CN105791668A - Automatic focusing device - Google Patents
Automatic focusing device Download PDFInfo
- Publication number
- CN105791668A CN105791668A CN201410827938.0A CN201410827938A CN105791668A CN 105791668 A CN105791668 A CN 105791668A CN 201410827938 A CN201410827938 A CN 201410827938A CN 105791668 A CN105791668 A CN 105791668A
- Authority
- CN
- China
- Prior art keywords
- image
- focusing
- lens
- spatial domain
- pixel
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Landscapes
- Studio Devices (AREA)
- Automatic Focus Adjustment (AREA)
Abstract
The invention relates to an automatic focusing device. The automatic focusing device comprises a lens, a lens driving part, a noise reduction module and a focusing control module; on the basis of the lens, an original image can be generated through light guide of a shot object; the noise reduction module is suitable for obtaining a frequency response result of a pixel point of the original image and outputting a noise reduction image on the basis of the frequency response result of a passband part; the focusing control part is suitable for extracting a spatial domain edge operator in an interested image position of the noise reduction image, and correspondingly multiplying and summarizing the spatial domain edge operator and a pixel value in the interested image position according to the edge direction, such that a spatial domain result is obtained; and the lens driving part is suitable for driving the lens so as to obtain the expected spatial domain result, and driving the lens by taking the position of the lens when the expected spatial domain result is achieved as the focusing position. By means of the invention, the focusing accuracy and the focusing speed of the shot object can be increased under conditions of low contrast texture and weak light.
Description
Technical field
The present invention relates to imaging field, particularly to a kind of automatic focusing mechanism.
Background technology
Picture pick-up device, the such as mobile phone of camera and integral photographic function, usually implement auto-focusing.
Auto-focusing (AutoFocus) make use of the principle of subject luminous reflectance, the light of subject reflection is accepted by the imageing sensor (such as CCD/CMOS sensor) of the picture pick-up devices such as camera, by computer disposal, driving electric focusing mechanism is focused.
Complete automatic focusing mechanism comprises the steps that at least one imaging len, at least one focusing lens, a sensor devices, a focal length setting mechanism.
From ultimate principle, Atomatic focusing method is segmented into two big classes: a class is based on camera lens and the range finding auto-focusing being taken between target range measurement, the another kind of focus detection auto-focusing being based on focusing screen imaging clearly, namely active focusing method and passive type focusing method.
The focusing mode of active focusing method is as follows:
The infrared generator (or supersonic generator) having on picture pick-up device, infrared generator (or supersonic generator) sends infrared light (or ultrasound wave) to subject；
The infrared light (or ultrasound wave) that accepter on picture pick-up device accepts to reflect is focused, the similar range of triangle focusing method of its optical principle, can be used for the auto-focusing of low-grade popular camera (such as various look squarely camera of finding a view).
The shortcoming of active focusing method is as follows:
Owing to the infrared light sent or ultrasound wave may be reflected to other direction, or not reaching subject, the light wave that on picture pick-up device, receptor accepts is limited, particularly to inclined-plane or shiny surface, focus effects is poor, the problem that focusing difficulty also can occur greatly and apart from remote subject for brightness；Active focusing method is when picture pick-up device actively sends light wave, there is likely to be the low contrast of light wave, cause the problem of focusing under decreased light, when the subject of hachure and the subject of motion are focused it also occur that the situation of physical ability absorbing light ripple being shot, additionally, subject by glass-reflected, thus can be carried out focusing through glass and difficulty can occur by light wave when through glass.
The focusing mode of passive type focusing method is as follows:
Directly receive and analyze reflective from subject self, thus auto-focusing.
Passive type Atomatic focusing method compared to the advantage of active auto-focusing is: self do not want emission system, thus power consumption is few, is conducive to miniaturization, is applicable to the small hand held devices such as mobile phone；To the physical ability being shot with certain brightness ideal carry out auto-focusing, focusing that also can be good under backlight, the object energy auto-focusing that distant place brightness is big, also can pass through glass focusing, and above-mentioned focus effects is comparatively good.
Passive type Atomatic focusing method has himself shortcoming:
To hachure and only there is the subject of weak texture cannot be carried out good focusing, it is easy to cause focusing mistake and focusing time long；Under low contrast and low light level situation, focusing performance is also unable to reach well；Additionally, to the subject of motion, subject containing polarisation, black subject or the subject with minute surface, all cannot accurately focus.
The focusing system of prior art and focusing method all cannot obtain preferably focus effects, thus affecting the imaging performance of picture pick-up device.
Summary of the invention
Technical solution of the present invention solved the technical problem that into, under low contrast texture and low light level situation, how to improve the focusing accuracy of subject.
In order to solve above-mentioned technical problem, technical solution of the present invention provides a kind of automatic focusing mechanism, including: lens, lens driver part, noise reduction module and focusing control module, based on described lens, the photoconduction of subject can be produced original image；
Described noise reduction module is suitable to obtain the frequency response results of original image pixels point, and the frequency response results based on passband portion exports noise-reduced image；
Described focusing controls module and is suitable to extract the spatial domain boundary operator on noise-reduced image image of interest position, is multiplied according to edge direction corresponding with the pixel value on described image of interest position for described spatial domain boundary operator and sue for peace, to obtain spatial domain result；Described lens driver part is suitable to drive described lens in the hope of expected spatial domain result, and lens position when being up to described expected spatial domain result drives described lens as focusing position.
Optionally, described noise reduction module includes filter unit, and described filter unit is suitable to the view data to original image pixels point and processes to obtain the frequency response values of pixel view data according to following filter function expression formula:
Wherein, m, n are the exponent number of described filter unit, a1To an+1、b1To bm+1For the adjustable parameter of described filter unit, described adjustable parameter is relevant with history original image；
Described frequency response results is the set of described frequency response values.
Optionally, described filter unit is four-step filter.
Optionally, described adjustable parameter is determined based on the frequency response curve of some history original images, and described history original image at least includes a width correct focusing position image and an incorrect focusing position image.
Optionally, the frequency response curve based on each history original image determines described filter unit band connection frequency, thus generating described adjustable parameter.
Optionally, described filter unit is low pass filter or band filter.
Optionally, described spatial domain boundary operator is Laplace operator or Sobel Operator, and described edge direction includes vertical direction and horizontal direction；
It is multiplied corresponding with the pixel value on described image of interest position for described spatial domain boundary operator and sue for peace in the vertical direction, obtaining vertical direction result；
It is multiplied corresponding with the pixel value on described image of interest position for described spatial domain boundary operator and sue for peace in the horizontal direction, obtaining horizontal direction result；Described spatial domain result is the absolute value absolute value sum with horizontal direction result of vertical direction result.
Optionally, described image of interest position is made up of the pixel around interested pixel point and this interested pixel point.
Optionally, described device also includes: focusing position search module: described focusing position search module is suitable to drive at described lens driver part constantly obtain described spatial domain result in the process of described lens, to generate the corresponding relation of lens position and described spatial domain result, and identify described expected spatial domain result based on described corresponding relation.
Optionally, described focusing position search module is integrated in described focusing control module.
Optionally, described device also includes: image sensing module；
Described image sensing module is suitable to reception, from lens, the photoconduction of subject is produced picture signal, and described original image is generated based on described picture signal.
Optionally, described image sensing module is made up of the C-MOS sensor as imaging apparatus.
In order to solve above-mentioned technical problem, technical solution of the present invention additionally provides a kind of automatic focusing mechanism, including: lens, lens driver part, noise reduction module and focusing control module, based on described lens, the photoconduction of subject can be produced original image；
Described noise reduction module is suitable to obtain the frequency response results on original image image of interest position, and the frequency response results based on passband portion exports noise-reduced image；
Described focusing controls module and is suitable to extract the spatial domain boundary operator of noise-reduced image, is multiplied according to edge direction corresponding with the pixel value of described noise-reduced image for described spatial domain boundary operator and sue for peace, to obtain spatial domain result；Described lens driver part is suitable to drive described lens in the hope of expected spatial domain result, and lens position when being up to described expected spatial domain result drives described lens as focusing position.
Optionally, described image of interest position is made up of the pixel around interested pixel point and this interested pixel point, is (1+2t) for pixel number2Pixel region, t is the natural number more than or equal to 1, and wherein, the pixel of t+1 row in this pixel region, t+1 row is described interested pixel point, and rest of pixels point is described pixel around.
Optionally, described noise reduction module includes filter unit, and described filter unit is suitable to the view data to image of interest position pixel and processes to obtain the frequency response values of pixel view data according to following filter function expression formula:
Wherein, m, n are the exponent number of described filter unit, a1To an+1、b1To bm+1For the adjustable parameter of described filter unit, described adjustable parameter is relevant with history original image；
Described frequency response results is the set of described frequency response values.
Optionally, described noise reduction module has 1+2t described filter unit, and kth filter unit is suitable to the view data of kth row/column pixel in image of interest position is processed, and k is more than or equal to 1 and less than or equal to the natural number of (1+2t).
Optionally, described noise reduction module only has a filter unit, and described filter unit is suitable to the view data of image of interest position pixel is carried out serial process.
The beneficial effect of technical solution of the present invention at least includes:
Technical solution of the present invention provides a kind of automatic focusing mechanism, it is different from prior art, it can pass through noise reduction module and obtain the frequency response results of original image pixels point, and the decrease of noise functions of original image is realized based on the frequency response results of passband portion, stablizing of picture signal can be realized under weak texture or low light level situation, thus the accuracy of auxiliary space territory focusing statistical result, automatic focusing mechanism focusing speed under low-light (level) and accuracy can be promoted, it is also possible to solve photographic subjects in weak texture situation and cannot focus or the problem of focusing time length.
The automatic focusing mechanism that technical solution of the present invention provides can configure more stable image information for auto-focusing process, from the frequency domain of view data, the valid data of image are first processed, view data is made to be subject to the impact of the factors such as ambient less, secondly by spatial domain operator, stable view data is reprocessed, export more good spatial domain result, statistical data of focusing accurately can be extracted, it is thus achieved that better focusing result.
In the alternative of technical solution of the present invention, technical solution of the present invention is further adapted for the enhancing and the noise removal function that use filter unit to realize picture signal, employ one filtering mode more flexibly, the configuration making filtering parameter is relevant to the focusing situation of history original image, the parameter that the filter function of filter unit is differently configured, for received image signal being made the Filtering Processing (low pass or bandpass) of different modes, it is derived from frequency information more preferably and reaches better noise reduction, and there is adjustability neatly.
Based on the motility that filter unit parameter is regulated by technical solution of the present invention, the accuracy that weak texture object is focused can be conducive to further, and reduce owing to focusing position can not Fast Convergent and the flating that causes, focus under the low light level also more stable, faster, the consuming time of entirety of focusing being decreased considerably focus process convergence rate on the whole.
In the alternative of technical solution of the present invention, input picture is also defined by technical solution of the present invention, input picture is carried out interested position choose, and according to the characteristic distributions of selected interested position pixel, multiple filter unit is set view data is carried out parallel processing, thus extension and the compatibility of image filtering mode can be realized, can directly promote the systematic function of focusing mechanism.
Accompanying drawing explanation
The structural representation of a kind of picture pick-up device that Fig. 1 provides for technical solution of the present invention；
Fig. 2 is the schematic diagram of lens system in the provided picture pick-up device of technical solution of the present invention and the concrete structure of shooting agent set；
The structural representation of a kind of automatic focusing mechanism that Fig. 3 provides for technical solution of the present invention；
A kind of filter unit that Fig. 4 provides for technical solution of the present invention implement structural representation；
A kind of frequency response curve adopting history original image that Fig. 5 provides for technical solution of the present invention is to obtain the frequency procedure schematic diagram of filter unit passband and stopband；
The schematic diagram choosing image of interest position pixel example first that Fig. 6 provides for technical solution of the present invention；
The structural representation of output focusing statistic curve under a kind of example that Fig. 7 provides for technical solution of the present invention；
The structural representation of the another kind of automatic focusing mechanism that Fig. 8 provides for technical solution of the present invention；
The structural representation of another automatic focusing mechanism that Fig. 9 provides for technical solution of the present invention.
Detailed description of the invention
Understandable in order to enable the purpose of the present invention, feature and effect to become apparent from, below in conjunction with accompanying drawing, the specific embodiment of the present invention is elaborated.
Elaborating a lot of detail in the following description so that fully understanding the present invention, but the present invention can also adopt other to be different from mode described here to be implemented, therefore the present invention is not by the restriction of following public specific embodiment.
The automatic focusing mechanism of technical solution of the present invention is applicable to following picture pick-up device.
With reference to Fig. 1, picture pick-up device 1 can be a kind of mobile phone camera module or camera, it includes lens system (LensSystem) 10 and shooting agent set 20, and wherein, shooting agent set 20 can obtain the view data of subject based on the photoconduction of lens system 10.
In conjunction with Fig. 2, lens system 10 is configured with and includes the battery of lens 11 by photographic lenss multiple in optical axis L and lens driver part 12, the plurality of photographic lens includes for realizing the lens (hereinafter referred to as to focus lens) that focusing controls, and lens driver part 12 includes voice loop (VoiceCoil) 13 and spring 14.Wherein, voice loop 13 can be focused signal (focusing voltage or focusing electric current) and drive, and can pass through self to rotate thus driving the lens in lens system to move the change of physical location between lens along optical axis L；Voice loop 13 coordinates with spring 14, and lens driver part 12 can be made to complete accurately to be controlled the action of battery of lens physical location by described focusing signal.The change of described control battery of lens lens physical location can only to the change that focus lens carries out physical location.
With continued reference to Fig. 1 and Fig. 2, shooting agent set 20 includes view finder (LVF, LiveViewFinder) 21 and imageing sensor 22 (ImageSensor, C-MOS sensor can be used to realize), by the battery of lens 11 of lens system 10 can by the light directing imageing sensor 22 from subject, it is hereby achieved that about the original image signal of subject.The original image signal obtained by imageing sensor 22 can generate the original image for recording, and described original image is applicable to auto-focusing, for the original image described in technical solution of the present invention.
Embodiment one
A kind of automatic focusing mechanism a1 as shown in Figure 3, it is adaptable to picture pick-up device (the similar picture pick-up device 1 of structure), including: lens 101, lens driver part 102, noise reduction module 103 and focusing control module 104.
Lens 101 are located in lens system 10, refer to the picture pick-up device 1 multiple lens in same optical axis L, and lens 101 include for realizing the lens that auto-focusing controls, also referred to as to focus lens.Focus lens can be moved forward and backward along the optical axis of lens, driven by the lens driver part 102 response to making with the focusing relevant signal (i.e. described focusing signal) of result, thus it is mobile to focus lens, change lens position, lens position can be detected by the lens position detection piece in or beyond automatic focusing mechanism, is conducive to the follow-up driving again to focus lens.Based on lens 101, the photoconduction of subject can be produced described original image.
With continued reference to the automatic focusing mechanism shown in Fig. 3.
Noise reduction module 103 is suitable to obtain the frequency response results of original image pixels point, and the frequency response results based on passband portion exports noise-reduced image.
Noise reduction module 103 can remove the image information of the non-effective frequency band of pixel view data original image from frequency domain angle, can the frequency band picture noise that cause under weak texture and low light level situation be filtered, thus exporting the image information of effective frequency section.
In the present embodiment, described original image pixels point refers to the process that image overall pixel directly carries out noise reduction module, but in other embodiments, this process object can also be image local pixel, such as the set of selected several pixels of subject subject area, i.e. the pixel of area-of-interest.
Concrete, obtaining described frequency response results in noise reduction module 103 can perform according to following process:
Asking for the frequency response values of described original image, described frequency response values can be image data representation of amplitude spectrum on frequency domain；
Choose the effective frequency section of original image based on the pass-band performance of wave filter, and export the frequency response values of effective frequency section；
The frequency response value set of original image pixels point effective frequency section is described frequency response results.
Described frequency response results is the result of image denoising at the present embodiment, and it is for removing the noise of the original image formed under a certain lens position, for stablizing picture signal, thus auxiliary focusing.
In the present embodiment, it is possible to a filter unit is set in noise reduction module 103 to realize the frequency response process of above-mentioned pixel: described filter unit is suitable to process to obtain pixel view data frequency response values according to following filter function expression formula to described original image pixels dot image data:
Wherein, m, n are the exponent number of described filter unit, a1To an+1、b1To bm+1Adjustable parameter for described filter unit.
Described filter unit uses the iir filter of biquadratic filter (dualbiquadfilter) form to realize, and is four-step filter, and the concrete value of m and n can be 4.
What Fig. 4 illustrated the general biquadratic filter form iir filter of a kind of MATLAB of utilization tool design implements structure, comprising:
Signal summation (SUM) module, signal cutout (CLIP) module, signal delay (Z-1) module and adjustable signal gain module (GAIN), comprising adjustable gain coefficient is c0～c10, adjusts gain coefficient c0～c10 module and adjustable parameter a1To an+1、b1To bm+1Equivalent, i.e. fixed adjustable parameter a1To an+1、b1To bm+1Numerical value, it is determined that the numerical value of regulation coefficient c0～c10；Or, it has been determined that the numerical value of regulation coefficient c0～c10, it is determined that adjustable parameter a1To an+1、b1To bm+1Numerical value.
When using MATLAB instrument, the design tool case of available wave filter supports that described iir filter performance design and parameter generate, can be changed filter passband frequency and cut-off frequency by regulation coefficient c0-c10, c0～c10 parameter specific design mode is referred to the conventional filter design method used of MATLAB instrument and is designed.
No matter it is to adjustable parameter a1To an+1、b1To bm+1Numerical value determine, or the design to c0～c10 parameter, it is all relevant to the frequency response situation of original image.The frequency response situation of original image is then relevant to described imageing sensor 22, consider the picture signal that the generation of described original image exports based on the photoconduction of imageing sensor reception, the present embodiment utilizes the picture signal that picture pick-up device obtains based on imageing sensor and the raw image data recorded, obtain the frequency response situation of raw image data, so that it is determined that in noise reduction module 103 filtering parameter of pixel frequency response values institute foundation.
The original image that above-mentioned picture pick-up device obtains picture signal based on imageing sensor and records can also be referred to as history original image.
The filtering parameter determined based on history original image, relates generally to the frequency of filter unit passband and stopband:
Determine that the frequency of filter unit passband and stopband is really the scope in order to obtain effective frequency section, so-called effective frequency section is really a technical solution of the present invention index to the possessed picture noise of original image, the quantity of information situation of pixel effective frequency section can help to reflection effective image information, and effective frequency section itself has filtered picture noise frequency range and other invalid band informations, aid in the focusing accuracy promoting focusing mechanism and focusing efficiency.
The present embodiment specifically adopts the frequency response curve of history original image to obtain the frequency of filter unit passband and stopband, the history original image adopted at least includes a width correct focusing position image and an incorrect focusing position image, in conjunction with Fig. 5, only to use two images to determine the frequency of filter unit passband and stopband, its acquisition process is as follows:
Input view data from imageing sensor 22 (picture signal, namely the original image that records of picture pick-up device or history original image)；
Use and regarded as the first image produced by correct focusing position and regarded as the second image produced by the fuzzyyest focusing position；
First image and the second image are made two-dimensional fast fourier transform respectively, the first image and the second image are converted to frequency domain；
Take data on the diagonal of the first image and the second two-dimensional image Fourier transform results matrix, as described first image and the second image graph frequency response curve, the frequency response curve of the first image and the second image is plotted in together；
Based on said frequencies response curve, the frequency of filter passband and stopband can be determined by analyzing the difference of the first image and the second image frequency response；
Typical filter design method (Butterworth filter design/other, available MATLAB workbox realizes) is finally used to generate the filter coefficient (can be above-mentioned adjustable parameter or regulation coefficient) of required exponent number.
In other embodiments, can be extended according to the method for above-mentioned acquisition process: as multiple images can be used, can also adding other kinds of reference picture to input, above-mentioned image all can embody the image conditions on focusing tram and the image conditions in focusing ambiguous location.After obtaining input picture, each image is done fast two-dimensional fourier transformation respectively, image is transformed into frequency domain；Take data on the diagonal of piece image two-dimension fourier transform matrix of consequence, as this width image frequency response curve, the frequency response curve of multiple figure is plotted in together；The frequency of filter passband stopband it is assured that by analyzing the difference of each figure frequency response；Finally use typical filter design method to generate the filter coefficient of required exponent number.
The coefficient feature of above-mentioned filter parameter is obtained, it is adaptable to the wave filter of the present embodiment is low pass filter or band filter according to the present embodiment.
With continued reference to the automatic focusing mechanism shown in Fig. 3.
Focusing controls module 104 and is suitable to according to the output focusing statistical result of noise-reduced image.
Above-mentioned focusing statistical result is the evaluation of estimate to original image focusing situation, namely for evaluating under a certain lens position, whether the original image formed has good focusing performance, meets acceptable focusing requirement accurately.
The pixel frequency response results that described noise-reduced image exports based on noise reduction module 103 produces.
The present embodiment realizes above-mentioned focusing in the following way and controls module 104, and namely focusing controls to realize in module 104:
Extract the spatial domain boundary operator on noise-reduced image image of interest position；
It is multiplied according to edge direction corresponding with the pixel value on described image of interest position for described spatial domain boundary operator and sue for peace, to obtain spatial domain result.
Described image of interest position is made up of the pixel around interested pixel point and this interested pixel point, it is believed that be pixel number for (1+2t)2Pixel region, wherein, t is the natural number more than or equal to 1, and the pixel of t+1 row in this pixel region, t+1 row is described interested pixel point, and rest of pixels point is described pixel around.
Fig. 6 illustrates the example first of above-mentioned image of interest position pixel, and it includes 3 × 3 pixels, wherein, minimum grid represents a pixel, the pixel of the 2nd row the 2nd row is described interested pixel point, and all the other are the pixel around described interested pixel point, the element I in grid00～I22Represent the pixel value of pixel representated by this grid successively.
In the present embodiment, if described spatial domain boundary operator is Laplace operator (Laplace), described edge direction includes corresponding vertical direction and horizontal direction, in concrete spatial-domain algorithm, obtains spatial domain result according to following manner:
It is multiplied corresponding with the pixel value on described image of interest position for described Laplace operator and sue for peace in the vertical direction, obtaining vertical direction result；
It is multiplied corresponding with the pixel value on described image of interest position for described Laplace operator and sue for peace in the horizontal direction, obtaining horizontal direction result；
Described spatial domain result is the absolute value absolute value sum with horizontal direction result of vertical direction result.
Still for the image of interest position pixel shown in Fig. 6, the pixel value on the pixel of known image of interest position is followed successively by I as shown in Figure 600～I22, corresponding above-mentioned pixel I in horizontal direction00～I22The Laplace operator of position respectively 0,0,0,1 ,-2,1,0,0 and 0, corresponding above-mentioned pixel I in vertical direction00～I22The Laplace operator of position respectively 0,1,0,0 ,-2,0,0,1 and 0, then have space field result FVLMeet:
FVL=abs (I10-2×I11+I12)+abs(I01-2×I11+I21)；
Wherein, (I10-2×I11+I12) for described horizontal direction result, abs (I01-2×I11+I21) for described vertical direction result.
In other embodiments, described spatial domain boundary operator is not limited to Laplace operator, it is possible to use any other spatial domain arithmetic operators, such as Sobel Operator (Sobel).
Still for the image of interest position pixel shown in Fig. 6, edge direction includes corresponding vertical direction and horizontal direction, corresponding above-mentioned pixel I in horizontal direction00～I22The Sobel Operator of position respectively 1,2,1,0,0,0 ,-1 ,-2 and-1, corresponding above-mentioned pixel I in vertical direction00～I22The Sobel Operator of position respectively 1,0 ,-1,2,0 ,-2,1,0 and-1, then adopt Sobel Operator then to obtain spatial domain result FVsMeet:
FVs=abs (I00+2×I01+I02-I20-2×I21-I22)+abs(I00+2×I10+I20-I02-2×I12-I22)；Wherein, (I00+2×I01+I02-I20-2×I21-I22) for horizontal direction result under Sobel Operator, (I00+2×I10+I20-I02-2×I12-I22) for vertical direction result under Sobel Operator.
Certainly, in other embodiments, described spatial domain operator can also is that other spatial domain operators；It addition, described edge direction is also not necessarily limited to vertical direction and horizontal direction, can being determined by spatial domain arithmetic operators, if in other embodiments, spatial domain arithmetic operators also comprises the edge extracting on other directions, then analogize according to upper method and be calculated.
In the process that lens driver part 102 drives focus lens, lens position is continually changing, in lens 101, the position between multiple lens is also changed, thus imageing sensor 22 continues to receive in the light of subject from associated lens photoconduction, obtain and continue different picture signals, so producing the multiple original images for recording.
In the present embodiment, focusing is controlled the spatial domain result output of relevant multiple original images processed through noise reduction module 103 that module 104 exports, described spatial domain result is the focusing statistical information of the present embodiment, for evaluation of focusing, the mode of available focusing statistic curve (FV curve) is intuitively illustrated.
Fig. 7 gives the schematic diagram of the focusing statistic curve of output in a kind of situation of focusing of the present embodiment, wherein, abscissa is that lens driver part 102 drives lower focusing position change (or being interpreted as focusing lens position change) situation (unit, mm), vertical coordinate is described spatial domain result (unit on the original image produced under certain focusing position, nothing, it will be appreciated that for being scalar, with the numerical value of data volume for comparing between the numerical value of vertical coordinate).
Known based on Fig. 7, can from focusing statistic curve, namely each original image correspondence spatial domain result searches expected spatial domain result, based on this focusing statistic curve, described expected spatial domain result can be the global maximum of described focusing statistic curve.This expected lens position corresponding to spatial domain result is the focusing position finally assert.Lens driver part 102 receives the described focusing signal sent according to the focusing position assert, thus driving focus lens, completes focusing.In the figure 7, on focusing statistic curve, abscissa value 17mm corresponding for some o is described focusing position.
The automatic focusing mechanism of the present embodiment adds noise reduction module 103, and use filter cell to realize noise reduction module 103, utilize bandpass and the low-pass characteristic of filter unit, information on the non-effective frequency band of original image is filtered, thus obtaining stable picture signal, the spatial domain result of module 104 is controlled by focusing, autofocus system focusing speed under low-light (level) and accuracy can be promoted, it is also possible to solving photographic subjects is cannot focus in weak texture situation or the problem of focusing time length.
From Fig. 7 focusing statistic curve generated, the focusing statistic curve that the present embodiment automatic focusing mechanism produces possesses good stability, it is substantially not present the situation of curve shake, and the steep of curve itself is more suitable for the focusing of weak texture subject, focusing position is made to be easier to searched acquisition.
The filter factor (adjustable parameter and regulation coefficient) of the filter unit of the present embodiment can also be for the focusing situation of different imageing sensors and history original image and is optimized, the selection making filter unit band connection frequency has more motility, it is possible to be greatly promoted focusing accuracy and the adaptability of focusing mechanism.
The structure of the present embodiment lens driver part 102 is consistent with lens driver part 12.
Embodiment two
A kind of automatic focusing mechanism a2 as shown in Figure 8, is different from embodiment one, also includes: focusing position search module 105.
Focusing position search module 105 is suitable to drive, at described lens driver part 102, the spatial domain result constantly obtaining focusing control module 104 output in the described process to focus lens, the corresponding relation of search lens position and described spatial domain result, and identify described expected frequency response results based on described corresponding relation.
Focusing statistic curve corresponding to embodiment one, described corresponding relation intuitively can be obtained based on focusing statistic curve, focusing position search module 105 identifies that the process of expected spatial domain result is exactly ask for the process of focusing statistic curve global maximum, by focusing position output corresponding for global maximum, then can send corresponding signal to lens driver part 102, make lens driver part 102 drive focus lens to described focusing position.
In other embodiments, focusing position search module 105 can also be integrated in focusing control module 104.
Embodiment three
The present embodiment is different from embodiment one, which show a kind of method concurrently original image pixels point being carried out noise reduction, and wherein, original image pixels point is to gather with the form of detection window of focusing.
A kind of automatic focusing mechanism a3 as shown in Figure 9, including: lens 101, lens driver part 102, noise reduction module 103 ' and focusing control module 104.Wherein, noise reduction module 103 ' is suitable to obtain the frequency response results on original image image of interest position, and the frequency response results based on passband portion exports noise-reduced image.Described focusing detection window is suitable to catch original image image of interest position.
Described image of interest position is made up of the pixel around interested pixel point and this interested pixel point, is (1+2t) for pixel number2Pixel region, t is the natural number more than or equal to 1, and wherein, the pixel of t+1 row in this pixel region, t+1 row is described interested pixel point, and rest of pixels point is described pixel around.
Noise reduction module 103 ' described in the present embodiment has 1+2t described filter unit, wherein, 1st filter unit is suitable to the view data of the pixel to the 1st row in image of interest position (or the 1st row) and processes, 2nd filter unit is suitable to the view data of the pixel to the 2nd row in image of interest position (or the 2nd row) and processes, by that analogy, the 1+2t filter unit is suitable to the view data of the pixel to 1+2t row in image of interest position (or 1+2t row) and processes, thus, the noise reduction module 103 ' of the present embodiment realizes the parallel processing of pixel in original image area-of-interest.
Shown in Fig. 9 is the situation of t=1, focusing detection window captures the pixel of 3 × 3 in original image location, then the noise reduction module 103 ' of automatic focusing mechanism a3 includes 3 and can be respectively used on this 3 × 3 picture position of parallel processing the filter unit of 1～3 row pixel frequency response values, each filter unit is suitable to process the pixel of a line input, the output of each filter unit then together as the frequency response results of this original image, and produce the original image after denoising be transferred to described focusing control module 104.
Although the present invention is with preferred embodiment openly as above; but it is not for limiting the present invention; any those skilled in the art are without departing from the spirit and scope of the present invention; may be by the method for the disclosure above and technology contents and technical solution of the present invention is made possible variation and amendment; therefore; every content without departing from technical solution of the present invention; according to any simple modification, equivalent variations and modification that above example is made by the technical spirit of the present invention, belong to the protection domain of technical solution of the present invention.
Claims (21)
1. an automatic focusing mechanism, it is characterised in that including: lens, lens driver part, noise reduction module and focusing control module, based on described lens, the photoconduction of subject can be produced original image；
Described noise reduction module is suitable to obtain the frequency response results of original image pixels point, and the frequency response results based on passband portion exports noise-reduced image；
Described focusing controls module and is suitable to extract the spatial domain boundary operator on noise-reduced image image of interest position, is multiplied according to edge direction corresponding with the pixel value on described image of interest position for described spatial domain boundary operator and sue for peace, to obtain spatial domain result；Described lens driver part is suitable to drive described lens in the hope of expected spatial domain result, and lens position when being up to described expected spatial domain result drives described lens as focusing position.
2. device as claimed in claim 1, it is characterized in that, described noise reduction module includes filter unit, and described filter unit is suitable to the view data to original image pixels point and processes to obtain the frequency response values of pixel view data according to following filter function expression formula:
Wherein, m, n are the exponent number of described filter unit, a1To an+1、b1To bm+1For the adjustable parameter of described filter unit, described adjustable parameter is relevant with history original image；
Described frequency response results is the set of described frequency response values.
3. device as claimed in claim 2, it is characterised in that described filter unit is four-step filter.
4. device as claimed in claim 2, it is characterised in that described adjustable parameter is determined based on the frequency response curve of some history original images, and described history original image at least includes a width correct focusing position image and an incorrect focusing position image.
5. device as claimed in claim 4, it is characterised in that the frequency response curve based on each history original image determines described filter unit band connection frequency, thus generating described adjustable parameter.
6. device as claimed in claim 2, it is characterised in that described filter unit is low pass filter or band filter.
7. device as claimed in claim 1, it is characterised in that described spatial domain boundary operator is Laplace operator or Sobel Operator, and described edge direction includes vertical direction and horizontal direction；
It is multiplied corresponding with the pixel value on described image of interest position for described spatial domain boundary operator and sue for peace in the vertical direction, obtaining vertical direction result；
It is multiplied corresponding with the pixel value on described image of interest position for described spatial domain boundary operator and sue for peace in the horizontal direction, obtaining horizontal direction result；Described spatial domain result is the absolute value absolute value sum with horizontal direction result of vertical direction result.
8. device as claimed in claim 1, it is characterised in that described image of interest position is made up of the pixel around interested pixel point and this interested pixel point.
9. device as claimed in claim 1, it is characterized in that, also include: focusing position search module: described focusing position search module is suitable to drive at described lens driver part constantly obtain described spatial domain result in the process of described lens, to generate the corresponding relation of lens position and described spatial domain result, and identify described expected spatial domain result based on described corresponding relation.
10. device as claimed in claim 9, it is characterised in that described focusing position search module is integrated in described focusing and controls module.
11. device as claimed in claim 1, it is characterised in that also include: image sensing module；
Described image sensing module is suitable to reception, from lens, the photoconduction of subject is produced picture signal, and described original image is generated based on described picture signal.
12. device as claimed in claim 11, it is characterised in that described image sensing module is made up of the C-MOS sensor as imaging apparatus.
13. an automatic focusing mechanism, it is characterised in that including: lens, lens driver part, noise reduction module and focusing control module, based on described lens, the photoconduction of subject can be produced original image；
Described noise reduction module is suitable to obtain the frequency response results on original image image of interest position, and the frequency response results based on passband portion exports noise-reduced image；
Described focusing controls module and is suitable to extract the spatial domain boundary operator of noise-reduced image, is multiplied according to edge direction corresponding with the pixel value of described noise-reduced image for described spatial domain boundary operator and sue for peace, to obtain spatial domain result；Described lens driver part is suitable to drive described lens in the hope of expected spatial domain result, and lens position when being up to described expected spatial domain result drives described lens as focusing position.
14. device as claimed in claim 13, it is characterized in that, described image of interest position is made up of the pixel around interested pixel point and this interested pixel point, for the pixel region that pixel number is (1+2t) 2, t is the natural number more than or equal to 1, wherein, the pixel that in this pixel region, t+1 row, t+1 arrange is described interested pixel point, and rest of pixels point is described pixel around.
15. device as claimed in claim 14, it is characterized in that, described noise reduction module includes filter unit, and described filter unit is suitable to the view data to image of interest position pixel and processes to obtain the frequency response values of pixel view data according to following filter function expression formula:
Wherein, m, n are the exponent number of described filter unit, a1To an+1、b1To bm+1For the adjustable parameter of described filter unit, described adjustable parameter is relevant with history original image；
Described frequency response results is the set of described frequency response values.
16. device as claimed in claim 15, it is characterized in that, described noise reduction module has 1+2t described filter unit, kth filter unit is suitable to the view data of kth row/column pixel in image of interest position is processed, and k is more than or equal to 1 and less than or equal to the natural number of (1+2t).
17. device as claimed in claim 15, it is characterised in that described noise reduction module only has a filter unit, and described filter unit is suitable to the view data of image of interest position pixel is carried out serial process.
18. device as claimed in claim 15, it is characterised in that described filter unit is four-step filter.
19. device as claimed in claim 15, it is characterised in that described adjustable parameter is determined based on the frequency response curve of some history original images, and described history original image at least includes a width correct focusing position image and an incorrect focusing position image.
20. device as claimed in claim 19, it is characterised in that the frequency response curve based on each history original image determines described filter unit band connection frequency, thus generating described adjustable parameter.
21. device as claimed in claim 15, it is characterised in that described filter unit is low pass filter or band filter.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410827938.0A CN105791668B (en) | 2014-12-25 | 2014-12-25 | Automatic focusing mechanism |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410827938.0A CN105791668B (en) | 2014-12-25 | 2014-12-25 | Automatic focusing mechanism |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105791668A true CN105791668A (en) | 2016-07-20 |
| CN105791668B CN105791668B (en) | 2019-03-05 |
Family
ID=56389392
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410827938.0A Active CN105791668B (en) | 2014-12-25 | 2014-12-25 | Automatic focusing mechanism |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105791668B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108765346A (en) * | 2018-05-30 | 2018-11-06 | 北京图森未来科技有限公司 | A kind of auxilary focusing method, device and readable medium |
| CN112689988A (en) * | 2020-05-28 | 2021-04-20 | 深圳市大疆创新科技有限公司 | Image processing method, image processing apparatus, removable platform, and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2008046351A (en) * | 2006-08-16 | 2008-02-28 | Canon Inc | Automatic focusing device and imaging apparatus |
| CN101175152A (en) * | 2006-10-30 | 2008-05-07 | 索尼株式会社 | Image processing apparatus, image capturing apparatus, image processing method |
| US20090047010A1 (en) * | 2007-07-31 | 2009-02-19 | Akihiro Yoshida | Imaging device and imaging method |
| CN101790031B (en) * | 2009-01-23 | 2012-07-04 | 索尼公司 | Image processing apparatus, image processing method, and imaging apparatus |
-
2014
- 2014-12-25 CN CN201410827938.0A patent/CN105791668B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2008046351A (en) * | 2006-08-16 | 2008-02-28 | Canon Inc | Automatic focusing device and imaging apparatus |
| CN101175152A (en) * | 2006-10-30 | 2008-05-07 | 索尼株式会社 | Image processing apparatus, image capturing apparatus, image processing method |
| US20090047010A1 (en) * | 2007-07-31 | 2009-02-19 | Akihiro Yoshida | Imaging device and imaging method |
| CN101790031B (en) * | 2009-01-23 | 2012-07-04 | 索尼公司 | Image processing apparatus, image processing method, and imaging apparatus |
Non-Patent Citations (1)
| Title |
|---|
| 《信息科技辑》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108765346A (en) * | 2018-05-30 | 2018-11-06 | 北京图森未来科技有限公司 | A kind of auxilary focusing method, device and readable medium |
| CN112689988A (en) * | 2020-05-28 | 2021-04-20 | 深圳市大疆创新科技有限公司 | Image processing method, image processing apparatus, removable platform, and storage medium |
| WO2021237577A1 (en) * | 2020-05-28 | 2021-12-02 | 深圳市大疆创新科技有限公司 | Image processing method and apparatus, and movable platform, and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105791668B (en) | 2019-03-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110691193B (en) | Camera switching method and device, storage medium and electronic equipment | |
| JP6271990B2 (en) | Image processing apparatus and image processing method | |
| CN102053339B (en) | Automatic focusing method | |
| CN102891960B (en) | Method and camera for determining an image adjustment parameter | |
| CN110121882A (en) | A kind of image processing method and device | |
| CN105141840B (en) | Information processing method and electronic equipment | |
| CN102113017A (en) | System and method to generate depth data using edge detection | |
| US20120300115A1 (en) | Image sensing device | |
| CN108076278A (en) | A kind of Atomatic focusing method, device and electronic equipment | |
| CN101729780A (en) | Automatic focusing method of optical imaging system | |
| CN104506762A (en) | Optical field collecting control method and device and optical field collecting equipment | |
| JP2013117969A (en) | System and method for generating robust depth map utilizing multiple resolution procedure | |
| CN104301596A (en) | Video processing method and device | |
| CN104270573A (en) | Multi-touch focus imaging system and method, as well as applicable mobile terminal | |
| CN104902182A (en) | Method and device for realizing continuous auto-focus | |
| CN106412423A (en) | Focusing method and device | |
| CN105446056B (en) | Automatic focusing mechanism and method | |
| CN105791666A (en) | Automatic focusing device | |
| CN105791668A (en) | Automatic focusing device | |
| US8026975B2 (en) | Image capturing apparatus and control method therefor | |
| CN106922181A (en) | Directional perception is focused on automatically | |
| JP6645711B2 (en) | Image processing apparatus, image processing method, and program | |
| CN106303199A (en) | focusing system and focusing method | |
| CN114500837A (en) | Shooting method and device and electronic equipment | |
| Srivastava et al. | Design and implementation of a real-time autofocus algorithm for thermal imagers |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CP03 | Change of name, title or address | ||
| CP03 | Change of name, title or address |
Address after: 100089 18 / F, block B, Zhizhen building, No.7, Zhichun Road, Haidian District, Beijing Patentee after: Beijing Ziguang zhanrui Communication Technology Co.,Ltd. Address before: 100084, Room 516, building A, Tsinghua Science Park, Beijing, Haidian District Patentee before: BEIJING SPREADTRUM HI-TECH COMMUNICATIONS TECHNOLOGY Co.,Ltd. |