CN116112793A

CN116112793A - Camera focusing method and system

Info

Publication number: CN116112793A
Application number: CN202211336039.1A
Authority: CN
Inventors: 胡露; 宋凯静; 袁永具; 张伟
Original assignee: Jiangxi Shinetech Precision Optical Company Ltd
Current assignee: Jiangxi Shinetech Precision Optical Company Ltd
Priority date: 2022-10-28
Filing date: 2022-10-28
Publication date: 2023-05-12

Abstract

The invention relates to the technical field of camera module focusing, in particular to a camera focusing method and a camera focusing system, wherein the method comprises the following steps: performing coarse inspection MTF focus finding test according to the coarse inspection searching range and coarse inspection stepping, and obtaining an optimal DAC value during coarse inspection through coarse inspection fitting and extremum judgment; according to the optimal DAC value during coarse detection and the fine detection stepping, performing fine detection SFR focus searching test, and obtaining the optimal DAC value during fine detection through fine detection fitting and extremum judgment; and acquiring SFR values of each view field of the image corresponding to the fine-detection optimal DAC value according to the fine-detection optimal DAC value, judging whether the SFR values meet the preset SFR value requirement, if so, taking the fine-detection optimal DAC value as a final optimal DAC value, and outputting. The method can avoid the problem of calibration of SFR and MTF, solve the problem of error found clear positions caused by the limitation of a fitting algorithm, and guarantee the accuracy of analytical force detection so as to improve the production efficiency.

Description

Camera focusing method and system

Technical Field

The invention relates to the technical field of camera module focusing, in particular to a camera focusing method and system.

Background

The camera module is divided into FF (Fixed Focus) and AF (Auto Focus) according to a focusing function; where the focal length of FF is fixed, and AF can change the focal length by moving the lens position to achieve autofocus. The camera module is mainly fixed focusing, and the camera module adopted in the fields of mobile phones, security protection, unmanned aerial vehicles, monitoring and the like at present has an automatic focusing function. The existing automatic focusing function mainly comprises a motor and a motor driver, wherein the motor and the motor driver are arranged in a camera module, and the motor driver drives the motor to enable the motor to push the camera module to move within a certain object distance range, so that automatic focusing is achieved.

In the production process of the camera module, the optimal DAC value (motor pushing position) is searched according to the project requirements of users (wherein the project requirements comprise the distance and even the middle focus object distance appointed by the users), and the optimal DAC value is burnt into the inside of the camera module or a storage medium, so that the produced camera module has an automatic focusing function meeting the project requirements. The DAC value can represent the pushing position of the motor, specifically, the DAC value can represent the magnitude of the current value for driving the lens motor, and different current values can control the motor to generate different pushing forces to push the camera module, so that the focal length of the camera module is adjusted, and the input DAC value and the output current are in a linear relationship, such as: the motor driver is typically 10 bits accurate, and the DAC value range is 2 ⁰ ～2 ¹⁰ -1＝0～1023；

In other words, the DAC value drives the camera module to focus, so to ensure the accuracy of the auto-focus function of the camera module, the accuracy of the burned optimal DAC value needs to be ensured, so in the production process, the accuracy of the searched optimal DAC value needs to be detected, and in the actual production process, the accuracy of the DAC value is detected by analyzing the resolving power corresponding to the DAC value.

In the prior art, the resolving power of the camera module is analyzed mainly according to three detection means of TVLine (line pair), MTF (modulation transfer function) and SFR (space frequency response);

TVLine is a subjective test, mainly characterized by convenient and visual operation, direct reading by naked eyes as shown in figure 1, but the result is easily influenced by subjective factors of operators, the resolving power cannot be accurately detected, and the actual production is not facilitated;

the MTF is calculated by searching the contrast ratio of the maximum brightness point and the minimum brightness point in the line pair, a black-and-white line pair chart is designed according to 1/4 frequency of a chip, as shown in fig. 2, the MTF value is between 0 and 1, and the higher the MTF value is, the better the resolution is;

SFR is used for measuring the influence on a single image caused by the increase of lines with spatial frequency, and the larger the SFR value is, the sharper the representative image is, and at the moment, the clearer the picture is; as shown in fig. 3, the chart is designed by the SFR test, and the SFR test can convert the MTFs approximately equal to those of all the spatial frequencies without photographing the line pairs of different spatial frequencies, so that the detection time is saved, the detection efficiency is greatly improved, and more users now require using the SFR to detect the camera module.

In summary, in the test of focusing of the camera module, the above detection means for analyzing the resolving power of the camera module is combined, and at present, the automatic focusing of the camera module is mainly performed in the following two ways:

one is to find focus by MTF detection mode and find out the best DAC value by MTF focusing mode; the other is to find focus by MTF detection and find the best DAC value by SFR focusing.

For users, the two modes have significance and value, manufacturers can select different detection modes according to different user demands, but in the second mode, the best DAC value is found through the SFR focusing mode, the analysis force detection is correspondingly carried out on the found best DAC value, the obtained best DAC value is higher in accuracy, and therefore the second mode is used more frequently;

however, in the second mode, the focus is found by using an MTF detection mode, and the optimal DAC value is found by using an SFR focusing mode, so that the standard matching problem of the MTF and the SFR exists, because the SFR can convert the MTF which is approximately equal to the MTF under all spatial frequencies, but the MTF is only approximately equal to the MTF, if the conversion is unreasonable or wrong, the resolution is poor, that is, the resolution detection is inaccurate, the real resolution response condition of the camera module is seriously affected, and the finding of the optimal DAC value is difficult and wrong; and combining the production process of camera module, the producer can be according to user's project demand, first sample of preparation, then carry out the heuristic, just carry out mass production at last, if first sample and heuristic in-process single camera module have not appeared to mark the problem, and in a large amount of camera modules when mass production, partly because have influenced analytic power detection to mark the problem, thereby cause the camera module of mass production not to accord with user's project demand, need test again, also can waste a large amount of test resources, seriously influence production efficiency, when the condition is serious, the user can require to change focusing method even, cause huge economic loss to production.

Disclosure of Invention

The invention aims to provide a camera focusing method which can avoid the problem of SFR and MTF standard alignment, ensure the accuracy of analysis force detection and improve the production efficiency.

The basic scheme provided by the invention is as follows: a camera focusing method comprises the following steps:

coarse detection MTF focus searching: performing coarse inspection MTF focus finding test according to the coarse inspection searching range and coarse inspection stepping, and obtaining an optimal DAC value during coarse inspection through coarse inspection fitting and extremum judgment;

fine detection SFR focus searching step: according to the optimal DAC value during coarse detection and the fine detection stepping, performing fine detection SFR focus searching test, and obtaining the optimal DAC value during fine detection through fine detection fitting and extremum judgment;

SFR focusing test step: and acquiring SFR values of each view field of the image corresponding to the fine-detection optimal DAC value according to the fine-detection optimal DAC value, judging whether the SFR values meet the preset SFR value requirement, if so, taking the fine-detection optimal DAC value as a final optimal DAC value, and outputting.

The first basic scheme has the beneficial effects that: in the scheme, coarse detection MTF and fine detection SFR are adopted to search focus, and then SFR is adopted to focus, so that an accurate optimal DAC value is obtained; specifically, in the scheme, a coarse inspection MTF focus searching step is firstly carried out, a coarse inspection MTF focus searching test is carried out according to a coarse inspection searching range and coarse inspection steps, and an optimal DAC value during coarse inspection is obtained through coarse inspection fitting and extremum judgment; then, performing fine detection SFR focus searching, namely performing fine detection SFR focus searching testing according to the optimal DAC value and fine detection stepping during coarse detection, and obtaining the optimal DAC value during fine detection through fine detection fitting and extremum judgment, namely obtaining the optimal clear position of the camera module under the current object distance;

the method comprises the steps of firstly carrying out rough detection on MTF focus searching before fine detection on SFR focus searching, obtaining an optimal DAC value during rough detection through rough detection fitting and extremum judgment, namely, roughly finding out a clear position, calculating a reasonable and proper fine detection searching range according to the optimal DAC value during rough detection in the fine detection on SFR focus searching so as to solve the problem that the found clear position is wrong due to the limitation of a fitting algorithm, wherein the DAC value collected in the focus searching is used as data, has certain data regularity, and a curve for fitting and obtaining is not necessarily a reasonable parabola and possibly a curve which is firstly level unchanged and then is declined, so that when the optimal DAC value is correspondingly found, the position on a horizontal line cannot be determined to be the optimal DAC value, and the found clear position is wrong; therefore, firstly, a coarse detection MTF focus searching test is carried out, and the optimal DAC value during coarse detection is obtained through coarse detection fitting and extremum judgment, so that a smaller fine detection searching range can be calculated, the regularity of data in the small range is reduced, the DAC value obtained by fine detection SFR focus searching in the range is not formed into a horizontal line, but a parabola, and the accurate optimal DAC value is obtained;

after finishing the rough detection MTF and the fine detection SFR focus searching, carrying out an SFR focusing test step, acquiring SFR values of each view field of the image corresponding to the fine detection optimal DAC value according to the fine detection optimal DAC value, judging whether the SFR values meet the preset SFR value requirement, if so, taking the fine detection optimal DAC value as a final optimal DAC value, and outputting; the SFR focusing test step is to detect the resolving power corresponding to the optimal DAC value during fine detection, so that the accuracy of the final optimal DAC value is ensured; in the SFR focusing test step, the optimal DAC value in fine detection is tested, and the optimal DAC value in fine detection is obtained by fine detection of SFR focusing, so that conversion standard alignment is not existed, the problem of standard alignment of SFR and MTF can be avoided, and the accuracy of analysis force detection is ensured. After the accuracy of the analysis force detection is guaranteed, the analysis force detection can be influenced by the standard problem during the reduction of production, so that the mass-produced camera module does not meet the project requirements of users and needs to be tested again, a large amount of test resources are saved, the production efficiency is improved, and huge economic loss is avoided.

In summary, the method can avoid the problem of the standard alignment of SFR and MTF, solve the problem of error in the found clear position caused by the limitation of the fitting algorithm, and ensure the accuracy of the analysis force detection so as to improve the production efficiency.

Further, the method further comprises the following steps: parameter setting: setting automatic focusing clamping control parameters;

an object distance adjusting step: and adjusting the object distance between the camera module and the chart.

The beneficial effects are that: the camera module is designed and manufactured according to the project requirements of users, focusing is one step in the production process, so that automatic focusing clamping control parameters are required to be set in advance, and follow-up focusing is guaranteed to meet the project requirements of the users.

Further, the auto-focus clamping control parameter includes: coarse search range [ RoughValue_L, roughValue_H]Coarse detection step Rough_step, coarse detection fitting-extremum Threshold Roughcheck_threshold, fine detection step Fine_step, and optimal DAC value clamping range [ RoughValue_L+m_range, roughValue_H-m_range ]]Center position of each field frame (ROI_X) _i ,ROI_Y _i ) And size (ROI_Width) _i ,ROI_Height _i ) MTF threshold center mtf_threshold for center field of view, fine vertical-horizontal SFR threshold Fine HV threshold, and individual field of view SFR threshold _i 。

The beneficial effects are that: the automatic focusing clamping control parameters are specifically set according to the project requirements of users and used for each focus searching and focusing step so as to finish focusing detection.

Further, the coarse detection MTF focus searching step comprises the following steps: s301, searching the range [ RoughValue_L, roughValue_H ] according to the rough detection]The DAC value DACVALue_Rough of the motor pushing position is sequentially set by Rough detection step Rough_step accumulation _m And at the motor according to DACVALue_Rough _m Acquiring an image raw of a chart when in a pushing position _m The method comprises the steps of carrying out a first treatment on the surface of the Wherein DACVALue_Rough _m ＝RoughValue_L+m*Rough_step；

M is the accumulated number in the rough detection;

and m is a positive integer which is a positive integer,

s302, according to raw _m Calculating the MTF of a field of view frame of a center _m A value;

s303, pushing the position DAC value DACVALue_Rough according to the motor _m Corresponding MTF _m Curve fitting is carried out to obtain the optimal DAC value DACVALue_best during rough detection _rough ；

S304, pushing the position DAC value DACVALue_Rough according to the motor _m Corresponding MTF _m Obtaining MTF _m Maximum MTF of (2) _max ＝max[MTF _m ]Corresponding motor push position DAC value DACVALue_Rough _m1 Wherein m1 ε m, and MTF _max ≥CenterMTF_threshold；

S305, judging DACVvalue_best _rough And DACVALue_Rough _m1 If the difference value of the two values belongs to the rough detection fitting-extremum Threshold range Roughcheck_threshold, executing a fine detection SFR focus searching step; wherein the difference is abs|DACVvalue_best _rough -DACValue_Rough _m1 |。

The beneficial effects are that: for the obtained DACVvalue_best _rough And DACVALue_Rough _m1 The difference value of (2) is judged to prevent the obtained optimal DAC value DACVvalue_best during coarse detection _rough Abnormality affects the progress of the subsequent steps.

Further, the step S302 includes: sequentially at each raw according to the set view field frame area _m Intercepting a center block image; converting the central block image into a gray scale image to form a central gray scale image; calculating the average pixel value of each center gray scale image; classifying pixels according to whether each pixel of the center gray map is higher than an average pixel value, and calculating an average value A of total pixels of the pixels higher than the average pixel value _m Average value B of total pixels of pixels lower than or equal to the average pixel value _m The method comprises the steps of carrying out a first treatment on the surface of the According to

Calculation of MTF _m Values.

The beneficial effects are that:

the pixels of each center block image can be effectively reflected.

Further, the fine detection SFR focus searching step comprises the following steps:

s401, according to the optimal DAC value DACVALue_best in coarse detection _rough Obtaining fine search range [ FineValue_L, fineValue_H]：

S402, according to the fine detection search range [ FineValue_L, fineValue_H]Sequentially setting the DAC value DACVALue_Fine of the motor push position by accumulating Fine-detection stepping Fine_step _n And at the motor according to DACVvalue_Fine _n Acquiring an image raw of a chart when in a pushing position _n The method comprises the steps of carrying out a first treatment on the surface of the Wherein, DACVvalue_Fine _n ＝FineValue_L+n*Fine_step；

N in the above formula is the cumulative number in fine detection;

and n is a positive integer which is a positive integer,

s403, according to raw _n Acquiring a vertical knife edge and a horizontal knife edge closest to the center position by adopting an image recognition algorithm, and acquiring a SFR value DACVALue_FineV in the vertical direction according to the vertical knife edge and the horizontal knife edge _n And SFR value dacvalue_fineh in horizontal direction _n ；

S404, according to SFR value DACVALue_FineV in vertical direction _n And SFR value dacvalue_fineh in horizontal direction _n Respectively performing curve fitting to obtain the optimal DAC value DACVALue_bestV in the vertical direction during fine detection _fine And DACVALue_bestH _fine ；

S405, judging DACVALue_FineV _n And DACVALue_FineH _n If the difference value of (b) belongs to the Fine detection vertical-horizontal SFR threshold value fine_HV_threshold, executing S406 if yes;

wherein the difference is abs|DACVALue_bestV _fine -DACValue_bestH _fine |；

S406, according to DACVALue_FineV _n And DACVALue_FineH _n Obtaining an optimal DAC value DACVvalue_best during fine detection; wherein DACVALue_best is DACVALue_FineV _n And DACVALue_FineH _n Average value of (2);

s407: judging whether DACVALue_best belongs to an optimal DAC value clamping range [ RoughValue_L+m_range ] or not;

if yes, i.e. DACVALue_best E [ RoughValue_L+m_range ], roughValue_H-m_range ], then the SFR focusing test step is executed.

The beneficial effects are that: for the obtained DACVvalue_FineV _n And DACVALue_FineH _n Is judged to prevent DACVALue_FineV _n And DACVALue_FineH _n The excessive difference causes inaccurate acquisition of the optimal DAC value DACVvalue_best during fine detection; and whether the DACVvalue_best belongs to the optimal DAC value clamping range is judged, so that the DAC value output subsequently is prevented from not meeting the project requirement of a user.

Further, the S403 includes:

sequentially at each raw according to the set view field frame area _n Intercepting a center block image;

performing Gaussian filtering and gray level conversion on the center block image to obtain a center block gray level image;

acquiring each angular point position of a gray level map of a center block through an angular point detection method of opencv and a sub-pixel level angular point detection method, and extracting four angular point positions closest to the center position;

according to the four corner positions closest to the central position, a knife edge in the vertical direction closest to the central position and a knife edge in the horizontal direction are obtained;

according to the knife edge in the vertical direction and the knife edge in the horizontal direction, calculating a SFR value DACVALue_FineV in the vertical direction _n And in the horizontal directionSFR value DACVALue_FineH _n 。

The beneficial effects are that: and carrying out Gaussian filtering on the central block image to reduce interference of the interference diagonal point position confirmation.

Further, the SFR focusing test step includes:

s501, obtaining an image raw corresponding to the optimal DAC value DACVALue_best in fine detection _n Calculating an image raw _n SFR values for each field of view; wherein the SFR values for each field of view comprise: SFR values in the vertical direction and SFR values in the horizontal direction of each view field frame;

s502, judging whether SFR values of all the fields are larger than SFR thresholds of all the fields, if so, taking the optimal DAC value during fine detection as a final optimal DAC value, and outputting.

The beneficial effects are that: and judging whether SFR values of all the fields are larger than SFR thresholds of all the fields, so as to analyze whether the resolving power corresponding to the optimal DAC value DACVvalue_best reaches the standard.

Further, the method further comprises the following steps: ending the step;

the ending step: ending the test;

the steps S305, S405, S407, and S502 each further include:

if not, executing the ending step.

The beneficial effects are that: the steps S305, S405, S407, and S502 each further include: if not, executing the ending step, wherein the steps are judging steps, judging whether the currently acquired result is reasonable or not, and if not, executing the ending step, thereby reducing the occupied memory in the actual application process.

The second objective of the present invention is to provide a camera focusing system, which can avoid the problem of SFR and MTF calibration, and ensure the accuracy of analysis force detection, so as to improve the production efficiency.

The invention provides a basic scheme II: a camera focusing system adopts the camera focusing method, which comprises the following steps: the system comprises a target, a light source and an image acquisition and processing subsystem;

the chart is arranged on the target;

the light source is used for providing preset illumination and preset color temperature for the target;

an image acquisition processing subsystem comprising: the camera shooting automatic focusing module and the upper computer;

the upper computer is electrically connected with the camera shooting automatic focusing module;

the camera automatic focusing module is used for automatically focusing the camera module and the chart, collecting image information and sending the image information to the upper computer;

and the upper computer is used for regulating and controlling the focusing of the camera module according to the automatic focusing clamping and controlling parameters, and processing and analyzing the image information to obtain a final optimal DAC value.

The second basic scheme has the beneficial effects that: the system executes the camera focusing method, adopts coarse detection MTF and fine detection SFR to search focus, and focuses through the SFR so as to obtain an accurate optimal DAC value; specifically, in the scheme, coarse inspection MTF focus searching is performed firstly, coarse inspection MTF focus searching test is performed according to a coarse inspection searching range and coarse inspection stepping, and the optimal DAC value during coarse inspection is obtained through coarse inspection fitting and extremum judgment; then fine detection SFR focus searching is carried out, fine detection SFR focus searching test is carried out according to the optimal DAC value and fine detection stepping in the coarse detection, and the optimal DAC value in the fine detection, namely the optimal clear position of the camera module under the current object distance, is obtained through fine detection fitting and extremum judgment;

after finishing the focus searching of the coarse detection MTF and the fine detection SFR, carrying out SFR focusing test, acquiring SFR values of each view field of the image corresponding to the fine detection optimal DAC value according to the fine detection optimal DAC value, judging whether the SFR values meet the preset SFR value requirement, if so, taking the fine detection optimal DAC value as a final optimal DAC value, and outputting; the SFR focusing test is to detect the resolving power corresponding to the optimal DAC value during fine detection, so that the accuracy of the final optimal DAC value is ensured; in the SFR focusing test, the optimal DAC value in fine detection is tested, and the optimal DAC value in fine detection is obtained by fine detection of SFR focus searching, and conversion calibration is not existed, so that the calibration problem of SFR and MTF can be avoided, and the accuracy of analysis force detection is ensured. After the accuracy of the analysis force detection is guaranteed, the analysis force detection can be influenced by the standard problem during the reduction of production, so that the mass-produced camera module does not meet the project requirements of users and needs to be tested again, a large amount of test resources are saved, the production efficiency is improved, and huge economic loss is avoided.

Drawings

FIG. 1 is a TVLine observation;

FIG. 2 is a black and white line versus chart;

FIG. 3 is a chart;

fig. 4 is a flowchart illustrating an embodiment of a camera focusing method according to the present invention;

fig. 5 is a schematic diagram of a corner position in an embodiment of a camera focusing method according to the present invention;

FIG. 6 is a schematic diagram of a knife edge in an embodiment of a camera focusing method according to the present invention;

fig. 7 is a schematic view of a field frame in an embodiment of a camera focusing method according to the present invention;

FIG. 8 is a graph of the result of curve fitting in the rough detection MTF focus finding step in an embodiment of a camera focusing method according to the present invention;

FIG. 9 is a graph of the result of curve fitting in the fine detection SFR focus finding step in an embodiment of a camera focusing method according to the present invention;

fig. 10 is a logic block diagram of an embodiment of a camera focusing system according to the present invention.

Detailed Description

The following is a further detailed description of the embodiments:

example 1

This embodiment is substantially as shown in fig. 4: a camera focusing method comprises the following steps:

parameter setting: setting automatic focusing clamping control parameters; wherein the auto-focus clamping control parameters comprise: coarse search range [ RoughValue_L, roughValue_H]Coarse detection step Rough_step, coarse detection fitting-extremum Threshold Roughcheck_threshold, fine detection step Fine_step, and optimal DAC value clamping range [ RoughValue_L+m_range, roughValue_H-m_range ]]Center position of each field frame (ROI_X) _i ,ROI_Y _i ) And size (ROI_Width) _i ,ROI_Height _i ) MTF threshold center mtf_threshold for center field of view, fine vertical-horizontal SFR threshold Fine HV threshold, and individual field of view SFR threshold _i ；

Wherein roughvalue_l is the lower limit of the coarse search range, and roughvalue_h is the upper limit of the coarse search range;

coarse detection fitting-extremum Threshold roughlock_threshold is the weight of coarse detection step fine_step;

i is the number of field frames to be tested of the image, the optimal DAC value clamping range is generally included in the coarse detection search range and is not empty, namely m_range is more than or equal to 0, and RoughValue_L+m_range is less than RoughValue_H-m_range.

An object distance adjusting step: adjusting the object distance between the camera module and the chart;

specifically, according to project requirements, the object distance between the camera module set formulated by the optical characteristics of the lens and the chart, namely the actual distance between the camera module set and the chart, is adjusted, and in order to influence the distance by the adjustable distance of the machine equipment in actual production of the surface, the actual object distance is simulated through the distance increasing mirror in the embodiment.

specifically, the coarse inspection MTF focus searching step comprises the following steps:

s301, searching the range [ RoughValue_L, roughValue_H ] according to the rough detection]The DAC value DACVALue_Rough of the motor pushing position is sequentially set by Rough detection step Rough_step accumulation _m And at the motor according to DACVALue_Rough _m Acquiring an image raw of a chart when in a pushing position _m ；

Wherein DACVALue_Rough _m ＝RoughValue_L+m*Rough_step；

M is the accumulated number in the rough detection;

and m is a positive integer which is a positive integer,

s302, according to raw _m Calculating the MTF of a field of view frame of a center _m A value; the method comprises the following steps: sequentially at each raw according to the set view field frame area _m A center block image is taken, wherein the center block image is an image of a field frame at the center, and the region of the field frame is set according to the center position (ROI_X of each field frame _i ,ROI_Y _i ) And size (ROI_Width) _i ,ROI_Height _i ) Determining, namely intercepting a center block image according to a set center view field frame area; converting the central block image into a gray scale image to form a central gray scale image; calculating the average pixel value of each center gray scale image; classifying pixels according to whether each pixel of the center gray map is higher than an average pixel value, and calculating an average value A of total pixels of the pixels higher than the average pixel value _m Average value B of total pixels of pixels lower than or equal to the average pixel value _m The method comprises the steps of carrying out a first treatment on the surface of the According to

Calculation of MTF _m A value;

s303, pushing the position DAC value DACVALue_Rough according to the motor _m Corresponding MTF _m Curve fitting is carried out to obtain the optimal DAC value DACVALue_best during rough detection _rough The method comprises the steps of carrying out a first treatment on the surface of the The method comprises the following steps: according to DACVALue_Rough _m With MTF _m Obtaining a curve fitting function, and calculating a maximum value DACVvalue_best of the curve fitting function _rough In which DACVvalue_best _rough ∈[RoughValue_L,RoughValue_H]Which is the best DAC value for coarse detection;

S305, judging DACVvalue_best _rough And DACVALue_Rough _m1 If the difference value of the two values belongs to the rough detection fitting-extremum Threshold range Roughcheck_threshold, executing a fine detection SFR focus searching step; if not, executing the ending step; wherein the difference is abs|DACVvalue_best _rough -DACValue_Rough _m1 |；

If abs|DACVvalue_best _rough -DACValue_Rough _m1 The difference value is within the Roughcheck_threshold of the rough detection fitting-extremum Threshold range, namely that the obtained optimal DAC value meets the preset requirement in rough detection after the rough detection MTF focus searching step, and the camera module set according to the user requirement can meet the requirement of the user on the camera module in rough detection MTF focus searching.

specifically, the fine detection SFR focus searching step comprises the following steps:

S402, according to the fine detection search range [ FineValue_L, fineValue_H]Sequentially setting the DAC value DACVALue_Fine of the motor push position by accumulating Fine-detection stepping Fine_step _n And at the motor according to DACVvalue_Fine _n Acquiring an image raw of a chart when in a pushing position _n ；

Wherein, DACVvalue_Fine _n ＝FineValue_L+n*Fine_step；

N in the above formula is the cumulative number in fine detection;

and n is a positive integer which is a positive integer,

s403, according to raw _n Acquiring a vertical knife edge and a horizontal knife edge closest to the center position by adopting an image recognition algorithm, and acquiring a SFR value DACVALue_FineV in the vertical direction according to the vertical knife edge and the horizontal knife edge _n And SFR value dacvalue_fineh in horizontal direction _n The method comprises the steps of carrying out a first treatment on the surface of the Wherein the knife edge is a black-white bevel edge; the method comprises the following steps:

obtaining each angular point position of a gray level diagram of a center block through an angular point detection method of opencv and a sub-pixel level angular point detection method, and extracting four angular point positions closest to the center position as shown in fig. 5; wherein the center position is according to (ROI_X _i ,ROI_Y _i ) Determining;

according to the four corner positions closest to the central position, a knife edge in the vertical direction closest to the central position and a knife edge in the horizontal direction are obtained, as shown in fig. 6;

according to the knife edge in the vertical direction and the knife edge in the horizontal direction, calculating a SFR value DACVALue_FineV in the vertical direction _n And SFR value dacvalue_fineh in horizontal direction _n The method comprises the steps of carrying out a first treatment on the surface of the The prior art is adopted for calculating SFR values, and supersampling is carried out on a knife edge in the vertical direction and a knife edge in the horizontal direction to obtain ESF (Edge Spread Function, edge expansion function), namely black-white conversion straight lines; thereby obtaining finer black-white conversion straight lines; deriving ESF to obtain LSF (Line Spread Function, linear expansion function), namely changing rate of black-white conversion straight line; performing Fourier transform on the LSF to obtain SFR values of all the spatial frequencies; and setting an abscissa according to the preset spatial frequency, and obtaining a SFR value of the corresponding ordinate through an interpolation algorithm.

In addition, if the obtained knife edge is not clear in the focus searching process, the SFR value is 0.

S405, judging DACVALue_FineV _n And DACVALue_FineH _n If the difference value of (b) belongs to the Fine detection vertical-horizontal SFR threshold value fine_HV_threshold, executing S406 if yes; if not, executing the ending step; wherein the difference is abs|DACVALue_bestV _fine -DACValue_bestH _fine |；

If abs|DACVALue_bestV _fine -DACValue_bestH _fine And if the I is less than or equal to the fine_HV_threshold, the difference value belongs to the Fine detection vertical-horizontal SFR threshold value fine_HV_threshold, namely, the vertical and horizontal SFR (analysis force) deviation meets the preset requirement, otherwise, the deviation is overlarge.

if yes, namely DACVALue_best E [ RoughValue_L+m_range ], executing an SFR focusing test step; if not, executing the ending step.

SFR focusing test step: acquiring SFR values of each view field of the image corresponding to the fine-detection optimal DAC value according to the fine-detection optimal DAC value, judging whether the SFR values meet the preset SFR value requirement, if so, taking the fine-detection optimal DAC value as a final optimal DAC value, and outputting the final optimal DAC value;

the method comprises the following steps:

s501, obtaining an image raw corresponding to the optimal DAC value DACVALue_best in fine detection _n Calculating an image raw _n SFR values for each field of view; wherein the SFR values for each field of view comprise: the specific calculation method of the SFR value in the vertical direction and the SFR value in the horizontal direction of each view field frame adopts the method described in S403, and will not be described in detail here;

s502, judging whether SFR values of all the fields are larger than SFR thresholds of all the fields, if so, taking the optimal DAC value during fine detection as a final optimal DAC value, and outputting the final optimal DAC value; if not, executing the ending step.

Ending the steps: ending the test.

The specific implementation process is as follows: taking a camera module of a certain 12Meg chip as an example, the resolution is 4000 x 3000, and automatic focusing is carried out when the object distance is 3 m; coarse search range [ RoughValue_L, roughValue_H]＝[600,960]Coarse detection step rough_step=40, coarse detection fitting-extremum Threshold roughheck_threshold=1.5×rough_step=60, fine detection step fine_step=8, optimum DAC value clamping range [ roughvalue_l+m_range, roughvalue_h-m_range]＝[600+30,960-30]＝[630,930]13 fields of view were tested, the center position of each field of view frame (ROI_X _i ,ROI_Y _i ) And size (ROI_Width) _i ,ROI_Height _i ) As shown in fig. 7;

MTF threshold center mtf_threshold=0.667 for center field of view, fine vertical-horizontal SFR threshold Fine hv_threshold=20, and each field of view SFR threshold sfr_threshold shown in table 1 below _i ：

Table 1: data relating to each field of view

Performing coarse detection MTF focus searching step to obtain DACValue_Rough _m And its MTF _m Values, as shown in table 2 below; curve fitting is performed, and as shown in fig. 8, the optimum DAC value dacvalue_best at the time of rough detection is obtained _rough =825, obtain MTF _m Maximum MTF of (2) _max ＝max[MTF _m ]Corresponding motor push position DAC value DACVALue_Rough _m1 =840, determine dacvalue_best _rough And DACVALue_Rough _m1 If the difference value of the two values belongs to the rough detection fitting-extremum Threshold range Roughcheck_threshold, executing a fine detection SFR focus searching step;

table 2: DACVALue_Rough _m And MTF _m Value of

DAC value	MTF value
			600	0.734746667
640	0.748166667
		680	0.758626667
720	0.7715
		760	0.780326667
800	0.788073333
		840	0.790093333
880	0.785573333
		920	0.77848
960	0.768313333

Performing fine detection SFR focus searching step according to the optimal DAC value DACVALue_best during coarse detection _rough Obtaining fine search range [ FineValue_L, fineValue_H]＝[825-40,825+40]＝[785,865]The method comprises the steps of carrying out a first treatment on the surface of the Further obtain DACVvalue_Fine _n SFR value DACVALue_FineV in vertical direction _n (SFR_V value) and SFR value DACVALue_Fineh in horizontal direction _n (SFR_H value) as shown in Table 3 below; curve fitting is performed, and as shown in fig. 9, the vertical direction optimum DAC value dacvalue_bestv at the time of fine inspection is obtained _fine =832 and dacvalue_besth _fine =838; judging DACVvalue_FineV _n And DACVALue_FineH _n Whether the difference value of (2) belongs to the Fine-detection vertical-horizontal SFR threshold value Fine_HV_threshold, if so, according to DACVvalue_FineV _n And DACVALue_FineH _n Obtaining the optimal DAC value DACVvalue_best=835 in fine detection; judging whether DACVALue_best belongs to the optimal DAC value clamping range [ RoughValue_L+m_range, roughValue_H-m_range]＝[630,930]If so, the method comprises, if so,then the SFR focus test step is performed;

table 3: DACVALue_Fine _n SFR value DACVALue_FineV in vertical direction _n And SFR value dacvalue_fineh in horizontal direction _n

DAC value	SFR value in vertical direction	SFR value in horizontal direction
			795	0	0.1
803	0.41	0.36
			811	0.421	0.44
819	0.567	0.56
			827	0.6	0.64
835	0.65	0.65
			843	0.73	0.74
851	0.68	0.678
			859	0.615	0.598
867	0.512	0.542
			875	0.21	0.32

SFR focusing test is carried out, and an image raw corresponding to the optimal DAC value DACVvalue_best in fine detection is obtained _n Calculating an image raw _n SFR values for each field of view; wherein the SFR values for each field of view comprise: SFR values in the vertical direction and SFR values in the horizontal direction of each view field frame; and judging whether the SFR values of the fields are larger than the SFR threshold of the fields, if so, taking the optimal DAC value during fine detection as the final optimal DAC value, and outputting DACVALue_best=835.

Example two

This embodiment is substantially as shown in fig. 10: a camera focusing system adopts the camera focusing method;

comprising the following steps: the system comprises a target, a light source and an image acquisition and processing subsystem;

the chart is arranged on the target plate, and is a grid chart formed by alternately spacing black-white oblique blocks, the grid has a certain gradient, the gradient of the grid is 8 degrees in the embodiment, and the side length of the grid is obtained according to the pixel size, the chip resolution, the FOV (field of view) angle, the test object distance and the EFL effective focal length in a conversion mode;

the light source is used for providing preset illumination and preset color temperature for the target; specifically, the light source adopts a light source plate, the light source plate is placed behind or in front of the target plate, and the preset illuminance and the preset color temperature are adjusted, wherein the illuminance is 650+/-50 Lux, and the color temperature is 6000+/-1000K in the embodiment;

the upper computer is electrically connected with the camera shooting automatic focusing module, and the upper computer is connected with the camera shooting automatic focusing module through a data connecting wire in the embodiment;

the camera shooting automatic focusing module is used for automatically focusing with the chart, collecting image information and sending the image information to the upper computer;

the upper computer is used for processing and analyzing the image information to obtain a final optimal DAC value; in this embodiment, the upper computer is disposed in the computer; the parameter setting step is to set automatic focusing clamping control parameters in an upper computer;

an automatic focusing module for photographing, comprising: the device comprises a camera module, a jig and an acquisition test box;

the jig is used for installing the camera module;

the camera module is a camera module which needs to be tested and is used for collecting image information; when the actual object distance between the camera module and the chart is adjusted according to project requirements, if the actual object distance cannot be reached by the environment of the actual machine where the automatic focusing module is located, setting a distance increasing mirror between the chart and the camera module so as to simulate the actual object distance;

and the acquisition test box is connected with the camera module and used for uploading the image information to the upper computer.

The foregoing is merely an embodiment of the present invention, and a specific structure and characteristics of common knowledge in the art, which are well known in the scheme, are not described herein, so that a person of ordinary skill in the art knows all the prior art in the application day or before the priority date of the present invention, and can know all the prior art in the field, and have the capability of applying the conventional experimental means before the date, so that a person of ordinary skill in the art can complete and implement the present embodiment in combination with his own capability in the light of the present application, and some typical known structures or known methods should not be an obstacle for a person of ordinary skill in the art to implement the present application. It should be noted that modifications and improvements can be made by those skilled in the art without departing from the structure of the present invention, and these should also be considered as the scope of the present invention, which does not affect the effect of the implementation of the present invention and the utility of the patent. The protection scope of the present application shall be subject to the content of the claims, and the description of the specific embodiments and the like in the specification can be used for explaining the content of the claims.

Claims

1. A camera focusing method is characterized in that: the method comprises the following steps:

2. The camera focusing method according to claim 1, characterized in that: further comprises: parameter setting: setting automatic focusing clamping control parameters;

3. The camera focusing method according to claim 2, characterized in that: the automatic focusing clamping control parameter comprises: coarse search range [ RoughV ]alue_L,RoughValue_H]Coarse detection step Rough_step, coarse detection fitting-extremum Threshold Roughcheck_threshold, fine detection step Fine_step, and optimal DAC value clamping range [ RoughValue_L+m_range, roughValue_H-m_range ]]Center position of each field frame (ROI_X) _i ,ROI_Y _i ) And size (ROI_Width) _i ,ROI_Height _i ) MTF threshold center mtf_threshold for center field of view, fine vertical-horizontal SFR threshold Fine HV threshold, and individual field of view SFR threshold _i 。

4. A camera focusing method according to claim 3, characterized in that: the coarse detection MTF focus searching step comprises the following steps: s301, searching the range [ RoughValue_L, roughValue_H ] according to the rough detection]The DAC value DACVALue_Rough of the motor pushing position is sequentially set by Rough detection step Rough_step accumulation _m And at the motor according to DACVALue_Rough _m Acquiring an image raw of a chart when in a pushing position _m The method comprises the steps of carrying out a first treatment on the surface of the Wherein DACVALue_Rough _m ＝RoughValue_L+m*Rough_step；

M is the accumulated number in the rough detection;

and m is a positive integer which is a positive integer,

S305, judging DACVvalue_best _rough And DACVALue_Rough _m1 Whether the difference of (2) is within the range of the rough detection fit-extremum thresholdSurrounding the Roughcheck_Threshold, if yes, executing a fine detection SFR focus searching step; wherein the difference is abs|DACVvalue_best _rough -DACValue_Rough _m1 |。

5. The camera focusing method according to claim 4, characterized in that: the step S302 includes: sequentially at each raw according to the set view field frame area _m A center block image is intercepted, wherein the center block image is an image of a field frame at the center; converting the central block image into a gray scale image to form a central gray scale image; calculating the average pixel value of each center gray scale image; classifying pixels according to whether each pixel of the center gray map is higher than an average pixel value, and calculating an average value A of total pixels of the pixels higher than the average pixel value _m Average value B of total pixels of pixels lower than or equal to the average pixel value _m The method comprises the steps of carrying out a first treatment on the surface of the According to

Calculation of MTF _m Values.

6. The camera focusing method according to claim 5, characterized in that: the fine detection SFR focus searching step comprises the following steps:

s401, according to the optimal DAC value DACVALue_best in coarse detection _rough Obtaining a fine detection search range

N in the above formula is the cumulative number in fine detection;

and n is a positive integer which is a positive integer,

wherein the difference is abs|DACVALue_bestV _fine -DACValue_bestH _fine |；

7. The camera focusing method according to claim 6, characterized in that: the S403 includes:

according to the knife edge in the vertical direction and the knife edge in the horizontal direction, calculating a SFR value DACVALue_FineV in the vertical direction _n And SFR value dacvalue_fineh in horizontal direction _n 。

8. The camera focusing method according to claim 7, characterized in that: the SFR focusing test step comprises the following steps:

9. The camera focusing method according to claim 8, characterized in that: further comprises: ending the step;

the ending step: ending the test;

the steps S305, S405, S407, and S502 each further include:

if not, executing the ending step.

10. A camera focusing system, characterized in that: a camera focusing method according to any one of claims 1 to 9, comprising: the system comprises a target, a light source and an image acquisition and processing subsystem;

the chart is arranged on the target;