates;

[0086] When it is determined that the object is not completely within the target filtering frame, the pan-tilt camera is controlled to move in the direction of the object so that the object is completely within the target filtering frame.

[0087] In a third aspect, an embodiment of the present application provides an electronic device, comprising: a processor and a memory, wherein the processor is configured to execute an object tracking program stored in the memory to implement the object tracking method described in any one of the first aspects.

[0088] The technical solution provided by the embodiment of the present application is to identify the attribute information of the object when the object is detected by an image acquisition device, determine the initial filter frame of the object based on the attribute information, determine the movement information of the object during the movement, adjust the initial filter frame based on the movement information, obtain the target filter frame, and track the above-mentioned object based on the target filter frame. This technical solution, when the object is detected by an image acquisition device, dynamically determines the filter frame based on the attribute information of the object and the movement information during the movement, and tracks the detected object based on the dynamically determined filter frame. It can ensure that when the detected object target shakes, the object is always within the dynamically determined filter frame, thereby preventing the pan-tilt camera from shaking along with the object frame coordinates, and realizing that the pan-tilt camera can move smoothly during the process of tracking the moving object, so that the pan-tilt camera can track the detected object in a timely and accurate manner, thereby improving the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0089] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0090] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0091] One or more embodiments are exemplarily illustrated by pictures in the corresponding drawings. These exemplifications do not constitute limitations on the embodiments. Elements with the same reference numerals in the drawings are represented as similar elements. Unless otherwise stated, the figures in the drawings do not constitute proportional limitations.

[0092] Figure 1 A flowchart of an embodiment of an object tracking method provided in an embodiment of the present application;

[0093] Figure 2 A flowchart of another object tracking method according to an embodiment of the present application;

[0094] Figure 3 A flowchart of another object tracking method according to an embodiment of the present application;

[0095] Figure 4 A schematic diagram of object tracking provided in an embodiment of the present application;

[0096] Figure 5 A schematic diagram of object detection provided in an embodiment of the present application;

[0097] Figure 6 A block diagram of an embodiment of an object tracking device provided in an embodiment of the present application;

[0098] Figure 7 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0099] To make the purpose, technical solutions, and advantages of the embodiments of this application more clear, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0100] The disclosure below provides many different embodiments or examples for implementing different configurations of the present invention. To simplify the disclosure of the present invention, the components and configurations of specific examples are described below. Of course, these are merely examples and are not intended to limit the present invention. In addition, the present invention may repeat reference numerals and / or letters in different examples. Such repetition is for the purpose of simplicity and clarity and does not in itself indicate the relationship between the various embodiments and / or configurations discussed.

[0101] In order to solve the problem in the existing technology that during the AI ​​detection process of the pan-tilt camera, the object frame coordinates of the object calculated by the image algorithm are prone to jitter due to changes in image light and the current algorithm uses low-bit inference (limited by chip computing power, in order to ensure the response speed of the algorithm model, a low-bit inference algorithm is required). In addition, since the filter frame of the gimbal camera in the prior art is an area of ​​fixed width and height, the filter frame jitters in order to keep the object within the fixed filter frame, causing the gimbal to jitter along with the object frame coordinates, thereby causing the gimbal camera to fail to track the detected object in a timely manner or even lose the object, which seriously affects the user experience. The present application provides an object tracking method, device and electronic device, which can dynamically determine the filter frame according to the attribute information of the object and the movement information during the movement when the object is detected by an image acquisition device, and track the detected object based on the dynamically determined filter frame. It can ensure that when the detected object target jitters, the object is always in the dynamically determined filter frame, thereby preventing the gimbal from jittering along with the object frame coordinates, and realizing that the gimbal can move smoothly during the process of the gimbal camera tracking the moving object, so that the gimbal camera can track the detected object in a timely and accurate manner, thereby improving the user experience.

[0102] The object tracking method provided in this application is further explained below with reference to specific embodiments in conjunction with the accompanying drawings. The embodiments do not constitute a limitation of the embodiments of the present invention.

[0103] See also Figure 1 , is a flow chart of an embodiment of an object tracking method provided in an embodiment of the present application. Figure 1 As shown, the process may include the following steps:

[0104] Step 101: When an object is detected by an image acquisition device, attribute information of the object is identified.

[0105] The image acquisition device is a device capable of capturing images, and may be a video camera, a camera, or other device with image acquisition capabilities, and the present application is not limited thereto. Furthermore, in order to track detected objects, the image acquisition device may be a rotatable device, such as a pan-tilt camera, which can track detected objects by rotating the camera via the pan-tilt.

[0106] The above-mentioned objects refer to movable objects detected by the image acquisition device, which may include but are not limited to: people, vehicles, animals, etc.

[0107] The above-mentioned attribute information refers to the attributes of the detected object itself, which may include but is not limited to: object size, object type, object color, etc.

[0108] In one embodiment, the execution subject of the embodiment of the present application can be an image acquisition device or an electronic device connected to the image acquisition device. Based on this, the execution subject of the embodiment of the present application can perform object detection through the image acquisition device.

[0109] Optionally, when an object is detected by an image acquisition device, the execution subject of the embodiment of the present application may identify the detected object to identify attribute information of the object.

[0110] As a possible implementation method, the execution subject of the embodiment of the present application can collect an image of the detected object and input the image into a preset image recognition model to obtain attribute information of the object output by the image recognition model.

[0111] Step 102: Determine the initial filter frame of the object based on the above attribute information.

[0112] The initial filter frame refers to the filter frame initially determined by the executing entity of the present embodiment. The filter frame is an area defined in the center of the entire captured image. When a detected object leaves this area, the image acquisition device must be rotated to bring the object within the filter frame, thereby enabling detection and tracking of the detected object.

[0113] In one embodiment, in order to ensure more accurate tracking of the detected object, the execution subject of the embodiment of the present application may first determine an initial filtering frame for detecting the above object based on the attribute information of the detected object.

[0114] As a possible implementation, the attribute information may include the object type of the object. Based on this, the execution subject of the embodiment of the present application may determine, based on the object type of the identified object, an initial filter frame size corresponding to the object type from a pre-set correspondence between the object type and the filter frame size. The initial filter frame size may include an initial width value and an initial height value of the initial filter frame. The object type refers to the type of the detected object, such as whether the object is a person, a vehicle, or an animal.

[0115] Furthermore, to make the detected object clearer and more complete, the execution subject of the embodiment of the present application may determine the center point of the captured image as the center point of the initial filter frame, and determine the center point coordinates of the center point of the initial filter frame (for example, if a coordinate system is established with the center point of the captured image as the origin, then the center point coordinates are (0, 0)). Based on the center point coordinates and the initial height and initial width of the initial filter frame, the initial vertex coordinates of the initial filter frame are obtained. Afterwards, the filter frame corresponding to the initial vertex coordinates, initial width, and initial height values ​​may be determined as the initial filter frame of the object.

[0116] As an exemplary embodiment, the execution entity of the embodiment of the present application can represent the initial filtering frame by Box[X1, Y1, W1, H1], where (X1, Y1) represents the initial vertex coordinates of the initial filtering frame, W1 represents the initial width value of the initial filtering frame, and H1 represents the initial height value of the initial filtering frame.

[0117] As another possible implementation, the attribute information may include the object size. Based on this, the execution entity of the embodiment of the present application may multiply the object size of the identified object by a preset size coefficient, that is, multiply the height and width of the object by the size coefficient respectively, thereby obtaining the initial width and initial height values ​​of the initial filter frame. The object size refers to the actual size of the detected object.

[0118] Furthermore, to make the detected object clearer and more complete, the execution subject of the embodiment of the present application may determine the center point of the captured image as the center point of the initial filter frame, and determine the center point coordinates of the center point of the initial filter frame (for example, if a coordinate system is established with the center point of the captured image as the origin, then the center point coordinates are (0, 0)). Based on the center point coordinates and the initial height and initial width of the initial filter frame, the initial vertex coordinates of the initial filter frame are obtained. Afterwards, the filter frame corresponding to the initial vertex coordinates, initial width, and initial height values ​​may be determined as the initial filter frame of the object.

[0119] As an exemplary implementation, the execution entity of the embodiment of the present application can represent the initial filtering frame by Box[X1, Y1, W1, H1], where (X1, Y1) represents the vertex coordinates of the initial filtering frame, W1 represents the initial width value of the initial filtering frame, and H1 represents the initial height value of the initial filtering frame.

[0120] As another optional implementation, the attribute information of the object may include the object type and object size of the object. Based on this, the execution entity of the embodiment of the present application may determine the type coefficient corresponding to the object based on the identified object type. Thereafter, the object width value included in the object size may be multiplied by the type coefficient and the preset size coefficient to obtain the initial width value of the initial filter frame, and the object height value included in the object size may be multiplied by the type coefficient and the size coefficient to obtain the initial height value of the initial filter frame.

[0121] Among them, the execution subject of the embodiment of the present application can pre-store the correspondence between the type of the object and the type coefficient, and after determining the object type, search the above correspondence according to the object type to determine the type coefficient corresponding to the object.

[0122] Finally, the initial filter frame of the object may be determined according to the initial width value and the initial height value.

[0123] As an exemplary embodiment, in order to make the detected object clearer and more complete, the execution subject of the embodiment of the present application can determine the center point of the captured image of the image acquisition device as the center point of the initial filter frame, and determine the center point coordinates of the center point of the initial filter frame (for example, if a coordinate system is established with the center point of the captured image as the origin, then the center point coordinates are (0, 0)). Based on the center point coordinates and the initial width and initial height values ​​of the initial filter frame, the initial vertex coordinates of the initial filter frame are determined. Thereafter, the filter frame corresponding to the initial vertex coordinates, initial width, and initial height values ​​can be determined as the initial filter frame of the object.

[0124] As a possible implementation method, the above-mentioned initial vertex coordinates can be the upper left vertex coordinates of the initial filter frame. Then, the horizontal coordinate value of the initial vertex coordinates can be the horizontal coordinate value of the center point coordinates of the shooting screen minus half of the above-mentioned initial width value; the vertical coordinate value of the initial vertex coordinates can be the vertical coordinate value of the center point coordinates of the shooting screen plus half of the initial height value, thereby obtaining the initial vertex coordinates.

[0125] As another possible implementation method, the above-mentioned initial vertex coordinates can be the lower left vertex coordinates of the initial filter frame. Then, the horizontal coordinate value of the initial vertex coordinates can be the horizontal coordinate value of the center point coordinates of the shooting screen minus half of the above-mentioned initial width value; the vertical coordinate value of the initial vertex coordinates can be the vertical coordinate value of the center point coordinates of the shooting screen minus half of the initial height value, thereby obtaining the initial vertex coordinates.

[0126] As another possible implementation method, the above-mentioned initial vertex coordinates can be the upper right vertex coordinates of the initial filter frame. Then, the horizontal coordinate value of the initial vertex coordinates can be the horizontal coordinate value of the center point coordinates of the shooting screen plus half of the above-mentioned initial width value; the vertical coordinate value of the initial vertex coordinates can be the vertical coordinate value of the center point coordinates of the shooting screen plus half of the initial height value, thereby obtaining the initial vertex coordinates.

[0127] As another possible implementation method, the above-mentioned initial vertex coordinates can be the lower right vertex coordinates of the initial filter frame. Then, the horizontal coordinate value of the initial vertex coordinates can be the horizontal coordinate value of the center point coordinates of the shooting screen plus half of the above-mentioned initial width value; the vertical coordinate value of the initial vertex coordinates can be the vertical coordinate value of the center point coordinates of the shooting screen minus half of the initial height value, thereby obtaining the initial vertex coordinates.

[0128] It can be understood that the above-mentioned initial vertex coordinates can also be a combination of any multiple (two, three or four) vertex coordinates of the four vertices (upper left, lower left, upper right, lower right) in the target filter box, and the embodiment of the present application does not limit this.

[0129] Step 103: Determine the movement information of the object during its movement.

[0130] Step 104: Adjust the initial filter frame according to the movement information to obtain the target filter frame.

[0131] The following is a unified description of step 103 and step 104:

[0132] The above-mentioned movement information refers to information related to the movement of the detected object, which may include but is not limited to: speed and acceleration. Furthermore, the above-mentioned speed may include speed magnitude and speed direction, and the above-mentioned acceleration may include acceleration magnitude and acceleration direction.

[0133] In one embodiment, since the object detected by the image acquisition device is generally in a moving state, in order to prevent the filter frame from shaking due to the target shaking of the detected object, thereby causing the pan-tilt head to shake continuously, the execution subject of the embodiment of the present application can determine the movement information of the object during the movement process, and adjust the initial filter frame determined above according to the movement information to obtain the target filter frame.

[0134] As an exemplary embodiment, the movement information may include the speed of the object. Based on this, the execution subject of the embodiment of the present application may determine the speed of the object during movement and adjust the initial filter frame according to the speed.

[0135] As another exemplary embodiment, the movement information may include the acceleration of the object. Based on this, the execution subject of the embodiment of the present application may determine the acceleration of the object during movement and adjust the initial filter frame according to the acceleration.

[0136] As another exemplary embodiment, the movement information may include the speed and acceleration of the object. Based on this, the execution subject of the embodiment of the present application may determine the speed and acceleration of the object during movement and adjust the initial filter frame according to the speed and acceleration.

[0137] As for how to determine the speed and acceleration of the object during its movement, we can refer to the following Figure 2 The process shown is explained below and will not be described in detail here.

[0138] Furthermore, as to how to adjust the initial filter frame according to the movement information to obtain the target filter frame, it can be described below. Figure 3 The process shown is explained below and will not be described in detail here.

[0139] Step 105: Track the object based on the target filter frame.

[0140] As can be seen from the above description, when an object is detected by an image acquisition device, the execution subject of the present embodiment can determine a target filter frame for the object based on the object's attribute information and the object's movement information during movement. Based on this, the execution subject of the present embodiment can track the detected object based on the determined target filter frame.

[0141] In one embodiment, the image acquisition device may be a pan-tilt camera. Based on this, the execution subject of the embodiment of the present application may obtain the object frame coordinates of the object during the process of tracking the detected object. The object frame coordinates may include the vertex coordinates, height value, and width value of the object frame.

[0142] It should be noted that the above-mentioned object frame refers to the detection frame used by the image acquisition device to identify and detect the detected object. When the image acquisition device detects the object, it can lock the object by including it in the object frame, and the image acquisition device will not move due to the movement of the object frame. Generally, the image acquisition device will only rotate when the object frame moves out of the filter frame.

[0143] Based on this, the execution subject of the embodiment of the present application can determine whether the object is completely located within the target filtering frame according to the above-mentioned object frame coordinates.

[0144] Optionally, when it is determined that the object is not completely within the target filtering frame, the pan-tilt camera can be controlled to move in the direction of the object so that the object is completely within the target filtering frame.

[0145] For example, Figure 4 As shown in FIG, a schematic diagram of object tracking provided by an embodiment of the present application is provided. Figure 4 As shown in the figure, the dotted line part is the target filter frame, and the solid line part is the shooting picture. n-1 At this moment, the object is within the target filter frame. At this moment, the execution subject of the embodiment of the present application may not control the pan-tilt rotation of the pan-tilt camera. n At this moment, the object has walked out of the target filter frame. At this time, the executive body of the embodiment of the present application can control the pan-tilt camera to move in the direction of the object (that is, the upper left of the shooting picture), thereby ensuring that the object is always within the target filter frame.

[0146] The technical solution provided by the embodiment of the present application is to identify the attribute information of the object when the object is detected by an image acquisition device, determine the initial filter frame of the object based on the attribute information, determine the movement information of the object during the movement, adjust the initial filter frame based on the movement information, obtain the target filter frame, and track the above-mentioned object based on the target filter frame. This technical solution, when the object is detected by an image acquisition device, dynamically determines the filter frame based on the attribute information of the object and the movement information during the movement, and tracks the detected object based on the dynamically determined filter frame. It can ensure that when the detected object target shakes, the object is always within the dynamically determined filter frame, thereby preventing the pan-tilt camera from shaking along with the object frame coordinates, and realizing that the pan-tilt camera can move smoothly during the process of tracking the moving object, so that the pan-tilt camera can track the detected object in a timely and accurate manner, thereby improving the user experience.

[0147] See also Figure 2 , which is a flow chart of an embodiment of another object tracking method provided in an embodiment of the present application. Figure 2 The process shown in Figure 1 Based on the process shown in FIG, the following describes how to determine the movement information of the object during its movement. Figure 2 As shown, the process may include the following steps:

[0148] Step 201: Obtain historical object frame coordinates and historical time points of an object detected within a preset time period. The historical object frame coordinates include vertex coordinates, height value, and width value of the historical object frame.

[0149] Step 202: Determine the historical speed and historical acceleration of the object corresponding to each historical time point based on the historical object frame coordinates and the historical time points, and obtain multiple historical speeds and multiple historical accelerations.

[0150] The following is a unified description of step 201 and step 202:

[0151] The above-mentioned preset time period refers to a preset historical time period before the current moment.

[0152] The historical object frame coordinates described above refer to the coordinates of the detection frame used by the image acquisition device during the historical time period when identifying and detecting the detected object. Upon detecting an object, the image acquisition device can lock onto the object by enclosing it within the object frame. The image acquisition device does not move due to the movement of the object frame; generally, the image acquisition device will only rotate after the object frame moves outside the filter frame.

[0153] The above historical time point refers to the time point when each historical object frame is detected.

[0154] The above historical speed refers to the speed of the object when each historical object frame is detected.

[0155] The above acceleration refers to the acceleration of the object when each historical object frame is detected.

[0156] In one embodiment, the execution subject of the embodiment of the present application can obtain the historical object frame coordinates and historical time points of the objects detected within a preset time period, wherein each of the above historical object frame coordinates may include the vertex coordinates, height value, and width value of the historical object frame.

[0157] For example, see Figure 5 , is a schematic diagram of object detection provided by an embodiment of the present application. Figure 5 As shown in the figure, when the image acquisition device detects an object, it can record and save N historical object frames detected by the AI ​​within a time period T. The coordinates of each of these historical object frames can be [X1, Y1, W1, H1], [X2, Y2, W2, H2], [X3, Y3, W3, H3], ..., [Xn, Yn, Wn, Hn]. Furthermore, a set of these historical object frame coordinates can be obtained.

[0158] Based on this, the execution subject of the embodiment of the present application can further determine the historical speed and historical acceleration of the above-mentioned object corresponding to each historical time point according to the above-mentioned historical object frame coordinates and historical time points, thereby obtaining multiple historical speeds and multiple historical accelerations.

[0159] Continue with Figure 5As shown in the example, after obtaining the coordinates and historical time points of each historical object frame, the execution subject of the embodiment of the present application can determine the distance and historical time difference between each two historical object frames. Based on this, the historical speed and historical acceleration corresponding to each historical object frame can be further determined, thereby obtaining a historical speed set and a historical acceleration set. Each historical speed set can include multiple historical speeds, and each historical acceleration set can include multiple historical accelerations.

[0160] Step 203: Determine an average speed of the multiple historical speeds and an average acceleration of the multiple historical accelerations.

[0161] Step 204: Determine the average speed as the speed of the object during the movement, and determine the average acceleration as the acceleration of the object during the movement.

[0162] Step 205: Determine the speed and the acceleration as movement information of the object during its movement.

[0163] The following is a unified description of steps 203 to 205:

[0164] The above-mentioned speed and acceleration may be the speed and acceleration corresponding to the object predicted by the embodiment of the present application at the current moment.

[0165] In an embodiment of the present application, in order to more accurately determine the speed and acceleration of the detected object at the current moment, the execution entity of the embodiment of the present application may determine the average speed of multiple historical speeds and the average acceleration of multiple historical accelerations.

[0166] Afterwards, the average speed can be determined as the speed of the object during the movement, and the average acceleration can be determined as the acceleration of the object during the movement. Furthermore, the speed and acceleration can be determined as the movement information of the object during the movement.

[0167] The technical solution provided by an embodiment of the present application obtains historical object frame coordinates and historical time points of an object detected within a preset time period, wherein the historical object frame coordinates include the vertex coordinates, height value, and width value of the historical object frame. Based on the historical object frame coordinates and historical time points, the historical speed and historical acceleration of the object corresponding to each historical time point are determined, thereby obtaining multiple historical speeds and multiple historical accelerations. The average speed of the multiple historical speeds and the average acceleration of the multiple historical accelerations are determined, and the average speed and the average acceleration of the multiple historical accelerations are determined. The average speed is determined as the speed of the object during movement, and the average acceleration is determined as the acceleration of the object during movement. The speed and the acceleration are determined as the movement information of the object during movement. This technical solution can quickly and accurately determine the movement information of the object during movement by determining multiple historical speeds and multiple historical accelerations of the object during movement based on the historical object frame coordinates of the object, and determining the average speed of the multiple historical speeds as the predicted speed of the object during movement, and determining the average acceleration of the multiple historical accelerations as the predicted acceleration of the object during movement.

[0168] See also Figure 3 , which is a flow chart of an embodiment of another object tracking method provided in an embodiment of the present application. Figure 3 The process shown in Figure 2 Based on the process shown in FIG, it describes how to adjust the initial filter frame according to the movement information to obtain the target filter frame. Figure 3 As shown, the process may include the following steps:

[0169] Step 301: Determine the speed variance of the object based on multiple historical speeds and the above speed.

[0170] Step 302: Determine the acceleration variance of the object based on the multiple historical accelerations and the above acceleration.

[0171] The following is a unified description of step 301 and step 302:

[0172] In an embodiment of the present application, in order to determine whether the detected object has frequent shaking or necessary acceleration and deceleration, the execution entity of the embodiment of the present application can determine the velocity variance and acceleration variance of the object, thereby determining whether the object has frequent shaking or necessary acceleration and deceleration based on the object's velocity, velocity variance, acceleration, and acceleration variance.

[0173] Alternatively, if the acceleration and acceleration variance are large, it means the object is moving frequently in a short period of time, with a lot of jumps in distance or direction changes. If the speed and speed variance are large, it means the target is moving a lot.

[0174] In one embodiment, the execution subject of the embodiment of the present application may determine the velocity variance of the object by the following formula (1):

[0175]

[0176] Where D(v) represents the speed variance, n represents the number of historical speeds, and v i Represents the historical speed, and v represents the above speed, that is, the average speed of multiple historical speeds.

[0177] In one embodiment, the execution subject of the embodiment of the present application may determine the acceleration variance of the object by the following formula (2):

[0178]

[0179] Where D(a) represents the acceleration variance, n represents the number of historical accelerations, and a i represents the historical acceleration, and a represents the above acceleration, that is, the average acceleration of multiple historical accelerations.

[0180] Step 303: Determine the velocity weight of the object according to the velocity, velocity variance, acceleration, and acceleration variance; wherein the velocity, velocity variance, acceleration, and acceleration variance are negatively correlated with the velocity weight.

[0181] In practical applications, when an object moves frequently within a short period of time, with abrupt movements, frequent changes in direction, and / or a large amplitude, that is, when the object's velocity, velocity variance, acceleration, and acceleration variance are large, the object's movement is noticeable. Therefore, when jitter is detected in the object frame, the jitter is not noticeable. Therefore, the object's filter frame can be reduced to ensure that the image acquisition device can rotate with the object's movement. Based on this, the velocity, velocity variance, acceleration, and acceleration variance are negatively correlated with the velocity weight.

[0182] In one embodiment, to more accurately determine the speed weight of an object, the execution entity of the embodiment of the present application may match the determined speed and speed variance with a preset speed interval to obtain a target speed interval to which the speed and speed variance belong. Each preset speed interval may correspond to at least one range of speeds and at least one range of speed variances, and each speed interval may correspond to a speed weight.

[0183] Afterwards, the speed weight corresponding to the target speed interval (hereinafter referred to as the first speed weight for ease of distinction) may be determined from the correspondence between the preset speed intervals and the speed weights.

[0184] Accordingly, the acceleration and acceleration variance can be matched with the preset acceleration intervals to obtain the target acceleration interval to which the acceleration and acceleration variance belong. Each preset acceleration interval can correspond to at least one range of acceleration and at least one range of acceleration variance, and each acceleration interval can correspond to a velocity weight.

[0185] Afterwards, the speed weight corresponding to the target acceleration interval (hereinafter referred to as the second speed weight for ease of description) may be determined from the correspondence between the preset acceleration intervals and the speed weights.

[0186] Finally, a weighted sum calculation may be performed on the first speed weight and the second speed weight to obtain the speed weight of the object.

[0187] In addition, an average value of the first speed weight and the second speed weight may be calculated, and the average value may be determined as the speed weight of the object.

[0188] Step 304: Adjust the initial filter frame according to the speed weight to obtain a target filter frame.

[0189] The above-mentioned target filter frame refers to the final target filter frame determined after the adjustment is completed.

[0190] In one embodiment, after the execution subject of the embodiment of the present application determines the speed weight of the object, the initial filter frame may be adjusted according to the determined speed weight to obtain the final target filter frame.

[0191] As an exemplary implementation, the initial width value and the initial height value of the initial filter frame object may be multiplied by the speed weight to obtain the target width value and the target height value.

[0192] Afterwards, the center point of the initial filter frame (i.e. Figure 1 The center point of the captured image determined by the process shown in the figure is used as the center point of the target filter frame, and the target center coordinates of the center point of the target filter frame in the captured image of the image acquisition device are determined (for example, if the center point of the initial filter frame is the center point of the captured image, and the center coordinates of the captured image are (0, 0), then the target center coordinates of the center point of the target filter frame are (0, 0)). Subsequently, the target vertex coordinates of the target filter frame can be determined based on the target center coordinates, as well as the target width and target height values ​​of the target filter frame.

[0193] As a possible implementation method, the above-mentioned target vertex coordinates can be the upper left vertex coordinates of the target filter box. Then, the horizontal coordinate value of the target vertex coordinates can be the horizontal coordinate value of the target center point coordinates minus half of the above-mentioned target width value; the vertical coordinate value of the target vertex coordinates can be the vertical coordinate value of the target center point coordinates plus half of the target height value, thereby obtaining the target vertex coordinates.

[0194] As another possible implementation method, the above-mentioned target vertex coordinates can be the upper right vertex coordinates of the target filter box. Then, the horizontal coordinate value of the target vertex coordinates can be the horizontal coordinate value of the target center point coordinates plus half of the above-mentioned target width value; the vertical coordinate value of the target vertex coordinates can be the vertical coordinate value of the target center point coordinates plus half of the target height value, thereby obtaining the target vertex coordinates.

[0195] As another possible implementation method, the above-mentioned target vertex coordinates can be the lower left vertex coordinates of the target filter box. Then, the horizontal coordinate value of the target vertex coordinates can be the horizontal coordinate value of the target center point coordinates minus half of the above-mentioned target width value; the vertical coordinate value of the target vertex coordinates can be the vertical coordinate value of the target center point coordinates minus half of the target height value, thereby obtaining the target vertex coordinates.

[0196] As another possible implementation method, the above-mentioned target vertex coordinates can be the lower right vertex coordinates of the target filter box. Then, the horizontal coordinate value of the target vertex coordinates can be the horizontal coordinate value of the target center point coordinates plus half of the above-mentioned target width value; the vertical coordinate value of the target vertex coordinates can be the vertical coordinate value of the target center point coordinates minus half of the target height value, thereby obtaining the target vertex coordinates.

[0197] It can be understood that the above-mentioned target vertex coordinates can also be a combination of any multiple (two, three or four) vertex coordinates of the four vertices (upper left, lower left, upper right, lower right) in the target filter box, and the embodiment of the present application does not limit this.

[0198] Finally, the filter frame corresponding to the target vertex coordinates, target width value, and target height value may be determined as the target filter frame.

[0199] The technical solution provided by the embodiment of the present application determines the speed variance of the object based on multiple historical speeds and the above speed, determines the acceleration variance of the object based on multiple historical accelerations and the above acceleration, and determines the speed weight of the object based on the speed, speed variance, acceleration, and acceleration variance; wherein the above speed, speed variance, acceleration, and acceleration variance are negatively correlated with the speed weight, and the initial filter frame is adjusted according to the above speed weight to obtain the target filter frame. This technical solution, by determining the speed weight of the object based on the speed, speed variance, acceleration, and acceleration variance of the object, and adjusting the initial filter frame according to the speed weight, adjusts the initial filter frame according to the movement information of the object, so that the determined target filter frame is more adaptable to the movement of the object, thereby achieving smooth movement of the gimbal camera during tracking of the moving object, so that the detected object can be tracked in a timely and accurate manner, thereby improving the user experience.

[0200] See also Figure 6 , is a block diagram of an embodiment of an object tracking device provided in an embodiment of the present application. Figure 6 As shown, the device may include:

[0201] The recognition module 61 is used to recognize attribute information of an object when the object is detected by the image acquisition device;

[0202] A first determining module 62 is configured to determine an initial filtering frame of the object based on the attribute information;

[0203] A second determining module 63 is used to determine movement information of the object during movement;

[0204] An adjustment module 64 is configured to adjust the initial filter frame according to the movement information to obtain a target filter frame;

[0205] The tracking module 65 is configured to track the object based on the target filter frame.

[0206] As a possible implementation, the attribute information includes the object type and object size of the object, and the first determining module 62 includes:

[0207] A first determining submodule, configured to determine a type coefficient corresponding to the object according to the object type;

[0208] A second determining submodule is configured to multiply the object width value included in the object size by the type coefficient and a preset size coefficient to obtain an initial width value of the initial filtering frame;

[0209] A third determining submodule is configured to multiply the object height value included in the object size by the type coefficient and the size coefficient to obtain an initial height value of the initial filtering frame;

[0210] The fourth determining submodule is configured to determine an initial filtering frame of the object according to the initial width value and the initial height value.

[0211] As a possible implementation, the third determining submodule is specifically configured to:

[0212] Taking the center point of the shooting picture of the image acquisition device as the center point of the initial filter frame, and determining the center point coordinates of the center point of the initial filter frame in the shooting picture;

[0213] Determining initial vertex coordinates of the filter frame according to the center point coordinates, and the initial width value and initial height value of the initial filter frame;

[0214] A filter frame corresponding to the initial vertex coordinates, the initial width value, and the initial height value is determined as the initial filter frame.

[0215] As a possible implementation, the second determining module 63 is specifically configured to:

[0216] Acquire historical object frame coordinates and historical time points of the object detected within a preset historical time period, wherein the historical object frame coordinates include vertex coordinates, height value, and width value of the historical object frame;

[0217] Determining, based on the historical object frame coordinates and the historical time points, a historical speed and a historical acceleration of the object corresponding to each historical time point, to obtain a plurality of historical speeds and a plurality of historical accelerations;

[0218] determining an average speed of the plurality of historical speeds, and an average acceleration of the plurality of historical accelerations;

[0219] determining the average speed as the speed of the object during the movement, and determining the average acceleration as the acceleration of the object during the movement;

[0220] The velocity and the acceleration are determined as movement information of the object during its movement.

[0221] As a possible implementation, the adjustment module 64 includes:

[0222] a fifth determining submodule, configured to determine a velocity variance of the object based on a plurality of historical velocities and the velocity;

[0223] a sixth determining submodule, configured to determine an acceleration variance of the object based on a plurality of historical accelerations and the acceleration;

[0224] a seventh determination submodule, configured to determine a velocity weight of the object based on the velocity, the velocity variance, the acceleration, and the acceleration variance; wherein the velocity, the velocity variance, the acceleration, and the acceleration variance are negatively correlated with the velocity weight;

[0225] The adjustment submodule is configured to adjust the initial filter frame according to the speed weight to obtain a target filter frame.

[0226] As a possible implementation, the seventh determining submodule is specifically configured to:

[0227] Matching the speed and the speed variance with a preset speed interval to obtain a target speed interval to which the speed and the speed variance belong;

[0228] Determining a first speed weight corresponding to the target speed interval from a correspondence between preset speed intervals and speed weights;

[0229] Matching the acceleration and the acceleration variance with a preset acceleration interval to obtain a target acceleration interval to which the acceleration and the acceleration variance belong;

[0230] Determining a second speed weight corresponding to the target acceleration interval from a preset correspondence between acceleration intervals and speed weights;

[0231] A weighted sum calculation is performed on the first speed weight and the second speed weight to obtain a speed weight of the object.

[0232] As a possible implementation, the adjustment submodule is specifically configured to:

[0233] Multiplying the initial width value and the initial height value corresponding to the initial filter frame by the speed weight respectively to obtain a target width value and a target height value;

[0234] Taking the center point of the initial filter frame as the center point of the target filter frame, and determining the target center coordinates of the center point of the target filter frame in the shooting picture of the image acquisition device;

[0235] Determining target vertex coordinates of the target filter frame according to the target center coordinates, and the target width value and the target height value of the target filter frame;

[0236] A filter frame corresponding to the target vertex coordinates, the target width value, and the target height value is determined as a target filter frame.

[0237] As a possible implementation, the tracking module 65 is specifically configured to:

[0238] In the process of tracking the object, obtaining the object frame coordinates of the object, the object frame coordinates including the vertex coordinates, height value, and width value of the object frame;

[0239] Determining whether the object is completely within the target filter frame according to the object frame coordinates;

[0240] When it is determined that the object is not completely within the target filtering frame, the pan-tilt camera is controlled to move in the direction of the object so that the object is completely within the target filtering frame.

[0241] like Figure 7As shown, it is a structural diagram of an electronic device provided in an embodiment of the present application, including a processor 71, a communication interface 72, a memory 73 and a communication bus 74, wherein the processor 71, the communication interface 72, and the memory 73 communicate with each other through the communication bus 74.

[0242] Memory 73, for storing computer programs;

[0243] In one embodiment of the present application, the processor 71 is configured to execute a program stored in the memory 73 to implement the object tracking method provided by any of the aforementioned method embodiments, including:

[0244] When an object is detected by an image acquisition device, identifying attribute information of the object;

[0245] determining an initial filtering frame of the object according to the attribute information;

[0246] Determining movement information of the object during movement;

[0247] Adjusting the initial filter frame according to the movement information to obtain a target filter frame;

[0248] The object is tracked based on the target filter frame.

[0249] An embodiment of the present application further provides a storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the object tracking method provided in any of the aforementioned method embodiments.

[0250] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of this embodiment.

[0251] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a general hardware platform, or of course, by hardware. Based on this understanding, the above technical solution, in essence, or the part that contributes to the relevant technology, can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or certain parts of the embodiment.

[0252] It should be understood that the terms used herein are for the purpose of describing specific example embodiments only and are not intended to be limiting. Unless the context clearly indicates otherwise, the singular forms "one", "an" and "said" as used herein may also be meant to include plural forms. The terms "comprise", "include", "contain" and "have" are inclusive and therefore specify the presence of stated features, steps, operations, elements and / or parts, but do not exclude the presence or addition of one or more other features, steps, operations, elements, parts, and / or combinations thereof. The method steps, processes, and operations described herein are not to be construed as necessarily requiring them to be performed in the specific order described or illustrated, unless the order of execution is clearly indicated. It should also be understood that additional or alternative steps may be used.

[0253] The foregoing description is intended only to provide specific embodiments of the present invention, which will enable those skilled in the art to understand and implement the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown herein, but is intended to be accorded the widest scope consistent with the principles and novel features claimed herein.

Claims

1. An object tracking method, characterized in that: The method comprises: When an object is detected by an image acquisition device, identifying attribute information of the object; determining an initial filtering frame of the object according to the attribute information; Determining movement information of the object during movement; Adjusting the initial filter frame according to the movement information to obtain a target filter frame; The object is tracked based on the target filter frame.

2. The method according to claim 1, characterized in that The attribute information includes an object type and an object size of the object, and determining an initial filter frame of the object according to the attribute information includes: Determining a type coefficient corresponding to the object according to the object type; Multiplying the object width value included in the object size by the type coefficient and the preset size coefficient to obtain an initial width value of the initial filter frame; Multiplying the object height value included in the object size by the type coefficient and the size coefficient to obtain an initial height value of the initial filter frame; An initial filter frame of the object is determined according to the initial width value and the initial height value.

3. The method according to claim 2, characterized in that The determining, according to the initial width value and the initial height value, an initial filter frame of the object includes: Taking the center point of the shooting picture of the image acquisition device as the center point of the initial filter frame, and determining the center point coordinates of the center point of the initial filter frame in the shooting picture; Determining initial vertex coordinates of the filter frame according to the center point coordinates, and the initial width value and initial height value of the initial filter frame; A filter frame corresponding to the initial vertex coordinates, the initial width value, and the initial height value is determined as the initial filter frame.

4. The method according to claim 1, wherein The determining movement information of the object during movement includes: Acquire historical object frame coordinates and historical time points of the object detected within a preset historical time period, wherein the historical object frame coordinates include vertex coordinates, height value, and width value of the historical object frame; Determining, based on the historical object frame coordinates and the historical time points, a historical speed and a historical acceleration of the object corresponding to each historical time point, to obtain a plurality of historical speeds and a plurality of historical accelerations; determining an average speed of the plurality of historical speeds, and an average acceleration of the plurality of historical accelerations; determining the average speed as the speed of the object during the movement, and determining the average acceleration as the acceleration of the object during the movement; The velocity and the acceleration are determined as movement information of the object during its movement.

5. The method according to claim 4, characterized in that The adjusting the initial filter frame according to the movement information to obtain a target filter frame includes: determining a velocity variance of the object based on a plurality of historical velocities and the velocity; determining an acceleration variance of the object based on a plurality of historical accelerations and the acceleration; Determining a velocity weight of the object according to the velocity, the velocity variance, the acceleration, and the acceleration variance; wherein the velocity, the velocity variance, the acceleration, and the acceleration variance are negatively correlated with the velocity weight; The initial filter frame is adjusted according to the speed weight to obtain a target filter frame.

6. The method according to claim 5, characterized in that Determining the velocity weight of the object according to the velocity, the velocity variance, the acceleration, and the acceleration variance includes: Matching the speed and the speed variance with a preset speed interval to obtain a target speed interval to which the speed and the speed variance belong; Determining a first speed weight corresponding to the target speed interval from a correspondence between preset speed intervals and speed weights; Matching the acceleration and the acceleration variance with a preset acceleration interval to obtain a target acceleration interval to which the acceleration and the acceleration variance belong; Determining a second speed weight corresponding to the target acceleration interval from a preset correspondence between acceleration intervals and speed weights; A weighted sum calculation is performed on the first speed weight and the second speed weight to obtain a speed weight of the object.

7. The method according to claim 5, characterized in that The adjusting the initial filter frame according to the speed weight to obtain a target filter frame includes: Multiplying the initial width value and the initial height value corresponding to the initial filter frame by the speed weight respectively to obtain a target width value and a target height value; Taking the center point of the initial filter frame as the center point of the target filter frame, and determining the target center coordinates of the center point of the target filter frame in the shooting picture of the image acquisition device; Determining target vertex coordinates of the target filter frame according to the target center coordinates, and the target width value and the target height value of the target filter frame; A filter frame corresponding to the target vertex coordinates, the target width value, and the target height value is determined as a target filter frame.

8. The method according to claim 1, characterized in that The image acquisition device is a pan-tilt camera, and tracking the object based on the target filter frame includes: In the process of tracking the object, obtaining the object frame coordinates of the object, the object frame coordinates including the vertex coordinates, height value, and width value of the object frame; Determining whether the object is completely within the target filter frame according to the object frame coordinates; When it is determined that the object is not completely within the target filtering frame, the pan-tilt camera is controlled to move in the direction of the object so that the object is completely within the target filtering frame.

9. An object tracking device, characterized in that: The device comprises: An identification module, configured to identify attribute information of an object when the object is detected by the image acquisition device; A first determining module, configured to determine an initial filtering frame of the object according to the attribute information; A second determining module is used to determine movement information of the object during movement; An adjustment module, configured to adjust the initial filter frame according to the movement information to obtain a target filter frame; A tracking module is used to track the object based on the target filtering frame.

10. An electronic device, characterized in that: include: A processor and a memory, wherein the processor is configured to execute an object tracking program stored in the memory to implement the object tracking method according to any one of claims 1 to 8.

Images