WO2022068466A1

WO2022068466A1 - Automatic focusing method and device, and camera

Info

Publication number: WO2022068466A1
Application number: PCT/CN2021/113969
Authority: WO
Inventors: 陶旭磊
Original assignee: 华为技术有限公司
Priority date: 2020-09-30
Filing date: 2021-08-23
Publication date: 2022-04-07

Abstract

Disclosed is an automatic focusing technology, which belongs to the field of video monitoring. When there is a target object in a photographed area, a camera acquires at least two images including the target object, and then calculates an FV of a target area where the target object is located in each image, so as to obtain a target position for clearly imaging the target object, thereby achieving automatic focusing. Said method avoids the influence of multiple object distances, and increases the focusing efficiency.

Description

Autofocus method, device and camera

technical field

The present application relates to the field of video surveillance, and in particular, to an automatic focusing method, device and camera.

Background technique

In recent years, with the continuous development of video surveillance, more and more challenging requirements have been placed on the camera, especially for the focusing ability of the camera, which is required to achieve all-weather precise focusing in dynamic scenes and scenes with multiple object distances. In this way, a clear image of the target in the scene monitored by the camera can be ensured. At present, many cameras are equipped with a special focusing module, which can control the moving direction and moving step of the focusing lens by controlling the focusing motor, so as to make the focus fall on the photosensitive plane and realize focusing.

In the related art, the focusing module of the camera usually adopts a "climbing algorithm" when focusing. In a fixed scene, the focus value (FV) of the image changes with the movement of the focusing motor, showing a hill-like appearance. Gaussian-like shape, the peak of the hill corresponds to the position of the focus motor corresponding to the sharp point of the image. The specific focusing process is as follows: the focusing module controls the focusing motor to move to the initial direction within a certain range, and compares the global FV of each frame of images before and after the movement. After the focusing motor moves, if the FV becomes larger, it continues to move to the initial direction. Focus motor, if FV becomes smaller, move the focus motor in the opposite direction. During the movement of the focus motor, the focus module continuously monitors the change of FV, and when the maximum value of FV is detected, it controls the focus motor to move to the position corresponding to the maximum value to complete the focus.

However, using the above method, the time of each focusing is at least 1 second or more. During this time period, the target in the dynamic scene is very likely to have a large displacement, so that the clear point of the image changes. The focused position of the focus motor is on a non-sharp point, resulting in a blurred image of the target. In addition, in a multi-object distance scenario, the focusing module in the above method is likely to focus on non-targets such as trees and the ground, which will also cause the image of the target to be blurred. It can be seen that the above-mentioned focusing method results in low imaging resolution and low focusing efficiency.

SUMMARY OF THE INVENTION

The present application provides an automatic focusing method, device and camera, which can ensure clear imaging of a target object in dynamic scenes and scenes with multiple object distances. The technical solution is as follows:

In a first aspect, an automatic focusing method is provided, the method includes: acquiring at least two frames of images including a target object, the at least two frames of images are obtained by capturing images of a captured area by the camera; for the at least two frames of images For each image, determine the target area where the target object is located in each image, and obtain the FV corresponding to each target area; based on the obtained at least two FVs, obtain the target position of the focusing motor in the camera , control the focusing motor in the camera to move to the target position, so that the focus of the light passing through the focusing lens falls on the photosensitive plane of the camera, and the FV corresponding to the target position meets the target focusing condition.

In the automatic focusing method, when there is a target object in the shooting area of the camera, the automatic focusing is triggered. At this time, the camera acquires at least two frames of images including the target object, and respectively determines the target where the target object is located in each frame of image. Then only the FV of the target area in each frame of image is calculated to obtain the target position to make the target object clearly imaged. Based on this, the focus motor is controlled to move to achieve focusing. That is to say, the automatic focusing method is based on the clarity of the target area where the target object is located, directly obtains the target position of the focusing motor, and then controls the focusing motor to move to the target position, which can avoid the influence of multiple object distances. The camera always takes the clear imaging of the target object as the goal to automatically focus, so as to realize the clear imaging of the target object, and this method can greatly shorten the focusing time and improve the focusing efficiency.

In the related art, when the camera is focusing, it needs to acquire multiple frames of images, and calculate the FV of the entire image of each frame of image. Control the focus motor to move and return to the motor position corresponding to the maximum FV. That is to say, in the related art, the camera is based on the sharpness of the entire shooting area, and does not consider whether the target object in the shooting area is clearly imaged, which can easily cause the image of the target object to be blurred. Moreover, this method needs to control the focus. The motor continuously moves to multiple motor positions, and then a motor position is determined from the multiple motor positions, and then the focusing motor is controlled to return to the determined motor position, which takes a long time and the focusing efficiency is not high.

Optionally, the acquisition includes at least two frames of images of the target object, including: acquiring the current position of the focus motor in the camera, and based on the current position, acquiring at least one reference position between the at least one reference position and the current position. The distance is the adjustment step; control the focus motor in the camera to move to the initial position and the at least one reference position respectively, when the focus motor in the camera is at the current position and the at least one reference position, the camera The at least two frames of images are acquired.

Optionally, the acquiring the current position of the focus motor in the camera includes one of the following:

In response to the camera detecting the target object for the first time, obtaining the current position through prediction;

In response to the camera not detecting the target object for the first time, the current position of the focus motor in the camera is acquired, and the current position is taken as the current position.

Based on the above optional implementation manner, no matter whether the camera detects the target object for the first time or not for the first time, a current position can be obtained, which is convenient for subsequent focusing cycles through the current position. That is to say, in this auto-focusing method, as long as the camera can detect the target object, it will continue to focus on the cycle, and a current position will be obtained during each focusing process. If the target object is detected for the first time, it will be obtained by prediction. a current position, and then perform the first focusing on the basis of the current position. If the target object is detected for the second time, the current position of the focusing motor is obtained, that is, when the focusing motor completes the first focusing position, and then perform a second focus based on this position, and so on, and so on, until the camera no longer detects the target object. Optionally, a detection period may also be set, and if the target object is detected within the detection period, the cycle will continue, and if the detection period ends, no further detection will be performed.

Optionally, the obtaining the current position through prediction includes: obtaining a predicted object distance between the camera and the target object; and obtaining the current position based on the predicted object distance and the focal length of the camera.

Based on the above optional implementation manner, the camera can give an estimated object distance by predicting the object distance between the camera and the target object, which is convenient to give an initial position of the focus motor, and the focus motor is at this position When the target object can achieve a certain degree of clear imaging.

Optionally, the adjustment step is half the focal depth.

Optionally, the obtaining at least one reference position based on the current position includes: obtaining one reference position on both sides of the current position based on the current position.

Based on the above optional implementation manner, the focusing module obtains two reference positions on the basis of the current position, which greatly reduces the focusing time. For example, it is only 0.16 seconds. During this time period, it can be approximately considered that the target object is stationary. Based on this, the accuracy of the target position can be ensured to the greatest extent, and the focusing efficiency can be improved.

Optionally, acquiring the target position of the focus motor in the camera based on the obtained at least two FVs includes: acquiring an association relationship between each FV and the position of the focus motor in the camera; based on each FV, The current position and the at least one reference position obtain the target position of the focus motor in the camera through the association relationship.

Optionally, the determining the target area where the target object is located in each image includes: inputting each image into a detection model, and the detection model is configured to output the each image according to the input each image. A detection result of an image, the detection result includes target area information of the target object in each image; the target area where the target object is located in each image is determined based on the detection result.

Based on the above optional implementation manner, after acquiring at least two frames of images, the target area where the target object is located in the image is determined in real time by the detection model, which improves the accuracy of determining the target area and further ensures the clarity of the target object. imaging.

In a second aspect, an automatic focusing device is provided, the device comprising:

an image acquisition module, configured to acquire at least two frames of images including the target object, the at least two frames of images are obtained by image capturing of the captured area by the camera;

The FV acquisition module is used for, for each of the at least two frames of images, to determine the target area where the target object is located in each of the images, and to acquire the focus value FV corresponding to each of the target areas;

The control module is configured to obtain the target position of the focusing motor in the camera based on the obtained at least two FVs, and control the focusing motor in the camera to move to the target position, so that the focus of the light passing through the focusing lens falls on the camera's On the photosensitive plane, the FV corresponding to the target position conforms to the target focusing condition.

Optionally, the image acquisition module includes:

a first obtaining unit, configured to obtain the current position of the focusing motor in the camera, and based on the current position, obtain at least one reference position, and the distance between the at least one reference position and the current position is the adjustment step;

a second acquisition unit, configured to control the focus motor in the camera to move to the initial position and the at least one reference position respectively, when the focus motor in the camera is located at the current position and the at least one reference position, the camera The at least two frames of images are acquired.

Optionally, the first acquiring unit is configured to acquire the current position by prediction in response to the camera detecting the target object for the first time; in response to the camera detecting the target object for the first time, acquire the current position of the focus motor in the camera. The current position is taken as the current position.

Optionally, the first obtaining unit is configured to obtain the predicted object distance between the camera and the target object; and obtain the current position based on the predicted object distance and the focal length of the camera.

Optionally, the adjustment step is half the focal depth.

Optionally, the first obtaining unit is configured to obtain the reference position on both sides of the current position based on the current position.

Optionally, the control module is used to obtain the association relationship between each FV and the position of the focus motor in the camera; based on each FV, the current position and the at least one reference position, obtain the association relationship through the association relationship. The target position of the focus motor in the camera.

Optionally, the FV acquisition module is used to input each image into a detection model, and the detection model is used to output the detection result of each image according to the inputted image, and the detection result includes the target. Target area information of the object in each image; determining the target area where the target object is located in each image based on the detection result.

In a third aspect, an auto-focus camera is provided, the auto-focus camera includes an image sensor and a processor, and the image sensor is configured to acquire at least two frames in the first aspect or any optional implementation manner of the first aspect image, the processor is configured to execute the autofocus method provided in the first aspect or any optional manner of the first aspect.

In a fourth aspect, an automatic focusing device is provided, the automatic focusing device includes a processor and a memory for storing at least one piece of program code, the at least one piece of program code is loaded and executed by the processor to make the automatic focusing The device executes the autofocus method provided in the first aspect or any optional manner of the first aspect.

In a fifth aspect, a computer-readable storage medium is provided, the computer-readable storage medium is used to store at least one piece of program code, and the at least one piece of program code is loaded and executed by a processor, so that the computer executes the first aspect or the first aspect. The automatic focusing method provided by any one of the optional manners in one aspect.

In a sixth aspect, a computer program product or computer program is provided, the computer program product or computer program comprising program code which, when run on an autofocus device, causes the autofocus device to perform the above-mentioned first aspect or the first aspect The autofocus method provided in various optional implementations of .

Description of drawings

1 is a schematic diagram of the relationship between a focus value and a lens position provided by an embodiment of the present application;

FIG. 2 is a schematic diagram of a system architecture of an autofocus method provided by an embodiment of the present application;

FIG. 3 is a flowchart of an automatic focusing method provided by an embodiment of the present application;

4 is a schematic diagram of obtaining a reference position provided by an embodiment of the present application;

FIG. 5 is a flowchart of another autofocus method provided by an embodiment of the present application;

6 is a schematic structural diagram of an automatic focusing device provided by an embodiment of the present application;

FIG. 7 is a schematic structural diagram of an automatic focusing device provided by an embodiment of the present application.

Detailed ways

In order to make the objectives, technical solutions and advantages of the present application clearer, the embodiments of the present application will be further described in detail below with reference to the accompanying drawings.

In the field of video surveillance, how to focus is the key to clear imaging. In this regard, the embodiments of the present application provide an auto-focusing method, and the auto-focusing methods provided by the embodiments of the present application can be applied to scenes such as road monitoring and safe cities where the shooting environment needs to be focused to ensure that clear images can be collected . Exemplarily, scenarios to which the auto-focusing method provided in the embodiments of the present application can be applied include but are not limited to the following.

Scenario 1. Road monitoring

In order to monitor traffic conditions and vehicles passing on the road in real time, a bayonet device is usually installed on the road to realize real-time monitoring of all vehicles passing through the current road. A camera is installed on the bayonet device to monitor the monitoring range. real-time image acquisition of the vehicle. In this scenario, using the automatic focusing method provided by the embodiment of the present application, the camera can detect in real time whether a target vehicle enters the monitoring range according to the image collected by the camera, so as to trigger automatic focusing and realize real-time tracking and monitoring of the target vehicle. , to ensure a clear image of the target vehicle.

Scene 2. Safe City

Safe City is a management system that can meet the needs of urban management, traffic management, emergency command, etc., and can also take into account the needs of image monitoring in aspects such as accident early warning and safety production monitoring. Through the image monitoring system, online visual video monitoring can be implemented. In this scenario, the automatic focusing method provided by the embodiment of the present application is applied to a camera in an image monitoring system, and the camera can monitor in real time whether a target object enters the monitoring range, so as to trigger automatic focusing and realize real-time tracking of the target object and monitoring to ensure clear imaging of the target object.

It should be noted that the above scenarios are only exemplary descriptions, the auto-focusing method provided in the embodiments of the present application can be applied to various scenarios that require focusing, and the embodiments of the present application do not limit the application scenarios of the auto-focusing method.

For easier understanding, before introducing the technical solutions provided by the embodiments of the present application, the technical terms involved in the embodiments of the present application are described below.

Focusing, the imaging principle of the camera is to use the lens to focus the light for imaging. In this process, the focus of the light focusing does not necessarily fall on the photosensitive plane. The focusing lens needs to be adjusted by the focusing motor so that the imaging point falls on the photosensitive plane, so that the camera picture is clear. Imaging, this process is the focusing process.

Depth of focus is the abbreviation for depth of focus. The range of the focal point and the photosensitive plane is within a certain range, which is clear to the human eye, and the clear range of the focus is called the depth of focus.

Depth of field, when an object is in focus, all the scenery from a certain distance in front of the object to a certain distance behind it should also be in focus. This distance from front to back is called the depth of field.

Object distance, the distance from the object to the imaging plane. In this application, relative to the object distance, the internal parameters of the camera lens can be ignored. Therefore, the object distance is approximately considered to be the distance from the object to be photographed to the first lens in front of the camera.

The focal length, the camera parameter, refers to the distance from the optical center of the lens in the camera to the imaging plane such as the film and the image sensor.

In a multi-object-distance scene, in the same scene, objects at different positions on the screen have different distances from the camera. This kind of scene is called a multi-object-distance scene.

Focus value (FV), also known as sharpness statistic, is used to evaluate the sharpness of an image. Schematically, referring to FIG. 1, FIG. 1 is a schematic diagram of the relationship between a focus value and a lens position provided by an embodiment of the present application. As shown in FIG. 1, the focus evaluation function in FIG. 1 is also the focus value FV, which is usually in In a fixed scene, the FV of the image changes with the movement of the lens position, showing a Gaussian-like function shape like a hillside, and the lens position corresponding to the peak of the hillside corresponds to the clear point of the image. Among them, the lens position is controlled by the focus motor. In the related art, the calculation methods of FV include gray gradient method and high frequency component method.

The climbing algorithm is a focusing algorithm that searches for the maximum value of FV within a certain range. At the beginning of the climbing algorithm, an initial motor movement direction will be given. After the motor moves, the FV before and after the movement are compared. If the FV becomes larger, the motor will continue to move in this direction, and if the FV becomes smaller, it will move in the opposite direction. motor. During the movement of the motor, the change of FV is constantly monitored. When the FV continues to increase, it means that the focus is constantly moving towards the photosensitive plane. When the FV drops, it means that the clear point has been reached, and the focusing process is over.

FIG. 2 is a system architecture of an auto-focusing method provided by an embodiment of the present application. As shown in FIG. 2 , the automatic focusing method provided in this embodiment of the present application is applied to the camera 200 . The cameras 200 include, but are not limited to, analog cameras, network cameras, and high-definition cameras. The camera 200 includes a lens 210 , an image sensor 220 , an image signal processing unit 230 , a focus motor 240 and a central processing unit 250 .

The lens 210 is mainly used to collect light on the photographed area.

The image sensor 220 is used to convert the optical signal into an image analog electrical signal. Optionally, the image sensor 220 involved in the embodiment of the present application includes a charge-coupled device (CCD) image sensor. It also includes image sensors based on complementary metal oxide semiconductor (complementary metal oxide semiconductor, CMOS) processes, also known as CMOS image sensors (CMOS image sensors), which are used in network cameras, digital cameras and other electronic optical devices. Core components. Certainly, other types of image sensors may also be included in some embodiments, and the types of the involved image sensors 220 are not limited in this embodiment of the present application.

The image signal processing unit 230 (image signal process, ISP) is mainly used for processing the signal output by the image sensor 220, and processing the image signal through a series of digital image processing algorithms to obtain the FV of the image.

The focusing motor 240 is used to drive the lens 210 to move, so that the imaging point falls on the image sensor 220, so as to realize clear image of the picture.

A central processing unit (CPU) 250 is used to control the movement of the focusing motor 240 according to the FV, so as to realize focusing.

The autofocus method provided by the embodiments of the present application is exemplarily described below. FIG. 3 is a flowchart of an auto-focusing method provided by an embodiment of the present application. In the embodiment of the present application, the auto-focusing method is performed by the camera 200 shown in FIG. 2 . For example, the auto-focusing method can be applied to a camera 200 in the central processing unit 250. Exemplarily, the auto-focusing method includes the following steps.

301. The camera acquires at least two frames of images including the target object, and the at least two frames of images are acquired by the camera through image acquisition of the captured area.

In the embodiments of the present application, the shooting area refers to an area where image acquisition needs to be performed. The target object is the object that meets the autofocus conditions. Optionally, the target object is a moving object in the shooting area. For example, the target object is a motor vehicle, a non-motor vehicle, a pedestrian, etc., which is not limited in this application. Optionally, the target object is a preset designated object. For example, in a scenario involving motor vehicle monitoring, the vehicle that needs to be tracked in real time is set as the target object.

In a possible implementation manner, the manner in which the camera acquires at least two frames of images including the target object includes but is not limited to the following steps 311-314.

311. The camera detects that a target object exists in the shooting area of the camera.

In some embodiments, the camera includes an image storage module and a detection module, the image storage module is used to store each frame of image obtained by the camera on the captured area, and the detection module stores the image in the image storage module. Detection is performed to determine if a target object is present in the shot area.

Optionally, when the camera collects images of the captured area, N frames of images are collected every second, and each frame of images is stored in the image storage module, and the images in the image storage module are detected by the detection module every M frames. To determine whether there is a target object in the shooting area; wherein, N and M are both positive integers, and N is greater than M. For example, the camera collects 25 frames of images per second, and stores each frame of the collected images in the image storage module in real time, and the detection module detects the images in the image storage module every 3 frames. It should be noted that, usually within the time interval of 3 frames of images, it is approximately considered that the position of the target object in the moving state in the photographed area does not change. Therefore, in the embodiment of the present application, the detection module detects every 3 frames. once. In practical applications, the developer can set the detection module to detect each frame of images according to requirements, and the application does not limit the detection frequency of the detection module.

Optionally, the camera is connected to the terminal through a wireless network or a wired network. A designated application program runs on the terminal, and the designated application program is used to set the functions of the camera. For example, the designated application program provides a setting option of the smart detection mode, the user can operate the terminal to turn on the smart detection mode, and the terminal sends a corresponding turn-on instruction to the camera in response to the user's operation of turning on the smart detection mode, and the turn-on instruction It is used to trigger the detection module to perform intelligent detection on the images stored in the image storage module.

Optionally, the camera detects whether there is a target object in the shooting area by using the detection model. This optional implementation is described in detail below, including the following steps 3111 and 3112:

3111. The camera inputs at least two frames of images into the detection model, where the at least two frames of images are acquired by the camera through image acquisition of the captured area.

The detection model is used to output the detection result of each image according to each input image. The detection result is used to indicate whether the target object exists in each image. If there is a target object in the image, the detection result includes the target area information of the target object in each image. Optionally, the target area information includes the identification of the target object, the size of the target area, and the location of the target area in each image.

In some embodiments, the detection model performs image semantic segmentation on an input frame of image, and classifies each pixel of the image to obtain the semantic category of each pixel to obtain an image segmentation result, and then based on the image segmentation result , and output the detection result of the image. Wherein, if there is a target object in the image, the image segmentation result includes the segmented image of the target object, and the detection model can determine the target area information of the target object in the image based on the segmented image to output the detection result.

In other embodiments, the detection model constructs a template for information such as edge contour, texture and gray level of each object in the image according to an input frame of image, and detects the target object by a template matching method. If the template of the object in the image matches the target object, the detection model determines the target area information based on the target area where the object is located in the image, so as to output the detection result.

It should be noted that the operation process of the above detection model is only illustrative, and the present application does not limit the method of detecting the target object of the detection model.

3112. In response to the detection result including the target area information of the target object in each image, the camera acquires the target area information.

Wherein, if the detection result output by the detection model includes the target area information of the target object in each image, it means that the target object currently exists in the area captured by the camera, the camera obtains the target area information, and executes the following step 312 . On the contrary, if the detection result output by the detection model does not include the target area information of the target object in each image, it means that there is currently no target object in the area captured by the camera, and the camera repeats this step 311 until the target is acquired Regional information.

It should be noted that, in the above steps 3111 and 3112, the detection model may be a detection model based on a specified scene. Optionally, the detection model is used to detect pedestrians with certain characteristics in the image to output detection results. For example, this detection model is used to detect males appearing in images. Optionally, the detection model is used to detect specific vehicles in the image to output detection results. For example, this detection model is used to detect public transit vehicles that appear in images. In practical applications, developers can train detection models according to requirements to obtain corresponding detection models, which are not limited in this application.

Through the above step 311, the camera detects whether there is a target object in the shooting area according to the image collected by the camera, and can determine the target area in the image in a targeted manner to obtain the target area information, which can not only become a priori for triggering automatic focusing information, and can also ensure clear imaging of the target object in the subsequent focusing process.

312. The camera acquires the current position of the focus motor in the camera.

In this embodiment of the present application, this step 312 is performed by the focusing module in the camera. Wherein, this step 312 includes any of the following situations:

Case 1: In response to the camera detecting the target object for the first time, the camera obtains the current position of the focus motor in the camera through prediction.

Case 2: In response to the camera not detecting the target object for the first time, the camera obtains the current position of the focus motor in the camera, and uses the current position as the current position.

It should be noted that, in this embodiment of the present application, regardless of whether the camera detects the target object for the first time or not for the first time, a current position can be obtained, which is convenient for subsequent focusing cycles through the current position. That is to say, as long as the camera can detect the target object, it will continue to focus on the cycle, and will obtain a current position during each focusing process. If the target object is detected for the first time, it will obtain a current position through prediction, and then in the The first focusing is performed on the basis of the current position. If the target object is detected for the second time, the current position of the focusing motor is obtained, that is, the position of the focusing motor when the first focusing is completed, and then the current position of the focusing motor is obtained. Based on this position, a second focus is performed, and so on, and the cycle continues until the camera no longer detects the target object. Optionally, a detection period may also be set, and if the target object is detected within the detection period, the cycle will continue, and if the detection period ends, no further detection will be performed.

313. The camera acquires at least one reference position based on the current position, and the distance between the at least one reference position and the current position is an adjustment step.

In this embodiment of the present application, this step 313 is performed by the focusing module in the camera. The adjustment step length refers to the moving step length of the focus motor of the camera. Optionally, the adjustment step length is half the focal depth, which is not limited in this application. Optionally, based on the current position, the camera obtains a reference position on both sides of the current position, wherein the adjustment step is half a focal depth, and the focal depth is calculated based on the current position.

Taking the camera to obtain two reference positions based on the current position as an example, this step will be described in detail below:

Optionally, referring to FIG. 4 , FIG. 4 is a schematic diagram of obtaining a reference position according to an embodiment of the present application. As shown in FIG. 4 , the current position acquired by the camera is P1, and the positions on both sides of the position P1, which are separated by half the focal depth, are the reference positions P2 and P3, respectively. Among them, the depth of focus is calculated by the following formula (1):

Depth of focus = a×F×L×f

In formula (1), a is a fixed parameter of the camera, and the value of a is different for different models of cameras; F is the aperture value of the camera; L is the object distance, which can be determined by formula (7) and formula (8) in the subsequent step 3121. Calculated; f is the focal length.

It should be noted that, in some embodiments, the camera can obtain W reference positions based on the current position, where W is a positive integer greater than 2. For example, on both sides of the current position, at every half depth of focus, obtain one reference position. With reference to the motor position, a total of 4 reference positions are obtained. This application does not limit this.

314. The camera controls the focus motor in the camera to move to the current position and the at least one reference position respectively, and acquires at least two frames of images when the focus motor in the camera is at the current position and the at least one reference position.

In this embodiment of the present application, this step 314 is performed by the focusing module in the camera. Wherein, the at least two frames of images are obtained by image capturing of the captured area by the camera. Wherein, each time the focus motor in the camera is located at a position, the camera acquires a frame of image of the corresponding position.

302. For each image in the at least two frames of images, the camera determines a target area where the target object is located in each image, and acquires the FV corresponding to each target area.

In this embodiment of the present application, this step 302 is performed by the focusing module in the camera. Among them, FV is used to evaluate the image sharpness of the target area. Optionally, the calculation method of FV may adopt a gray gradient method or a high frequency component method, etc., which is not limited in this application.

In some embodiments, for each image in the at least two frames of images, the camera determines the target area where the target object is located in each image according to the target area information obtained in step 301, so as to obtain the target area corresponding to each area. the FV.

In other embodiments, for each image in the at least two frames of images, the camera inputs each image into a detection model, and the detection model is configured to output a detection result of each image according to each input image, the The detection result includes the target area information of the target object in each image. Then, according to the target area information in the detection result, the camera determines the target area where the target object is located in each image, so as to obtain the FV corresponding to each target area. It should be noted that, the implementation of the detection model has been described in the above step 301, so it will not be repeated here.

The first point that needs to be explained is that the embodiments of the present application do not limit the manner in which the camera determines the target area where the target object is located in each image. That is, the camera can determine according to the target area information obtained from step 301, and can also input each frame of image into a detection model for detection, so as to determine the target area information.

The second point that needs to be explained is that the

above steps

301 and 302 can be executed synchronously, that is, whenever the camera controls the focus motor to move to a motor position, a frame of image is immediately acquired, and it is determined that the target object is located in the image. located in the target area, and calculate the FV corresponding to a target area.

Optionally, the above steps are specifically described below by taking the camera acquiring two reference positions and adjusting the step size as half the focal depth as an example. For example, with continued reference to FIG. 4 , the

above steps

301 and 302 can be replaced with the following steps 1 to 3:

Step 1. The camera controls the focus motor to move to the current position P1, obtains the first frame of image, and determines the target area where the target object is located in the first frame of image based on the target area information, and then calculates the first target. The FV corresponding to the region.

Step 2: The camera controls the focusing motor with P1 as the starting point and half the focal depth as the adjustment step, moves forward to reach the reference position P2, obtains the second frame of image, and determines that the target object is in the second frame based on the target area information. The target area located in the frame image, and then calculate the FV corresponding to the second target area.

Step 3: The camera controls the focusing motor with P2 as the starting point and a focal depth as the adjustment step, moves in the reverse direction to reach the reference position P3, obtains the third frame of image, and determines the target object in the third frame based on the target area information The target area located in the image, and then calculate the FV corresponding to the third target area.

It should be noted that, what is shown in FIG. 4 is only schematic. In practical applications, the camera can control the focus motor to first move in the reverse direction to reach P3, and then move forward to reach P2, which is not limited in this application. In addition, it should be noted that the forward and reverse directions in the above steps are relative to the default moving direction of the focusing motor, and the default moving direction may be preset, which is not limited in this embodiment of the present application.

The following takes the camera to obtain the FV corresponding to a target area based on one frame of image as an example to illustrate:

For example, when the camera acquires a frame of image, it also acquires the FV of the entire area of the image. Usually, a frame of image is divided into 255 sub-areas, and each sub-area corresponds to a sub-FV. Determine the target area where the target object is located in the image, and then add the FVs of all sub-areas corresponding to the target area to obtain the FV corresponding to the target area. It should be noted that, in practical applications, the FV corresponding to the target area can be calculated in various ways. In some embodiments, the camera can also obtain the average value of the FV of all sub-areas corresponding to the target area to calculate The average value is taken as the FV corresponding to the target area. The embodiment of the present application does not limit the calculation method of the FV of the target area.

303. The camera acquires the target position of the focusing motor in the camera based on the obtained at least two FVs, and controls the focusing motor in the camera to move to the target position, so that the focus of the light passing through the focusing lens falls on the photosensitive plane of the camera, The FV corresponding to the target position meets the target focus condition.

In this embodiment of the present application, this step 303 is performed by the focusing module in the camera. The target focusing condition means that the imaging of the target object meets the definition requirement. In this step, the camera takes the obtained at least two FVs as the ordinate and the position of the focusing motor corresponding to the FVs as the abscissa, and obtains the target position through fitting.

Optionally, in this step 303, the camera obtains the target position of the focus motor in the camera based on the obtained at least two FVs, including the following steps 3031 and 3032:

3031. The camera acquires an association relationship between each FV and the position of the focus motor in the camera.

3032. The camera obtains the target position of the focus motor in the camera based on each FV, the current position, and the at least one reference position and through the association relationship.

The following steps 3031 and 3032 are described in detail by taking two reference positions as an example:

For example, continuing to refer to FIG. 4 , as shown in FIG. 4 , the current position of the focus motor of the camera is P1, and the reference positions are P2 and P3 respectively, and these three positions correspond to the FV of a target area respectively. The camera obtains the relationship between each FV and the position of the focus motor in the camera, that is, the following formula (2). Further, based on the formula (2), the camera obtains the relational expressions between the motor positions P1, P2, P3 and the FV corresponding to each motor position, as shown in formulas (3) to (5).

In formula (2), y represents FV, x represents the motor position, a and c are variable parameters, and b is the target position; in the above formulas (3) to (5), x ₁ , x ₂ and x ₃ respectively represent The current position P1 and the reference positions P2 and P3, y ₁ , y ₂ and y ₃ respectively represent the FVs of the target areas corresponding to P1 , P2 and P3 .

By fitting the above formula (3) to formula (5), the following formula (6) can be obtained:

According to the above formula (6), the value of b can be calculated, and b is the target position of the focusing motor.

The first point that needs to be explained is that in the above steps, by calculating the FV corresponding to the target area, the influence of multiple object distances can be avoided, so that the camera can only focus on the target object, which greatly increases the imaging clarity of the target object. It is also possible to reduce the amount of data for image processing.

The second point that needs to be explained is that, in some embodiments, the camera obtains two reference positions on the basis of the current position, so that the time required for the camera to perform the above steps 301 to 303 is only 4 frames of time. For example, it is only 0.16 seconds. During this time period, the target object can be approximately considered to be in a stationary state. Based on this, the accuracy of the target position can be guaranteed to the greatest extent. At the same time, the focusing time is greatly reduced and the focusing efficiency is improved.

The third point that needs to be explained is that, after the above steps 301 to 303, when the camera detects that there is a target object in the area captured by the camera, automatic focusing is triggered, and the focusing module in the camera is switched from the idle state to the focusing state.

The two situations in the above step 312 are described in detail below:

Taking case 1 as an example, the above step 312 is replaced with: in response to the camera detecting the target object for the first time, the camera obtains the current position of the focus motor in the camera through prediction.

In the embodiment of the present application, the camera further includes a focusing module, which can be used to perform step 312, wherein, when the focusing module is in an idle state and the camera detects the target object, it is determined that the camera detects the target object for the first time, and the focusing module is idle Status means that the focus module is not currently in the focus process. Optionally, the focusing module detects whether the camera acquires the target area information every frame. If the camera acquires the target area information when the focusing module is in an idle state, it is determined that the camera detects the target object for the first time.

The following describes the specific implementation of the camera to obtain the current position of the focus motor in the camera through prediction, including the following steps 3121 and 3122:

3121. The camera acquires a predicted object distance between the camera and the target object.

Among them, the predicted object distance is calculated by the following formulas (7) and (8):

Object distance = target guess length / ratio of detection frame to the entire screen / tangent function of field of view

(7)

field of view = acrtan (sensor size/focal length)

(8)

The formula (7) and formula (8) are elaborated below, including the following three points:

First, in Equation (7), the target guess length is the predicted value of the target object length. Optionally, the camera pre-sets the value of the target guess length. For example, when the mark of the target object in the target area information indicates that the target object is a vehicle, the target guess length is 2.2 meters (m). The specific value of the target guess length in this application is Not limited. It should be noted that the target guess length in the above formula (7) may also be the target guess height. Optionally, the camera pre-sets the value of the target guessing height. For example, when the target object identifier in the target area information indicates that the target object is a pedestrian, the target guessing height is 1.7m. In practical applications, a corresponding target guess length can be set according to requirements, which is not limited in this embodiment of the present application.

Second, in formula (7), the ratio of the detection frame to the entire image refers to the ratio of the size of the target area to the entire image. Optionally, the camera calculates the area of the target area according to the size of the target area in the target area information, so as to obtain the ratio of the detection frame to the entire screen.

Third, in formula (7), the field of view is calculated by formula (8), where the sensor size refers to the size of the imaging plane of the image sensor.

3122. The camera acquires the current position of the focus motor in the camera based on the predicted object distance and the focal length of the camera.

The focal length of the camera is a real-time camera parameter that can be obtained directly. In some embodiments, the camera further includes a parameter storage module for storing various parameter information of the camera. Among them, the parameter storage module stores a table for querying the position of the focusing motor, and the camera sends a query request to the parameter storage module according to the calculated predicted object distance and the current focal length of the camera, and the query request is used to predict the object distance and focal length according to The corresponding focusing motor position is queried, and the parameter storage module sends the corresponding focusing motor position to the focusing module based on the query request.

Through the above step 312, the camera gives an estimated object distance by predicting the object distance between the camera and the target object, so as to obtain an initial position of the focus motor, and when the focus motor is at the initial position, the target object A certain degree of clear imaging can be achieved.

It should be noted that, in case 1 as an example, after the above steps 301 to 303, the camera responds to the camera detecting the target object for the first time, and calculates the predicted object distance of the target object according to the target area where the target object is located. Based on this , obtain the current position of the focusing motor through prediction, and determine at least one reference position, finally obtain at least two FVs, and obtain the target position through fitting. Through the above method, when the camera controls the focusing motor to move to the target position, the light focus of the focusing lens can fall on the photosensitive plane of the camera. Clear imaging can be achieved.

Taking case 2 as an example, the above step 312 is replaced with: the camera, in response to the camera detecting the target object for the first time, obtains the current position of the focus motor in the camera, and uses the current position as the current position.

In the embodiment of the present application, this step 312 is specifically performed by the focusing module in the camera, wherein when the focusing module is in the focusing state and the camera obtains the target area information, it is determined that the camera has not detected the target object for the first time. Optionally, the focusing module detects whether the target area information is received every frame, and if the focusing module continues to obtain the target area information when it is in the focused state, it is determined that the camera has not detected the target object for the first time.

Optionally, the following takes the second acquisition of the target area information by the camera as an example for description:

In this step, in response to the camera not detecting the target object for the first time, the camera obtains the current position of the focus motor, that is, in case 1, the target position in the above step 303 is taken as the target position. The current position, then step 313 and step 314 in the above-mentioned step 301 are executed, and then step 302 and step 303 are executed to finally obtain the target position.

In some embodiments, this step 312 can also be replaced by the following steps 312A to 312C:

312A. The camera controls the focus motor in the camera to move to the current position, and acquires a frame of image. At the same time, the focus module acquires a frame of image captured by the camera when the focus motor was at the current position during the last focusing process.

312B. For each image in the two frames of images, the camera determines a target area where the target object is located in each image, and obtains the FV corresponding to each target area.

312C, the camera obtains the target position of the focusing motor in the camera based on the obtained two FVs, and controls the focusing motor in the camera to move to the target position, so that the focus of the light passing through the focusing lens falls on the photosensitive plane of the camera, the The FV corresponding to the target position meets the target focus condition.

Through the above steps 312A to 312C, the camera obtains two frames of images, one frame is the image when the focus motor is at the current position during the current focusing process, and the other frame is when the focus motor is at the current position during the last focusing process. Based on this, the camera obtains the FV corresponding to the two target areas, and obtains the target position by comparing the numerical values of the two FVs. For example, if the FV obtained by the camera during the current focusing process is smaller than the FV during the previous focusing process, the camera controls the focusing motor to move forward by half the focal depth adjustment step to obtain the target position; The FV in this focusing process is greater than the FV in the previous focusing process, and the camera controls the focusing motor to reversely move half the focal depth adjustment step to obtain the target position. It should be noted that the above examples are only illustrative, and in practical applications, developers can set different moving directions and adjust the step length to obtain the target position according to requirements, which is not limited in this embodiment of the present application.

In addition, it should be noted that, in this embodiment of the present application, as long as the camera can detect the target object, the target area information will be obtained, the camera will always perform automatic focusing, and the target position will be obtained each time the focusing is completed. A new focus is started again, so as to quickly adjust the position of the focus motor, so that when the target object is moving, the target object in the image captured by the camera can still be clearly imaged. When the camera no longer detects the target object, the target area information cannot be obtained. At this time, the focusing module in the camera switches from the focused state to the idle state until the camera obtains the target area information again.

In the automatic focusing method provided by the embodiment of the present application, when there is a target object in the shooting area of the camera, the automatic focusing is triggered. At this time, the camera acquires at least two frames of images including the target object, and respectively determines that the target object is in each frame of the image. Then, only the FV of the target area in each frame of image is calculated to obtain the target position for clear imaging of the target object. Based on this, the focus motor is controlled to move to achieve focusing. That is to say, the automatic focusing method is based on the clarity of the target area where the target object is located, directly obtains the target position of the focusing motor, and then controls the focusing motor to move to the target position, which can avoid the influence of multiple object distances. The camera always takes the clear imaging of the target object as the goal to automatically focus, so as to realize the clear imaging of the target object, and this method can greatly shorten the focusing time and improve the focusing efficiency. Further, as long as the camera can continue to detect the target object, the camera will focus again on the basis of the previous focus, so as to continuously refresh the motor position of the focusing motor, and realize real-time tracking of the moving target object, thus ensuring dynamic scenes. clear imaging of the target object.

FIG. 5 is a flowchart of another automatic focusing method provided by an embodiment of the present application. The automatic focusing method provided by the embodiment of the present application will be schematically described below with reference to FIG. 5 . As shown in FIG. 5 , the automatic focusing method includes the following steps 501 to 505 .

501. The camera acquires 25 frames of images per second, and sends each frame of image to the camera platform, where the camera platform is also the CPU in the camera. When the intelligent detection mode is turned on, the detection module in the camera platform detects the target object in the image every 3 frames. Among them, the detection module will import the image as input into a trained detection model, and finally output the detection result of the detection model for the target object in the whole image. If the target object is detected, the detected target will be in the output result. area information, and transmit the detected target area information to the focusing module in the camera platform.

502. The focusing module in the camera platform detects whether the target area information of the detection module is received in each frame. If the focusing module receives the target area information of the detection module when it is in an idle state, it will trigger automatic focusing; after the focusing module completes one cycle, it will continue to detect whether it can still receive the prior information of the intelligent algorithm. , then continue the loop, if it cannot receive, then exit the loop and enter the idle state.

503. If the focusing module receives the target area information when it is in an idle state, that is, the target object is detected for the first time, the predicted object distance will be estimated first according to the above formula (7), and then, according to the predicted object distance and the current focal length. Look up the table to obtain the corresponding motor position, and use the motor position as the current position.

504. The focusing module obtains a reference position on the left and right sides of the current position, respectively. Then the focusing module controls the focusing motor to move to the three positions in turn, and obtains the FV of the target area corresponding to the three positions at the same time, and then fits the target position of the clearest point according to the correlation shown in the above formula (2). , the focusing module controls the focusing motor to move to the target position to complete a focusing cycle.

505. If the focusing module still receives the target area information after one cycle, that is to say, the target object is not detected for the first time, the focusing module will obtain the position decided by the focusing module after focusing last time, and use the position as the new current position again. , and repeat step 504 to obtain the FV of the target area, obtain the target position of the clearest point by fitting, and control the focusing motor to move to the target position again to complete the focusing cycle again.

FIG. 6 is a schematic structural diagram of an auto-focusing apparatus provided by an embodiment of the present application. The auto-focusing apparatus is used to execute the steps of the above-mentioned auto-focusing method. Referring to FIG. 6 , the auto-focusing apparatus 600 includes: an image acquisition module 601, FV acquisition module 602 and control module 603 .

An image acquisition module 601, configured to acquire at least two frames of images including the target object, the at least two frames of images are obtained by the camera on the captured area;

The FV acquisition module 602 is configured to, for each of the at least two frames of images, determine the target area where the target object is located in each of the images, and acquire the FV corresponding to each of the target areas;

The control module 603 is configured to obtain the target position of the focusing motor in the camera based on the obtained at least two FVs, and control the focusing motor in the camera to move to the target position, so that the focus of the light passing through the focusing lens falls on the camera On the photosensitive plane of , the FV corresponding to the target position meets the target focusing conditions.

In an optional implementation manner, the image acquisition module 601 includes:

In an optional implementation manner, the first acquisition unit is configured to acquire the current position by prediction in response to the camera detecting the target object for the first time; acquire the current position in response to the camera detecting the target object for the first time The current position of the focus motor in the camera is taken as the current position.

In an optional implementation manner, the first obtaining unit is configured to obtain the predicted object distance between the camera and the target object; and obtain the current position based on the predicted object distance and the focal length of the camera.

In an optional implementation manner, the adjustment step is half the focal depth.

In an optional implementation manner, the first obtaining unit is configured to obtain the reference position on both sides of the current position based on the current position.

In an optional implementation manner, the control module 603 is configured to acquire the association relationship between each FV and the position of the focus motor in the camera; based on each FV, the current position and the at least one reference position , and through the relationship, the target position of the focus motor in the camera is obtained.

In an optional implementation manner, the FV acquisition module 602 is configured to input each image into a detection model, and the detection model is configured to output the detection result of each image according to the inputted each image , the detection result includes the target area information of the target object in each image; the target area where the target object is located in each image is determined based on the detection result.

In the auto-focusing device, when there is a target object in the shooting area of the camera, auto-focusing is triggered. At this time, the camera acquires at least two frames of images including the target object, and respectively determines the target where the target object is located in each frame of image. Then only the FV of the target area in each frame of image is calculated to obtain the target position to make the target object clearly imaged. Based on this, the focus motor is controlled to move to achieve focusing. That is to say, the automatic focusing device is based on the clarity of the target area where the target object is located, directly obtains the target position of the focusing motor, and then controls the focusing motor to move to the target position, which can avoid the influence of multiple object distances. The camera is made to automatically focus on the clear imaging of the target object at all times, so as to realize the clear imaging of the target object, which can greatly shorten the focusing time and improve the focusing efficiency. Further, as long as the target object can continue to be detected, the automatic focusing device will focus again on the basis of the previous focusing, so as to continuously refresh the motor position of the focusing motor, and realize the real-time tracking of the moving target object, so as to ensure dynamic Clear imaging of target objects in the scene.

FIG. 7 is a schematic structural diagram of an auto-focusing device provided by an embodiment of the present application. The auto-focusing device 700 may vary due to different configurations or performances, and includes one or more processors 701 and one or more memories 702 , wherein at least one piece of program code is stored in the memory 702, and the at least one piece of program code is loaded and executed by the processor 701 to implement the steps in the above method embodiments. Of course, the auto-focusing device 700 can also have components such as a wired or wireless network interface, a keyboard, and an input-output interface for input and output. The auto-focusing device 700 also includes other components for realizing device functions, which will not be repeated here. .

In an exemplary embodiment, a computer-readable storage medium, such as a memory including program codes, is also provided, and the program codes can be executed by a processor in the terminal, so that the computer can perform the auto-focusing method in the above-mentioned embodiments. For example, the computer-readable storage medium is read-only memory (ROM), random access memory (RAM), compact disc read-only memory (CD-ROM), magnetic tape , floppy disks and optical data storage devices.

In an exemplary embodiment, an auto-focusing camera is also provided, for example comprising an image sensor and a processor, wherein the image sensor is used to acquire at least two frames of images in the above-mentioned embodiments, and the processor is used to execute the above-mentioned embodiments Autofocus method in .

In this application, the terms "first", "second" and other words are used to distinguish the same or similar items with basically the same function and function, and it should be understood that between "first", "second" and "nth" There are no logical or timing dependencies, and no restrictions on the number and execution order. It will also be understood that, although the following description uses the terms first, second, etc. to describe various elements, these elements should not be limited by the terms. These terms are only used to distinguish one element from another. For example, a first image may be referred to as a second image, and, similarly, a second image may be referred to as a first image, without departing from the scope of various described examples. Both the first image and the second image may be images, and in some cases, may be separate and distinct images.

The term "at least one" in this application means one or more, and the term "plurality" in this application means two or more, for example, a plurality of second images means two or more Second image above. The terms "system" and "network" are often used interchangeably herein.

It should also be understood that the term "if" may be interpreted to mean "when" or "upon" or "in response to determining" or "in response to detecting." Similarly, depending on the context, the phrases "if it is determined..." or "if a [statement or event] is detected" can be interpreted to mean "when determining..." or "in response to determining... ” or “on detection of [recited condition or event]” or “in response to detection of [recited condition or event]”.

The above descriptions are only specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of various equivalent modifications within the technical scope disclosed in the present application. or replacement, these modifications or replacements should be covered within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

In the above-mentioned embodiments, it may be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented in software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer program instructions. When the computer program instructions are loaded and executed on a computer, the procedures or functions according to the embodiments of the present application are generated in whole or in part. The computer may be a general purpose computer, a special purpose computer, a computer network, or other programmable device.

The computer instructions may be stored in or transmitted from one computer readable storage medium to another computer readable storage medium, for example, the computer program instructions may be transmitted from a website site, computer, server or data center via Wired or wireless transmission to another website site, computer, server or data center. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. that includes one or more available media integrated. The available media may be magnetic media (eg, floppy disks, hard disks, magnetic tapes), optical media (eg, digital video discs (DVDs), or semiconductor media (eg, solid state drives), and the like.

Those of ordinary skill in the art can understand that all or part of the steps of implementing the above embodiments can be completed by hardware, or can be completed by instructing relevant hardware through a program, and the program can be stored in a computer-readable storage medium. The storage medium can be read-only memory, magnetic disk or optical disk, etc.

As mentioned above, the above embodiments are only used to illustrate the technical solutions of the present application, but not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand: The technical solutions described in the embodiments are modified, or some technical features thereof are equivalently replaced; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

An automatic focusing method, characterized in that, applied to a camera, the method comprising:

Acquiring at least two frames of images including the target object, the at least two frames of images are obtained by image capturing of the captured area by the camera;

for each of the at least two frames of images, determining a target area where the target object is located in each of the images, and obtaining a focus value FV corresponding to each of the target areas;

Based on the obtained at least two FVs, the target position of the focusing motor in the camera is acquired, and the focusing motor in the camera is controlled to move to the target position, so that the focus of the light passing through the focusing lens falls on the light-sensing of the camera On the plane, the FV corresponding to the target position meets the target focusing condition.
The method according to claim 1, wherein the acquiring at least two frames of images including the target object comprises:

Acquire the current position of the focus motor in the camera, obtain at least one reference position based on the current position, and the distance between the at least one reference position and the current position is the adjustment step;

controlling the focus motor in the camera to move to the initial position and the at least one reference position, respectively, when the focus motor in the camera is at the current position and the at least one reference position, the camera The at least two frames of images are acquired.
The method according to claim 2, wherein the acquiring the current position of the focus motor in the camera comprises one of the following:

obtaining the current position by prediction in response to the camera detecting the target object for the first time;

In response to the camera not detecting the target object for the first time, the current position of the focus motor in the camera is acquired, and the current position is used as the current position.
The method according to claim 3, wherein the obtaining the current position by prediction comprises:

obtaining the predicted object distance between the camera and the target object;

The current position is acquired based on the predicted object distance and the focal length of the camera.
The method according to any one of claims 2-4, wherein the adjustment step is half a focal depth.
The method according to any one of claims 2-5, wherein the acquiring at least one reference position based on the current position comprises:

Based on the current position, one of the reference positions is acquired on both sides of the current position.
The method according to any one of claims 1-6, wherein the obtaining the target position of the focus motor in the camera based on the obtained at least two FVs comprises:

obtaining the relationship between each FV and the position of the focus motor in the camera;

Based on each FV, the current position, and the at least one reference position, the target position of the focus motor in the camera is acquired through the association relationship.
The method according to any one of claims 1-7, wherein the determining a target area where the target object is located in each image comprises:

The each image is input into the detection model, and the detection model is used for outputting the detection result of each image according to the inputted each image, and the detection result includes the target object in the target area information in each image;

A target area where the target object is located in each of the images is determined based on the detection result.
An automatic focusing device, characterized in that the device comprises:

an image acquisition module, configured to acquire at least two frames of images including the target object, the at least two frames of images are acquired by the camera on the captured area;

The FV acquisition module is configured to, for each of the at least two frames of images, determine a target area where the target object is located in each of the images, and acquire a focus value FV corresponding to each of the target areas ;

The control module is configured to obtain the target position of the focusing motor in the camera based on the obtained at least two FVs, and control the focusing motor in the camera to move to the target position, so that the focus of the light passing through the focusing lens falls on the target position On the photosensitive plane of the camera, the FV corresponding to the target position meets the target focusing condition.
The device according to claim 9, wherein the image acquisition module comprises:

a first obtaining unit, configured to obtain the current position of the focus motor in the camera, and based on the current position, obtain at least one reference position, and the distance between the at least one reference position and the current position is the adjustment step size ;

a second acquiring unit, configured to control the focus motor in the camera to move to the initial position and the at least one reference position, respectively, and the focus motor in the camera is located at the current position and the at least one reference position When in position, the camera acquires the at least two frames of images.
The device according to claim 10, wherein the first acquiring unit is configured to acquire the current position by prediction in response to the camera detecting the target object for the first time; in response to the camera not detecting the target object for the first time; The target object is detected, the current position of the focus motor in the camera is acquired, and the current position is used as the current position.
The device according to claim 11, wherein the first obtaining unit is configured to obtain a predicted object distance between the camera and the target object; based on the predicted object distance and the focal length of the camera , to obtain the current position.
The device according to any one of claims 10-12, wherein the adjustment step is half a focal depth.
The apparatus according to any one of claims 10-13, wherein the first obtaining unit is configured to obtain one of the reference positions on both sides of the current position based on the current position.
The device according to any one of claims 9-14, wherein the control module is configured to obtain the association relationship between each FV and the position of the focus motor in the camera; The current position and the at least one reference position are obtained, and the target position of the focus motor in the camera is acquired through the association relationship.
The device according to any one of claims 9-15, wherein the FV acquisition module is configured to input the each image into a detection model, and the detection model is configured to input each image according to the input image, output the detection result of each image, the detection result includes target area information of the target object in each image; determine the target object in each image based on the detection result in the target area.
An auto-focusing camera, characterized in that the auto-focusing camera comprises an image sensor and a processor, the image sensor is used to collect at least two frames of images according to any one of claims 1 to 8, and the processor uses for performing the autofocusing method as claimed in any one of claims 1 to 8 .