A method for quickly determining the optimal focal plane position for imaging in large-span scenes

Abstract

The present invention provides a method for quickly determining the optimal focal plane position for imaging in a large-span scene, relates to the field of automatic focusing technology, and includes an automatic calibration device for a focus tracking curve, the automatic calibration device includes a calibration plate, a fixing device, a slide rail, a supporting device, a depth detection module, and an image acquisition and processing module, wherein the fixing device is used to support the calibration plate and rotate it, the slide rail is perpendicular to the calibration plate and fixed on the fixing device, the slide rail extends in a horizontal direction, the supporting device is slidably arranged on the slide rail, the depth detection module is used to obtain depth distribution data and transmit the data to the image acquisition and processing module, and the image acquisition and processing module includes a camera and a microcontroller for image processing and drive control. The automatic calibration device proposed in the present invention can calibrate different fixed-focus cameras and quickly and accurately obtain a focus tracking curve.

Description

Technical Field

[0001] The invention relates to the field of automatic focusing technology, and in particular to a method for quickly determining the best focal plane position for imaging in a large-span scene. Background Art

[0002] Autofocus technology is a technology that uses sensors and algorithms to automatically adjust the focal length of the lens. It allows the camera to automatically focus on the subject when shooting, so as to obtain a clear and sharp image. Therefore, how to determine the best focal plane position to obtain a clear image is the key to autofocus technology. The existing methods for determining the best focal plane position are mainly divided into active focusing and passive focusing. Active focusing measures the distance between the objective lens and the target through laser, infrared, etc., and drives the motor to focus according to the feedback distance. It is not affected by lighting conditions and is simple to operate, with a wide range of applications and high clarity. Passive focusing uses an image sensor to image the object in the scene, determines the current focus state through the focus evaluation function and the focus strategy, and then adjusts the focal length of the lens to achieve automatic focusing of the object. It has a low cost, a stable focusing process and high imaging efficiency. Global search algorithms and hill climbing algorithms are often used to achieve this. The global search method evaluates and determines the best focus position based on the image clarity of the camera at each focus position. Although it has high accuracy, the search efficiency is low. The hill climbing algorithm searches for the best focus position corresponding to the maximum image clarity with a fixed step size, which has the advantage of good real-time performance. However, when the focal distance is far, the hill climbing algorithm is prone to fall into the false peak phenomenon, resulting in focusing errors.

[0003] However, when imaging the surface in a scene with a large span and depth, traditional active or passive focusing methods often have problems such as insufficient depth of field or depth of field matching errors, resulting in poor imaging quality and slow speed. Wang et al. proposed a method of imaging with a large depth of field interval, which achieved focused imaging based on the intermediate depth in the divided interval in a large span scene, and obtained an image with a larger depth of field by fusing multiple focused images. On this basis, how to quickly obtain the optimal focal plane position in a large span scene and obtain a clear image requires a corresponding solution. Summary of the invention

[0004] The purpose of the present invention is to solve the problem that the prior art lacks a method for quickly obtaining the optimal focal plane position in a large-span scene and obtaining a clear image.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] An automatic calibration device for a focus tracking curve comprises a calibration plate, a fixing device, a slide rail, a supporting device, a depth detection module and an image acquisition and processing module, wherein the fixing device is used to support the calibration plate and rotate it, the slide rail is perpendicular to the calibration plate and fixed on the fixing device, the slide rail extends in a horizontal direction, the supporting device is slidably arranged on the slide rail, the supporting module is used to fix the depth detection module and the image acquisition and processing module, the depth detection module is used to obtain depth distribution data and transmit the data to the image acquisition and processing module, and the image acquisition and processing module comprises a camera and a microcontroller for image processing and drive control.

[0007] Preferably, the calibration plate comprises a large calibration plate and a small calibration plate, wherein the small calibration plate is fixed on the large calibration plate, and the small variable top plate is perpendicular to the direction of the camera visual axis, and the small calibration plate is laid with a sample set.

[0008] Preferably, the fixing device includes a base and a bracket located on the base, the bracket is fixed at both ends of the large calibration plate in the length direction, and a rotating device is provided on the side of the two brackets away from the large calibration plate, and the rotating device is used to control the rotation of the large calibration plate.

[0009] Preferably, one end of the slide rail is fixed on the base, and the other end extends outward in a horizontal direction. A slider and a stepper motor are provided on the slide rail. The stepper motor is used to drive the slider to slide on the slide rail. The supporting device is fixed on the slider. A pull rope encoder is also provided on the slide rail. The main body of the pull rope encoder is located at the connection between the slide rail and the base, and one end of the pull rope of the pull rope encoder is fixed on the slider.

[0010] Preferably, the supporting device includes a supporting frame, a control box and a camera fixing device, wherein a microcontroller is placed in the control box, the supporting frame is extended upward in a vertical direction, the control box and the camera fixing device are both fixed on the supporting frame, the camera fixing device includes a camera housing and a cross rod frame, the cross rod frame is used to connect the camera housing and the supporting frame, the camera is arranged in the camera, and a depth sensor is also arranged on the camera housing.

[0011] Preferably, the slider remains parallel to the camera visual axis when moving horizontally; the camera visual axis and a vertically placed large calibration plate remain perpendicular to each other, and the intersection of the straight line where the camera visual axis is located and the plane where the calibration plate is located is located at the geometric center of the calibration plate; the rope of the rope encoder remains horizontal with the camera visual axis.

[0012] The present application also provides a method for quickly determining the best focal plane position of an imaging camera in a large-span scene, comprising the following contents: S1: using the automatic calibration device for implementing the focus tracking curve to obtain a focus tracking curve of a distance-best focal plane position mapping relationship;

[0013] S2: Provide an extreme value search algorithm based on an improved hill climbing algorithm.

[0014] Preferably, the extreme value search algorithm uses an improved hill climbing search method to determine six image recording points, and applies a curve fitting method to these recording points, and uses the peak point obtained by fitting the curve as the optimal focal plane position.

[0015] Preferably, the specific steps of the extreme value search algorithm are as follows:

[0016] Step S1: Given the initial focus position P0, control the camera to drive the focus value to search along the direction of increasing image clarity with a constant step length ΔVF, using the Laplace focus evaluation function value La Pi To characterize the image clarity at the corresponding position.

[0017] Step S2: Continuous imaging, and judging the interval of the true peak value by the focus evaluation function value La: If La P3 <La P2 <La P1 (Judgment condition (1)), that is, the focus evaluation function value decreases continuously, then the corresponding focal plane position is recorded as P1 (the corresponding focus evaluation function value La P1 ), the focal plane positions corresponding to the two consecutive decreases in image clarity are P2 and P3.

[0018] Step S3: If La P2 <La P1 And La P3 >La P2 , that is, a local peak appears, then La P2 is the focal plane position demarcation point, the step length ΔVF is changed, and steps (1-2) are repeated for the left and right partitions to search for the focal plane position until the judgment condition (1) is met.

[0019] Step S4: Search in the opposite direction at a focal plane position of (P3-P2) / 2 with a constant step length ΔVF, repeating steps 1 to 2. P6 <La P5 <La P4 When , the corresponding focal plane position is recorded as P4. At this time, the focal plane positions where the clarity drops twice in succession are P5 and P6 respectively.

[0020] Step S5: During the hill climbing search process, 6 focus evaluation function values ​​(LaPi, i=1, 2, 3…6) and corresponding focal plane positions (P1-P6) are recorded, and a curve fitting is performed using discrete focus evaluation function values ​​and focal plane position parameters to obtain the optimal focal plane position.

[0021] Preferably, the experimental steps of constructing the focus tracking curve by the automatic calibration device are as follows:

[0022] Step S1: setting a depth range, and then dividing the depth range into a set of depth intervals according to a large depth of field interval division imaging method;

[0023] Step S2: placing a small calibration plate in the divided depth of field interval;

[0024] Step S3: The microprocessor drives the stepper motor to adjust the camera to a specified distance according to the depth of field interval, and the microprocessor controls the camera to focus and image, and evaluates the image clarity of the image collected by the camera, and obtains the best focal plane position corresponding to the distance;

[0025] Step S4: repeating steps S2-S3 for the remaining depth of field interval sets;

[0026] Step S5: Change the depth range, the depth range must overlap with the previous depth range by no less than 50% and the depth range must be the same length as the previous depth range, and repeat the above steps S2-S4;

[0027] Step S6: obtaining multiple sets of distances and corresponding optimal focal plane positions, and the microprocessor performs curve fitting using the distances and the optimal focal plane positions as coordinate points to obtain a distance-optimal focal plane position focus tracking curve.

[0028] Compared with the prior art, this application has the following beneficial effects:

[0029] 1) The extreme value search algorithm proposed in the present invention can avoid the problem of falling into a local peak in the hill climbing algorithm, and at the same time has a fast focusing speed and high focusing accuracy, and can provide a reliable method support for camera calibration to obtain a distance-optimal focal plane position focusing tracking curve.

[0030] 2) The present invention can be applied to a variety of fixed-focus cameras with integrated ranging functions, and the camera can be selected according to the needs. It has great universality and a wide range of applications.

[0031] 3) The automatic calibration device proposed in the present invention can calibrate different fixed-focus cameras and quickly and accurately obtain the focus tracking curve. After the depth sensor obtains the depth data, the optimal focal plane position can be quickly obtained according to the focus tracking curve in a large-span scene. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1It is a schematic diagram of the overall structure of an automatic calibration device for a focus tracking curve in one embodiment of the present invention;

[0033] Figure 2 It is a schematic structural diagram of a camera housing on an automatic calibration device for a focus tracking curve in one embodiment of the present invention;

[0034] Figure 3 This is a flow chart of an extreme value search algorithm in one embodiment of the present invention.

[0035] Legend:

[0036] 1. Large calibration plate; 2. Servo; 3. Small calibration plate; 4. Bracket; 5. Base; 6. Cable encoder; 7. Linear guide rail; 8. Camera housing; 9. Cross rod bracket; 10. Support bracket; 11. Control box;

[0037] 12. Slider; 13. Stepper motor; 14. Coupling; 15. Camera; 16. Depth sensor. DETAILED DESCRIPTION

[0038] The present invention is further described in detail below in conjunction with specific embodiments.

[0039] A method for quickly determining the optimal focal plane position of an imaging camera in a large-span scene is disclosed. The method adopts a focus tracking method combining active and passive focusing based on a range-finding camera, and determines the optimal focal plane position in the large-span scene by constructing a focus tracking curve, so as to ensure that a clear imaging effect can be quickly obtained in the scene.

[0040] Specifically, in one embodiment, the method includes the following contents:

[0041] (1) Obtaining the focus tracking curve of the distance-optimal focal plane position mapping relationship

[0042] Specifically, in one implementation, a mapping relationship between distance and optimal focal plane position, ie, a focus tracking curve, is established through the ranging function of the camera itself.

[0043] By establishing a mapping relationship of the focus tracking curve, the position of the optimal focal plane in a large-span scene can be quickly determined, thereby improving the clarity and quality of imaging.

[0044] In order to focus the tracking curve, in one embodiment, an automatic calibration device for the focus tracking curve is provided.

[0045] See also Figure 1 and Figure 2 The automatic calibration device includes a rotating device, a fixing device, a calibration plate, a motion module, a supporting device, a depth detection module and an image acquisition and processing module.

[0046] The calibration plate includes a large calibration plate 1 and a small calibration plate 3, wherein the small calibration plate 3 is fixed on the large calibration plate 1 and is perpendicular to the direction of the visual axis of the camera 15, and the small calibration plate 3 is covered with a sample set. In one embodiment, the sample set uses 1951USAF (1951 United States Air Force resolution test chart) to characterize the camera imaging clarity, and the calibration plate is used to construct different depth of field ranges;

[0047] The fixing device includes a base 5 and a bracket 4 placed on the base 5, and the bracket 4 is used to fix and install the rotating device; the base 5 is placed horizontally, and the bracket 4 is vertically arranged on the base 5 and located at both ends of the length direction of the large calibration plate 1.

[0048] The rotating device includes two steering gears 2 equipped with steering gear arms, and the two steering gears 2 are used to control the rotation of the large calibration plate 1. In one embodiment, the two steering gears 2 are located on the side of the bracket 4 away from the large calibration plate 1.

[0049] The motion module includes a linear motion device and a supporting device, wherein the linear motion device includes a linear guide rail, a slider 12, a coupling 14, a stepper motor 13 and a draw wire encoder 6, wherein the linear guide rail is fixed to a side edge of the base 5 and extends in a direction perpendicular to the large calibration plate 1, and the slider 12 is slidably arranged along the length direction of the linear slide rail 7. In one embodiment, the slider 12 remains parallel to the visual axis of the camera 15 when moving horizontally, and the coupling 14 is located at the end of the slide rail away from the base 5. The coupling 14 is used to connect the linear guide rail and the stepper motor 13, and the stepper motor 13 is used to control the precise sliding of the slider 12. The stretch encoder is used to feedback the object distance to realize closed-loop control of the linear motion device, and the body of the draw wire encoder 6 is mounted on the base 5, and its draw wire is fixed on the slider 12.

[0050] The supporting device is arranged above the slider 12. In one embodiment, the supporting device includes a supporting frame 10, a control box 11 and a camera fixing device. The supporting frame 10 is vertically arranged above the slider 12. The control box 11 is sleeved on the supporting frame 10. A microcontroller is arranged in the control box 11 for controlling each module of the automatic calibration device. The camera fixing device is composed of a camera housing 8 and a cross rod frame 9, wherein the cross rod frame 9 is used to connect the camera housing 8 and the supporting frame 10, and realize the sliding of the camera housing 8 in the height direction of the supporting frame 10.

[0051] The depth detection module uses an ultra-micro area array depth sensor based on the time of flight principle (TOF) of ST Company to obtain depth distribution data and transmit the obtained depth data to the microprocessor. In one embodiment, the depth detection module includes a depth sensor 16, and the depth sensor 16 is fixed on the camera housing 8.

[0052] The image acquisition and processing module includes a camera 15 and a microcontroller for image processing and drive control, wherein the camera 15 is fixed in the camera housing 8. In view of the advantages of a fixed-focus camera such as a wide field of view and high imaging quality, the camera 15 is a fixed-focus camera. The camera 15 is used for image acquisition and transmits the acquired information to the microcontroller. The microcontroller uses Raspberry Pi 4B as an image acquisition processor to realize image acquisition and clarity evaluation, and at the same time realizes the distance adjustment of the motion module and the control of the rotating device. In addition, the microcontroller can perform image processing and is a high-performance processing module with a large storage. In one embodiment, the visual axis of the camera 15 maintains a mutually perpendicular relationship with the vertically placed large calibration plate 1, and the intersection of the straight line where the visual axis of the camera 15 is located and the plane where the calibration plate is located is located at the geometric center of the large calibration plate 1. The visual axis of the camera 15 is arranged parallel to the rope of the rope encoder 6.

[0053] In one embodiment, in order to achieve control over the camera 15, the microcontroller quickly searches for the optimal focal plane position by controlling the adjustment of the voice coil motor of the camera 15, and when the microcontroller receives that the image taken by the camera 15 does not include all the small calibration plates 3 laid out within the depth range, the microcontroller will control the rotating device to change the inclination angle of the large calibration plate 1 so that the camera field of view can cover the entire depth range, and control the camera to take a new photo at this position.

[0054] (2) In order to further improve the speed and accuracy of the mapping relationship, an extreme value search algorithm based on an improved hill climbing algorithm is proposed. This algorithm can effectively improve the search efficiency and ensure that the optimal focal plane position can be found quickly and accurately in complex scenes.

[0055] In one embodiment, the extremum search algorithm uses an improved hill climbing search method to determine six image recording points, and applies a curve fitting method to these recording points, and the peak point obtained by fitting the curve is used as the optimal focal plane position. The specific steps of the flowchart are as follows Figure 3 , the specific steps are as follows:

[0056] Step S1: Given the initial focus position P0, control the camera to drive the focus value to search along the direction of increasing image clarity with a constant step length ΔVF, using the Laplace focus evaluation function value La Pi To characterize the image clarity at the corresponding position.

[0057] Step S2: Continuous imaging, and judging the interval of the true peak value by the focus evaluation function value La: If La P3 <La P2 <La P1 (Judgment condition (1)), that is, the focus evaluation function value decreases continuously, then the corresponding focal plane position is recorded as P1 (the corresponding focus evaluation function value La P1 ), the focal plane positions corresponding to the two consecutive decreases in image clarity are P2 and P3.

[0058] Step S3: If La P2 <La P1 And La P3 >La P2 , that is, a local peak appears, then La P2 is the focal plane position demarcation point, the step length ΔVF is changed, and steps (1-2) are repeated for the left and right partitions to search for the focal plane position until the judgment condition (1) is met.

[0059] Step S4: Search in the opposite direction at a focal plane position of (P3-P2) / 2 with a constant step length ΔVF, repeating steps 1 to 2. P6 <La P5 <La P4 When , the corresponding focal plane position is recorded as P4. At this time, the focal plane positions where the clarity drops twice in succession are P5 and P6 respectively.

[0060] Step S5: During the hill climbing search process, 6 focus evaluation function values ​​(LaPi, i=1, 2, 3…6) and corresponding focal plane positions (P1-P6) are recorded, and a curve fitting is performed using discrete focus evaluation function values ​​and focal plane position parameters to obtain the optimal focal plane position.

[0061] At the same time, in order to fully reflect the clarity level of the best focus position, the Laplace operator that characterizes the edge and detail features of the image is used as the evaluation standard for the image clarity of each focus position, and its focus evaluation function is:

[0062] Among them, g(x,y) represents the grayscale value at (x,y), and the La value is the evaluation result of the current image focusing effect.

[0063] In one embodiment, the experimental steps of constructing a focus tracking curve by the automatic calibration device are as follows:

[0064] S1: setting a depth range, and then dividing the depth range into a set of depth intervals according to a large depth of field interval division imaging method.

[0065] S2: Place a small calibration plate in the divided depth of field range.

[0066] S3: The microprocessor drives the stepper motor to adjust the camera to a specified distance according to the depth of field range. The microprocessor controls the camera to focus the image and evaluates the image clarity of the image collected by the camera, and obtains the optimal focal plane position corresponding to the distance.

[0067] S4: Repeat S2-S3 for the remaining depth of field interval sets.

[0068] S5: Change the depth range. The depth range must overlap with the previous depth range by no less than 50% and the depth range must be the same length as the previous depth range. Repeat S2-S4 above.

[0069] S6: multiple sets of distances and corresponding optimal focal plane positions are obtained, and the microprocessor performs curve fitting using the distances and the optimal focal plane positions as coordinate points to obtain a distance-optimal focal plane position focus tracking curve.

[0070] The rangefinder camera calibrated by the calibration device can quickly and accurately determine the optimal focal plane position by using the curve mapping relationship after detecting the scene depth information in a large-span scene, providing an effective solution for the further development of imaging technology.

[0071] The extreme value search algorithm proposed in this application can avoid the problem of falling into a local peak in the hill climbing algorithm. At the same time, it has a fast focusing speed and high focusing accuracy, and can provide a reliable method support for camera calibration to obtain a distance-optimal focal plane position focusing tracking curve.

[0072] In addition, the automatic calibration device for the focus tracking curve provided by the present application can be applied to a variety of fixed-focus cameras with integrated ranging functions, and the camera can be selected according to the needs, with great universality and a wide range of applications. Moreover, the automatic calibration device proposed in the present invention can calibrate different fixed-focus cameras and quickly and accurately obtain the focus tracking curve. After the depth sensor obtains the depth data, the optimal focal plane position can be quickly obtained according to the focus tracking curve in a large-span scene.

Claims

1. An automatic calibration device for a focus tracking curve, characterized in that: It includes a calibration plate, a fixing device, a slide rail, a supporting device, a depth detection module and an image acquisition and processing module, wherein the fixing device is used to support the calibration plate and rotate it, the slide rail is perpendicular to the calibration plate and fixed on the fixing device, the slide rail is extended in the horizontal direction, the supporting device is slidably arranged on the slide rail, the supporting device is used to fix the depth detection module and the image acquisition and processing module, the depth detection module is used to obtain depth distribution data and transmit the data to the image acquisition and processing module, and the image acquisition and processing module includes a camera and a microcontroller for image processing and drive control; The calibration plate includes a large calibration plate and a small calibration plate, wherein the small calibration plate is fixed on the large calibration plate, and the small calibration plate is perpendicular to the direction of the camera visual axis, and a sample set is laid on the small calibration plate; the sample set adopts 1951USAF to characterize the camera imaging clarity, and the calibration plate is used to construct different depth of field ranges; The fixing device includes a base and a bracket located on the base, the bracket is fixed at both ends of the large calibration plate in the length direction, and a rotating device is provided on the side of the two brackets away from the large calibration plate, and the rotating device is used to control the rotation of the large calibration plate; The base is placed horizontally, and the bracket is vertically arranged on the base and located at both ends of the length direction of the large calibration plate; the rotating device includes two servos equipped with servo arms, the two servos are used to control the rotation of the large calibration plate, and the two servos are located on the side of the bracket away from the large calibration plate.

2. The automatic calibration device for focus tracking curve according to claim 1, characterized in that: One end of the slide rail is fixed on the base, and the other end extends outward in the horizontal direction. A slider and a stepper motor are provided on the slide rail. The stepper motor is used to drive the slider to slide on the slide rail. The supporting device is fixed on the slider. A pull rope encoder is also provided on the slide rail. The main body of the pull rope encoder is located at the connection between the slide rail and the base, and one end of the pull rope of the pull rope encoder is fixed on the slider.

3. The automatic calibration device for focus tracking curve according to claim 2, characterized in that: The supporting device includes a supporting frame, a control box and a camera fixing device, wherein a microcontroller is placed in the control box, the supporting frame is extended upward in a vertical direction, the control box and the camera fixing device are both fixed on the supporting frame, the camera fixing device includes a camera housing and a cross rod frame, the cross rod frame is used to connect the camera housing and the supporting frame, the camera is set in the camera housing, and a depth sensor is also set on the camera housing.

4. The automatic calibration device for focus tracking curve according to claim 3, characterized in that: The slider remains parallel to the camera visual axis when moving horizontally; the camera visual axis and the vertically placed large calibration plate remain perpendicular to each other, and the intersection of the straight line where the camera visual axis is located and the plane where the calibration plate is located is located at the geometric center of the calibration plate; the pull rope of the pull rope encoder remains horizontal with the camera visual axis.

5. A method for quickly determining the optimal focal plane position of an imaging camera in a large-span scene, characterized in that: Contains the following content: S1: using the automatic calibration device for focus tracking curve according to any one of claims 1 to 4 to obtain a focus tracking curve of a distance-optimal focal plane position mapping relationship; The experimental steps for constructing the focus tracking curve using the automatic calibration device are as follows: Step S11: setting a depth range, and then dividing the depth range into a set of depth intervals according to a large depth of field interval division imaging method; Step S12: placing a small calibration plate in the divided depth of field interval; Step S13: The microprocessor drives the stepper motor to adjust the camera to a specified distance according to the depth of field interval, and the microprocessor controls the camera to focus the image and evaluate the image clarity of the image collected by the camera, and obtains the best focal plane position corresponding to the distance; Step S14: repeating steps S12-S13 for the remaining depth of field interval sets; Step S15: Change the depth range, the depth range must overlap with the previous depth range by no less than 50% and the depth range must be the same length as the previous depth range, and repeat the above steps S12-S14; Step S16: obtaining multiple sets of distances and corresponding optimal focal plane positions, and the microprocessor performs curve fitting using the distances and the optimal focal plane positions as coordinate points to obtain a distance-optimal focal plane position focus tracking curve; S2: Provide an extreme value search algorithm based on an improved hill climbing algorithm; the extreme value search algorithm uses the improved hill climbing search method to determine 6 image recording points, and applies the curve fitting method to these recording points, and the peak point obtained by fitting the curve is used as the optimal focal plane position; the specific steps are as follows: Step S21: Given an initial focus position P0, control the camera to drive the focus value to search along the direction of increasing image clarity with a constant step length ΔVF, using the Laplace focus evaluation function value La Pi To characterize the image clarity at the corresponding position; Step S22: Continuous imaging, and focusing evaluation function value La P1 To determine the true peak interval: if La P3 <La P2 <La P1 , that is, the focus evaluation function value decreases continuously, then the corresponding focal plane position is recorded as P1, and the focus evaluation function value La corresponding to P1 P1 , the focal plane positions corresponding to the two consecutive decreases in image clarity are P2 and P3; Step S23: If La P2 <La P1 And La P3 >La P2 , that is, a local peak appears, then La P2 is the focal plane position demarcation point, the step size ΔVF is changed, and S21-S22 are repeated for the left and right partitions to search the focal plane position until La is satisfied. P3 <La P2 <La P1 ; Step S24: Search in the opposite direction at a focal plane position of (P3-P2) / 2 with a constant step length ΔVF, and repeat steps S21-22. P6 <La P5 <La P4 When , the corresponding focal plane position is recorded as P4. At this time, the focal plane positions where the clarity drops twice in succession are P5 and P6 respectively. Step S25: During the hill climbing search process, record the six focus evaluation function values ​​La Pi , i = 1, 2, 3 ... 6 and the corresponding focal plane positions P1 to P6, the discrete focus evaluation function value and the focal plane position parameters are used to perform curve fitting to obtain the optimal focal plane position.

Images