CN1719328A - Method of automatic exposure regulation and its control system - Google Patents

Abstract

The present invention relates to an automatic exposure regulation method and its control system. Said method mainly includes the following steps: setting exposure control basic parameters including target luminance standard value Ytarget, luminance threshold value Ythd, luminance gain value Gtarget, and luminance gain value threshold value, making exposure, extracting image and making image processing, reading exposure time T1, luminance gain G1 and image exposure intensity statistical value Y1, judging exposure intensity value, calculating optimum exposure time T2 and using it to regulate exposure time T1 of image data input module.

Description

A kind of method and control system that is used for the automatic exposure adjusting

Technical field

The present invention relates to a kind of method and control system that automatic exposure is regulated that be used for, particularly be applied to the method and the control system of the automatic exposure adjusting of film or digital photosensitive parts (CCD, CMOS etc.).

Background technology

Exposure is that film or digital photosensitive parts (CCD, CMOS etc.) reception are advanced the process that light forms image from camera lens.Time shutter promptly is the time that photosensitive-member receives light.Under same environment conditions such as (do not change) intensities of illumination, the time shutter is long more, and image is bright more; Otherwise the time shutter is short more, and image is dark more.If scenery is bright excessively in the image, and bright part do not have level or details, and it is over-exposed that Here it is; If photo is dark, can't truly reflect the color and luster of scenery, be exactly under-exposed.So, when environment changes, also to adjust the time shutter accordingly so that the brightness of image of output in suitable scope, i.e. suitable increase time shutter when environment is dark is in the brighter environment minimizing time shutter.But, different being provided with of following time shutter of Different Light.

Below see figures.1.and.2 and describe the characteristics of light source.Perfect light source can be divided into two kinds: a kind of is that the intensity of light source does not change in time, as shown in Figure 1; Another kind is that the intensity of light source is along with time generating period variation, as shown in Figure 2.In daily life, Xiang Guan light mainly contains: the light that sunshine and room lighting are used.Wherein, because that sunshine has is lasting, stable, research thinks that sunshine is perfect light source A for convenience.At present, room lighting is used up general relevant with local line frequency, and its light intensity is generally all along with the time changes.Certainly also having many illumination lights not have frequency characteristic or some substantially now is to adopt the direct current supply mode, but this is less after all, is not discussed here.Ac frequency on the our times substantially all is 50Hz or 60Hz, and according to the relation of supply voltage and power, the frequency that can obtain indoor light source is 100Hz or 120Hz.Simple derivation is as follows:

U(t)＝U ₀sin(ωt) (1)

P (t)=U (t) ²/ R R is a lamp resistance, is constant (2)

P(t)＝kU(t) ²＝kU ₀ ²sin ²(ωt)＝kU ₀ ²(1-cos(2ωt))/2(3)

When w=50Hz, the frequency of P (t) is 100Hz, and Y and P (t) are linear relationship, so the frequency of indoor light source is 100Hz or 120Hz.

Therefore just possesses exposure characteristics as described below.Change time shutter of view data load module (Sensor) according to the difference of environment, so that its brightness of image of importing into is suitable.The depth of exposure of view data load module and the relation of time shutter as shown in Figure 3, Fig. 3 shows the graph of relation of sensitometric characteristic and time shutter.

As can be seen from Figure 3, along with the increase of time shutter, the depth of exposure of image is also increasing.But the time shutter is big more, and its contribution to depth of exposure is more little, and the recruitment of promptly identical time shutter hour is greater than exposure time values when big to the influence of brightness in the value of time shutter.

Below be the test data of under the room lighting situation different incident intensities being tested, each data are all measured 2 groups, and twice difference all between ± 1, is therefore all got first group of data.

Wherein:

The unit of time shutter is 10ms

The unit of exposure numerical value is the electronics quantized value, and maximal value is 255, and minimum value is 0

Form 1

Time shutter	1	2	3	4	5	6	7	8	9
										For the first time	33	67	90	110	128	143	157	169	180
For the second time	30	62	86	105	121	134	146	156	166
										For the third time	37	73	99	118	133	146	159	172	183

In actual applications, if the time shutter is provided with under the indoor environment when not being the integral multiple in indoor light source cycle (the domestic ac frequency following cycle of 50HZ is 10 milliseconds), the light and shade striped can appear in the image of different frame, then can not occur when the time shutter is the integral multiple in light source cycle, this phenomenon is called the Flicker phenomenon.

Usually the view data load module that adopts has two kinds of CMOS and CCD, and its production technology has some differences, and therefore the mode of exposure is also different.The cmos image data input module adopts the row exposure technique, promptly each delegation that exposes, interval certain hour between capable and capable (be worth different vendor concrete interval time different regulations), therefore, the exposure starting point of each row is all inequality, if the time shutter is not the integral multiple in cycle, then each capable light intensity of being accepted might be different, though time shutter of each row is the same, also can cause the different variation of light and shade, i.e. our said Flicker phenomenon.What the ccd image data input module adopted is the face exposure technique, single exposure one entire image, if the time shutter that we are provided with is incorrect, though the light and shade that can not go out between existing and the row changes, but can cause the light and shade difference between consecutive image, give the visual experience of a kind of image flicker of people, this also is a kind of of Flicker phenomenon.

The reason that the Flicker phenomenon occurs is as follows:

, can know for its brightness of piece image to determine that by light intensity and time shutter under the certain situation of light intensity, the time shutter, long more image was bright more about the theoretical description that exposes by the front; Otherwise if the time shutter is certain, the strong more image of light intensity is also bright more.Therefore, under the condition of identical time shutter, intensity variations also can cause the light and shade inequality of image.

The voltage of having known indoor light source changed with the cycle, and it changes shown in (1), and intensity variations is shown in (3), and the light intensity computing formula in a period of time is as follows so:

${\∫}_1^{2}P\left(t\right)\mathrm{dt}={\∫}_1^2\mathrm{kU}{\left(t\right)}^2\mathrm{dt}=\frac{1}{2}{\mathrm{kU}}_0^2{\∫}_1^2(1-\cos \left(2\mathrm{\ωt}\right))\mathrm{dt}=\frac{1}{2}{{\mathrm{kU}}_0}^2[(t_2-t_1)-\frac{1}{2\mathrm{\ω}}\sin 2\mathrm{\ω}(t_2-t_1)]---\left(4\right)$

Make Δ t=t ₂-t ₁, then following formula becomes:

${\∫}_1^{2}P\left(t\right)\mathrm{dt}=\frac{1}{2}{{\mathrm{kU}}_0}^2[\mathrm{\Δt}-\frac{1}{2\mathrm{\ω}}\sin 2\mathrm{\ω\Δt}]---\left(5\right)$

As seen, if Δ t equals period T, following formula is a constant

Otherwise be variate.This

Illustrate when the time shutter is the integral multiple of one-period, though initial exposure station from which, light intensity act as a constant, promptly exposure effect is identical.Otherwise, then can cause light intensity effect inequality.

When the intensity of light source changed with the cycle (as indoor situation), the time shutter of setting must be the integral multiple of its period of change.When the intensity of light source is constant, not limited.

By preceding described, under indoor situation, the setting of time shutter is necessary for the integral multiple of indoor light source period of change T to avoid the generation of flicker phenomenon, and this has just caused another problem: the uncontinuity that the time shutter is set.The time shutter of She Zhiing is not the integral multiple (in fact most applications all is like this) of T just if desired, and that just can not obtain the brightness of image of needs.

Automatic exposure is the problem that all products that relate to Digital Image Processing (as digital camera etc.) all will solve, the effect that it is not to use the person the most directly to experience yet.

Past can only be adopted and approach fitting method owing to do not found the mathematical relation of depth of exposure and time shutter, and promptly increase little by little or minimizing time shutter and luminance gain are come the approximation theory value, and the efficient of this method obviously is very low.

Summary of the invention

The purpose of this invention is to provide method and control system that a kind of automatic exposure is regulated, the method that automatic exposure of the present invention is regulated comprises following steps:

(1) sets exposure control basic parameter, determine the target brightness value scope;

(2) after exposure and extracting the processing image, judge current brightness value Y ₁Whether in the target brightness value scope, if, continue by original parameter exposure and handle image, if not, then carry out following automatic exposure regulating step;

(3) calculate required luminance gain value G ₂, judge whether to utilize the adjustment of luminance gain value brightness value can be adjusted in the target brightness value scope; If, then with luminance gain value G ₁Be adjusted into G ₂Value, handle image by new luminance gain value again, if not, carry out next step;

(4) calculate optimum time shutter T2, adjust time shutter T1 with optimum time shutter T2.

The exposure control basic parameter of setting in step (1) comprises object brightness standard value Y _Target, luminance threshold Y _Thd, described definite target brightness value scope is: [Y _Target-Y _Thd, Y _Target+ Y _Thd].

Make when avoiding adjusting the luminance gain value brightness value change the discontinuous flicker that causes, earlier according to the luminance gain threshold value G of initial setting up _Thd, set different time and guarantee that down brightness changes continuous luminance gain threshold values: utilize formula G _Thdn+1=(G _Target-n * G _Thd.n) the luminance gain threshold value G that allows when calculating different exposure time of ÷ (n+1)+offset _Thdn, and result of calculation is stored in the luminance gain threshold value table of comparisons, wherein n is the multiple of time shutter to intensity of illumination period of change time T, offset is an arbitrary integer, G _Thd.1=G _Thd,

In step (3), the described concrete grammar that utilizes the adjustment of luminance gain value brightness value can be adjusted in the target brightness value scope that judges whether is:

(3.1) utilize formula $G_{}$ ${}_2\frac{\ln (M-Y_{t\arg \mathrm{et}})\ln \left(M\right)}{\ln (M-Y_1)-\ln \left(M\right)}\&times;{\mathrm{<figure-callout\; id="1"\; label="G"\; filenames="A20051008032700231.png"\; state="\{\{state\}\}">G</figure-callout>}}_1$

Calculate required luminance gain value

G ₂, M is the brightness range maximal value in the formula;

(3.2) judge whether G _Target-G _Thd.n≤ G ₂≤ G _Target+ G _Thd.n

In step (4), the optimum time shutter T of described calculating ₂Be to utilize formula

$f\left(Y_1\right)=\frac{\ln (M-Y_1)-\ln \left(M\right)}{\ln (M-Y_{t\arg \mathrm{et}})-\ln \left(M\right)}$

Calculate f (Y ₁); Utilize formula then $T_{t\arg \mathrm{et}}=\frac{G_1{T}_1}{f\left(Y_1\right)G_{t\arg \mathrm{et}}}$

Calculate target time shutter T _Target, get T _TargetBe optimum time shutter T ₂

The method for optimizing that calculates the target time shutter is earlier different exposure intensities and logarithm value thereof to be stored in the table of comparisons, and maximum exposure intensity is set at constant M, when the system handles data are 8 integers, and constant M=256; Calculate according to the following steps then:

(4.1) read Y ₁, M-Y ₁, M-Y _TargetAnd M is to the logarithm value among the Gui Zhaobiao;

(4.2) utilize formula $f\left(Y_1\right)=\frac{\ln (M-Y_1)-\ln \left(M\right)}{\ln (M-Y_{t\arg \mathrm{et}})-\ln \left(M\right)}$ Calculate f (Y ₁);

(4.3) utilize formula $T_{t\arg \mathrm{et}}=\frac{G_1{T}_1}{f\left(Y_1\right)G_{t\arg \mathrm{et}}}$ Calculate T _Target

For avoiding the flicker phenomenon, intensity of illumination period of change time T is set also in step (1); In step (4), utilize formula to calculate target time shutter T _TargetAfter also carry out following processing: if the target time shutter T that calculates _TargetFor intensity of illumination period of change time T integral multiple, then get time T _TargetBe T ₂, if the target time shutter T that calculates _TargetBe not the integral multiple of intensity of illumination period of change time T, the time of then getting is near T _TargetThe integer multiple of T be T ₂

Because the optimum time shutter of selecting is not target time shutter (time shutter that object brightness requires), calculating optimum time shutter T ₂After can also carry out following steps and adjust the luminance gain value and carry out exposure compensating:

Use formula $G_{}$ ${}_2=\frac{G_1{T}_1}{f\left(Y_1\right){\mathrm{<figure-callout\; id="2"\; label="T"\; filenames="A20051008032700231.png"\; state="\{\{state\}\}">T</figure-callout>}}_2}$

Calculate optimum luminance gain value G ₂, and adjust with result of calculation

Luminance gain value G in the image processing module ₁

Calculating G ₂After, according to time shutter T ₂Read corresponding luminance gain value threshold value G in the luminance gain threshold value table of comparisons _Thdn, judge whether G earlier ₂At G _TargetIn the scope that allows: if G ₂＞G _Target+ G _Thdn, then get G ₂=G _Target+ G _Thd.nIf G ₂＜G _Target-G _Thd.n, then get G ₂=G _Target-G _Thd.nIf G _Target-G _Thd.n≤ G ₂≤ G _Target+ G _Thd.n, then do not revise G ₂And then use G ₂Adjust luminance gain value G ₁

The control system that also provides a kind of automatic exposure to regulate of the present invention, comprise view data load module (sensor), control device, view data output module (LCD), the view data load module view data that obtains of will exposing sends control device to, behind the control device image data processing from view data output module output image, is furnished with the register that to revise view data load module time shutter and/or Flame Image Process luminance gain value in the control device, and be provided with register that target brightness value and automatic exposure produce the brightness change threshold that interrupts and the formula of realizing the calculation exposure time $f\left(Y_1\right)=\frac{\ln (M-Y_1)-\ln \left(M\right)}{\ln (M-Y_{t\arg \mathrm{et}})-\ln \left(M\right)}$ And

${\mathrm{<figure-callout\; id="2"\; label="T"\; filenames="A20051008032700231.png"\; state="\{\{state\}\}">T</figure-callout>}}_2=\frac{G_1{T}_1}{f\left(Y_1\right)G_{t\arg \mathrm{et}}}$ The device of algorithm.

In control device, include view data load module control interface module (SIF), image processing module (ISP) and processor (HOST), the device of realization time shutter computing formula algorithm can be realized by in processor (HOST) software that can realize described computing formula algorithm being installed, control interface module (SIF) reads image data from the view data load module sends image processing module (ISP) to, and reads controlled variable such as current time shutter T ₁Send processor to, processor is adjusted the controlled variable of view data load module by the control interface module.

Dispose the register of storing the luminance gain threshold value table of comparisons, the register of storing the table of comparisons of brightness logarithm value in the control device, image processing module statistical picture processing parameter such as current brightness value Y ₁And luminance gain value G ₁Deng, and sending it in the control device processor, processor will calculate adjusted Flame Image Process parameter and send image processing module back to, adjust the setting of Flame Image Process parameter.

Automatic exposure regulating method of the present invention and control system utilize the exposure algorithm formula to make the process of automatic exposure become very simple, and can be adjusted to desired value relatively quickly.Simultaneously, the parameter of required setting is considerably less, has simplified user's use.

Description of drawings

Fig. 1 shows the situation that the intensity of light source does not change in time;

Fig. 2 shows the situation of intensity of light source generating period variation in time;

Fig. 3 shows the graph of relation of sensitometric characteristic and time shutter;

Fig. 4 shows the graph of a relation of incident light and exposure;

Fig. 5 shows the primary structure of auto exposure system;

Fig. 6 shows the treatment scheme of automatic exposure regulating method.

Embodiment

Following examples are used to illustrate the present invention, but are not used for limiting the scope of the invention.

1, mathematical model

In order better to study the rule of exposure variations, find out the calculation exposure time method, the probability model that is constructed as follows:

Modelling: the table tennis box A and the B of two equal numbers are arranged, put M table tennis in the box separately, the ball in two boxes all has the sequence number of oneself and sequence number unique, and the table tennis under original state in the B all is white.We take table tennis in the A, and according to the table tennis of taking out number, with changing yellow on the corresponding ball in the B, then the table tennis of taking out are put back to A.

Ask if from A, take out K ball at every turn, repeat to take N time, total what club's yellowings in the last B?

Answer as follows:

Because the table tennis among the A is as broad as long, the probability of any one ball is identical in taking when therefore vying for the ball at every turn, all is (K/M);

If after n time, total Sn yellow ball in the B then has:

When n=0, S0=0;

During n=1, S1=K;

During n=2, S2=(M-S1) (K/M)+S1;

During n=m, Sm=(M-Sm-1) (K/M)+Sm-1;

By above derivation, we as can be known:

(1) when K=0, S _N=0;

(2) when K=M, S ₁=S ₂=...=S _N=M

(3)S _m＞＝S _m-1；

(4)S _m-S _m-1＜S _m-1

(5) when N → ∞, S _NBe substantially equal to M

Set up according to top mathematical model, the relation of view data load module incident light and exposure, as shown in Figure 4.Among Fig. 4, dot matrix A is used to simulate the response to voltage U, and dot matrix B is used for analogue exposure degree S, and the quantity of vying for the ball is used for aanalogvoltage U, and the number of times of vying for the ball is used for the analogue exposure time, then:

If each quantity unanimity of vying for the ball, then the simulation be U constant be the constant light source radiation situation;

If each quantity of vying for the ball does not require unanimity, then simulation is that the U variation is the variable light source situation.

Use the formula of mathematical model that experimental result is calculated:

Form 2

Time shutter		1	2	3	4	5	6	7	8	9
											For the first time	Test result	33	67	90	110	128	143	157	169	180
Model value	33	62	91	111	129	144	157	170	180
										For the second time		Test result	30	62	86	105	121	134	146	156	166

	Model value	30	56	85	106	123	137	148	159	168
											For the third time	Test result	37	73	99	118	133	146	159	172	183
Model value	37	69	99	122	138	151	162	173	184

Therefore can see the extraordinary test result of approaching of model, can think that this exposure model is the exposure problem that the cmos image data input module can be described.

Have by mathematical model

S _m＝(M-S _m-1)(K/M)+S _m-1 (6)

Being mapped in continuous space has

$f(x+\mathrm{\&Delta;x})=(M-f\left(x\right))\frac{f(0+\mathrm{\&Delta;x})}{M}+f\left(x\right)---\left(7\right)$

Promptly

$f^{\&prime;}\left(x\right)=\frac{M-f\left(x\right)}{M}{f}^{\&prime;}\left(0\right)---\left(8\right)$

Make K=f ' (0), have

$f^{\&prime;}\left(x\right)+\frac{K}{M}f\left(x\right)=K---\left(9\right)$

Separating this differential equation obtains

$f\left(x\right)=M(1-e^{-\frac{K}{M}x})---\left(10\right)$

F (0)=0 wherein; F ' (0)=K

2, exposure compensating

Under indoor situation, the setting of time shutter is necessary for the integral multiple of indoor light source period of change T to avoid the generation of flicker phenomenon, and this has just caused another problem: the uncontinuity that the time shutter is set.The time shutter of She Zhiing is not the integral multiple (in fact most applications all is like this) of T just if desired, and that just can not obtain the brightness of image that we need.For this reason, introduced another parameter: luminance gain value G (Gain).

Luminance gain value G is the value that acts on the brightness of image curve, and the amplitude of the luminance component by regulating image can change the chiaroscuro effect of image.But luminance gain is regulated and can not be increased the entrained quantity of information of image, if therefore undue adjusting luminance gain value can cause image quality decrease.

The present invention adopts the method for regulating time shutter and luminance gain value simultaneously to carry out exposure compensating, guarantees the continuity of brightness regulation by the luminance gain value.

By the formula that Derivation of Mathematical Model goes out, the method that the present invention can obtain calculation exposure time and luminance gain is as follows:

Suppose:

The brightness maximal value is M, and this maximal value is by the data processing figure place decision of image processing module, and making M=256 here is constant;

The current time shutter is T ₁, luminance gain is G ₁, brightness is Y ₁

If change luminance gain into G ₂(object brightness gain), the brightness degree that image is obtained in hope is Y ₂, time shutter T then ₂Computation process is:

${\mathrm{<figure-callout\; id="1"\; label="Y"\; filenames="A20051008032700231.png"\; state="\{\{state\}\}">Y</figure-callout>}}_1=M(1-e^{-\frac{K{G}_1{T}_1}{M}})---\left(11\right)$

$\ln (M-Y_1)-\ln \left(M\right)=-\frac{K{G}_1{T}_1}{M}---\left(12\right)$

In like manner have

$\ln (M-Y_2)-\ln \left(M\right)=-\frac{K{G}_2{T}_2}{M}---\left(13\right)$

$\frac{\ln (M-Y_1)-\ln \left(M\right)}{\ln (M-Y_2)-\ln \left(M\right)}=\frac{G_1{T}_1}{G_2{G}_2}---\left(14\right)$

In actual applications, generally have the depth of exposure value (image brightness is suitable) and the luminance gain value (be generally 1, promptly the luminance gain value is not had an effect) of an expectation, therefore can determine a target exposure intensity Y _TargetWith object brightness yield value G _Target, (14) revised obtain:

$\frac{\ln (M-Y_1)-\ln \left(M\right)}{\ln (M-Y_{t\arg \mathrm{et}})-\ln \left(M\right)}=\frac{G_1{T}_1}{G_{t\arg \mathrm{et}}{T}_2}---\left(15\right)$

As seen, has only a unknown number T in the formula ₂, therefore can solve T ₂Value:

Order

$f\left(Y_1\right)=\frac{\ln (M-Y_1)-\ln \left(M\right)}{\ln (M-Y_{t\arg \mathrm{et}})-\ln \left(M\right)}---\left(16\right)$

Can obtain

$f\left(Y_1\right)=\frac{G_1{T}_1}{G_{t\arg \mathrm{et}}{\mathrm{<figure-callout\; id="2"\; label="T"\; filenames="A20051008032700231.png"\; state="\{\{state\}\}">T</figure-callout>}}_2}\&RightArrow;{\mathrm{<figure-callout\; id="2"\; label="T"\; filenames="A20051008032700231.png"\; state="\{\{state\}\}">T</figure-callout>}}_2=\frac{G_1{T}_1}{f\left(Y_1\right)G_{t\arg \mathrm{et}}}---\left(17\right)$

Obtained the mathematical method of calculation exposure time of the present invention thus.

3, method realizes

Automatic exposure control system realizes the scheme of automatic exposure control: utilize register to read the brightness value of the image of image processing module statistics, processor produced and interrupts when the brightness value that counts at image processing module exceeded the brightness section of setting, and system carries out automatic exposure and regulates; Target exposure intensity, object brightness yield value, luminance gain variation range are set, obtain parameters such as current luminance gain value, exposure intensity value (brightness statistics value) and time shutter, and utilize the formula of calculation exposure time to calculate the time shutter that makes new advances; Write the register of view data load module the time shutter that calculates by iic bus or Serials Bus communication mode.

A kind of specific embodiment of the present invention provides a kind of system that utilizes installed software to realize automatic exposure (AE) algorithm, and wherein this system comprises: the view data load module, can revise view data load module time shutter T ₁With image processing module luminance gain value G ₁Register, be provided with object brightness degree value Y _TargetProduce the luminance threshold Y that interrupts with automatic exposure _ThdRegister, and the software of automatic exposure algorithm; The workflow of automatic exposure control system as shown in Figure 6.After exposure produces, the current brightness value Y of systematic comparison ₁With can accept brightness section [Y _Target-Y _Thd, Y _Target+ Y _Thd], when current brightness value exceeds brightness section, producing and interrupt, auto exposure system carries out automatic exposure to be regulated;

Exposure intensity and luminance gain adjustment are depended in the brightness that obtains image, and exposure intensity is then by the time shutter decision, and therefore, brightness of image is by time shutter and the decision of luminance gain value.Can judgement luminance gain compensation earlier be adjusted to brightness and can accept brightness section:

[Y _target-Y _thd，Y _target+Y _thd]，

Utilize formula

${\mathrm{<figure-callout\; id="2"\; label="G"\; filenames="A20051008032700231.png"\; state="\{\{state\}\}">G</figure-callout>}}_2=\frac{\ln (M-Y_{t\arg \mathrm{et}})-\ln \left(M\right)}{\ln (M-Y_1)-\ln \left(M\right)}\&times;{\mathrm{<figure-callout\; id="1"\; label="G"\; filenames="A20051008032700231.png"\; state="\{\{state\}\}">G</figure-callout>}}_1$

Calculate required luminance gain value, M is the brightness maximal value in the formula, judges whether G _Target-G _Thd.n≤ G ₂≤ G _Target+ G _Thd.nIf,, then with luminance gain value G ₁Be adjusted into G ₂Value, and continue later exposure, if not,

Then carrying out the time shutter adjusts: the formula that utilizes the calculation exposure time $T_{t\arg \mathrm{et}}=\frac{G_1{T}_1}{f\left(Y_1\right)G_{t\arg \mathrm{et}}}$ And $f\left(Y_1\right)=\frac{\ln (M-Y_1)-\ln \left(M\right)}{\ln (M-Y_{t\arg \mathrm{et}})-\ln \left(M\right)}$ Calculate the target time shutter, T in the formula _TargetBe at target time shutter, G ₁Be current luminance gain, T ₁Be the current time shutter, Y ₁Be current brightness.With the target time shutter T that calculates _TargetAdjust time shutter T ₁Be provided with.

When light source is indoor light, because of intensity of illumination can be to change in the cycle with T, for avoiding the flicker phenomenon, if the target time shutter T that calculates _TargetFor intensity of illumination period of change time T integral multiple, then use time T _TargetAs optimum time shutter T ₂Adjust time shutter T ₁Be provided with, if the target time shutter T that calculates _TargetNot the integral multiple of intensity of illumination period of change time T, then use near T _TargetIntegral multiple time of T as optimum time shutter T ₂Adjust time shutter T ₁Be provided with.

Because the optimum time shutter of selecting above is not target time shutter T _Target(time shutter that object brightness requires) is so calculating optimum time shutter T ₂After carry out following steps again and adjust the luminance gain value and carry out exposure compensating:

Use formula $G_{}$ ${}_2=\frac{G_1{T}_1}{f\left(Y_1\right){\mathrm{<figure-callout\; id="2"\; label="T"\; filenames="A20051008032700231.png"\; state="\{\{state\}\}">T</figure-callout>}}_2}$ Calculate optimum luminance gain value G ₂, and adjust luminance gain value G in the image processing module with result of calculation ₁

Make when avoiding adjusting the luminance gain value brightness value change the discontinuous flicker that causes, earlier according to the luminance gain threshold value G of initial setting up _Thd, utilize formula G _Thdn+1=(G _Target-n * G _Thd.n) the luminance gain threshold value G that allows when calculating different exposure time of ÷ (n+1) _Thdn, and result of calculation is stored in the luminance gain threshold value table of comparisons, wherein the n time shutter is to the multiple of intensity of illumination period of change time T, G _Thd1=G _Thd, calculating G ₂After, according to time shutter T ₂Read corresponding luminance gain value threshold value G in the luminance gain threshold value table of comparisons _Thd, judge whether G earlier ₂At G _TargetIn the scope that allows: if G ₂＞G _Target+ G _Thd.n, then get G ₂=G _Target+ G _Thd.nIf G ₂＜G _Target-G _Thdn, then get G ₂=G _Target-G _Thd.nIf G _Target-G _Thd.n≤ G ₂≤ G _Target+ G _Thdn, then do not revise G ₂And then use G ₂Adjust the luminance gain value G in the image processing module ₁

Because the dirigibility when algorithm requirement and user use is considered, need dispose following parameter according to actual conditions in system by the user:

Y _Target: object brightness;

Y _Thd: the brightness change threshold of permission is top said interruption and produces interval;

G: the original intensity yield value can be set to object brightness yield value G _Target

G _Thd: acceptable luminance gain value change threshold;

G _Delay: Sensor might come into force behind several frames after having revised the time shutter, and this value of different Sensor is also different, G ₁Be provided with and need come into force simultaneously with the time shutter, dispose G herein ₁Came into force every a frame as the time shutter of sensor configuration in the position of coming into force, then G herein _DelayValue is made as 1

Maxet: the maximum exposure time that can be provided with since the time shutter be increased to certain value and can cause frame per second slack-off, so should limit the maximal value of setting.

Minet: the minimum exposure time is set.

Maxgain: when the time shutter has reached maximal value and still can not meet the demands,

G ₁The setting of value can exceed (G _Target-G _Thd, G _Target+ G _Thd) scope, maximal value can be made as Maxgain.

Mingain: when time shutter that calculates during, at first will be reduced to a T time shutter, and reduce G then less than a T ₁Value can arrive the scope that Mingain is provided with, if still can not satisfy, reduces the time shutter again to less than a T.

Speed: 4 values (0,1,2,3) can be set.When needs are adjusted the time shutter, the new value that might calculate and original be set with very big variation, if directly will adjust to the value that calculates the time shutter, might image can produce rank that apparent in view light and shade changes more, wish the speed adjusted, speed 0＜1＜2＜3 so can set the user herein.

When bigger variation takes place in environment (such as switching to brighter environment suddenly) by darker environment, since algorithm of the present invention can be very fast the exposure time values that obtains target, if calculated value directly is provided with, can make image have a bigger light and shade to change, effect is got on just so-called rank.This may make user's incompatibility, and therefore, the present invention can avoid this situation by parameter s peed value is set in actual applications.According to user's actual demand, can change this value to obtain different effects.For example, the current time shutter is n, and calculating the new time shutter of gained is m, when ambient brightness changes when violent, | m-n| can be a bigger value (surpassing 4 times of T), if we directly are provided with m, will produce rank and get over effect.In our implementation, be provided with 4 grades of governing speeds.Speed 0＜1＜2＜3.When being set to different governing speeds, with | the amplitude of variation of m-n|/s changes the time shutter.Time shutter promptly be set be n ± | m-n|/s.Governing speed is fast more, and the value of s is more little.No matter why speed setting is worth, and when being 4 times of T of less than as if the time shutter adjusting range, the mode of adjusting the time shutter is each T of adjustment.

Because the time shutter is at the indoor integral multiple that must be set to 50HZ/60HZ, then there is not this restriction outdoor, when indoor and outdoors switches when environment for use can go wrong.For example: when indoor light is more intense, the time shutter that is calculated might be less than a T, owing to be difficult to determine current environment (environment is variable at any time), so being difficult to decision is the time shutter to be set to calculated value, or be set to 1 T value, therefore, some model selections be can increase, the current environment and the effect of exposure determined by the user.The selection of pattern can or be revised the parameter value of listing above and finish fully by configuration.

Below be each mode declaration pattern specification, and the value of setting of recommending:

Auto

Each parameter setting values is: Minet=0, Maxet=0xe0, Y _Top=0xe0 is when time shutter that calculates during less than a T, if this moment Y ₁Value is less than y _TopValue is not then adjusted the time shutter setting, otherwise adjusts the time shutter setting.Under this pattern when indoor environment is brighter flicker might appear.

Outdoor

Each parameter setting values is: Minet=0, Maxet=0xe0, Y _Top=Y _Target+ Y _ThdWhen time shutter that calculates during, be set to calculated value less than 1T.Under this pattern

Flicker appears if be in indoor meeting.

Indoor

Each parameter setting values is: Minet=1, Maxet=0xe0, Y _TopNo matter how indoor environment flicker can not occur under this pattern of=0xe0, but when indoor light was brighter, the Y value may not reach best effects.

Dark

Each parameter setting values is: Minet=5, the minimum exposure time is 5T under this pattern of Maxet=255, may cause image shinny when being in brighter environment.

4, the basic hardware architecture of auto exposure system

Fig. 5 is the basic hardware architecture that an auto exposure system is used, mainly comprise 4 parts: view data load module (Sensor), view data output module (LCD), picture processing chip (Chip), processor (Host), the control device in picture processing chip (Chip) and processor (Host) the composition system.This also is a image processing system the most basic.View data load module (sensor) is accepted the external image signal, through the conversion of simulating signal to digital signal, imports the numerical data of changing into picture processing chip and is shown on the LCD display after treatment.In this system, picture processing chip is a most important parts, the quality of his processing power decision entire image processing system.

The auto exposure system of present embodiment can be applied among the picture processing chip VC0568 of a special use, and he can finish the processing to input picture, shows and storage.Hardware module relevant with auto exposure system in it has just been described in structure of last figure.

As can be seen from Figure 5, chip comprises two interface modules, finishes mutual with view data load module and output device respectively.The interface module SIF of major concern of the present invention and view data load module (Sensor), all are all finished by interface module SIF the operation of view data load module.The communication protocol of view data load module and interface module mainly adopts IIC or Serial bus, the value that can write the register of view data load module writes the particular register of interface module, is responsible for these data are write Sensor by interface module then.The ISP module is an image processing module, and analysis and processing to image are provided, and can offer the essential information of our image.Processor is generally a microprocessor (the present invention adopts the Arm7TDMI of Samsung).Software of the present invention (comprising the automatic exposure algorithm) all moves on processor.Below be the processing procedure of auto exposure system:

At first, picture processing chip obtains data, and the data of each width of cloth image are handled and added up, and has disposed two registers in the ISP module, as the Y that go out of ISP module to brightness (Y) the primary system meter of present image ₁When value exceeded the scope of configuration, chip sent automatic exposure and interrupts, and processor calls registration is interrupted in the present invention to automatic exposure when initialization interrupt handling routine (ISR AUTOEXPOSE).

Secondly, in interrupt handling routine, we obtain by the related register that reads ISP and carry out automatic exposure calculating parameters needed (as Y ₁, G ₁, G _TargetDeng), calculate the time shutter T that makes new advances by formula ₂And G ₂Value.

At last, processor Host is by the register of configuration interface module SIF, and the new time shutter that will dispose writes view data load module Sensor by interface module SIF, new luminance gain value G ₂The G of the ISP module of writing direct ₁Value register.

Though disclose the present invention with preferred example embodiment of the present invention, those skilled in the art should understand that do not breaking away under the scope of the invention, can carry out multiple other variation, omission and change to its process and details.

Claims (16)

1, a kind of method of automatic exposure adjusting is characterized in that, comprises following steps:

(1) sets exposure control basic parameter, determine the target brightness value scope;

(2) after exposure and extracting the processing image, judge current brightness value Y ₁Whether in the target brightness value scope, if, continue by original parameter exposure and handle image, if not, then carry out following automatic exposure regulating step;

(3) calculate required luminance gain value G ₂, judge whether to utilize the adjustment of luminance gain value brightness value can be adjusted in the target brightness value scope; If, then with luminance gain value G ₁Be adjusted into G ₂Value, handle image by new luminance gain value again, if not, carry out next step;

(4) calculate optimum time shutter T2, adjust time shutter T1 with optimum time shutter T2.

2, the method for claim 1 is characterized in that,

The exposure control basic parameter of setting in step (1) comprises object brightness yield value G _Target, luminance gain threshold value G _Thd, and according to the luminance gain threshold value G of initial setting up _Thd, utilize formula G _Thd.n+1=(G _Target-n * G _Thd.n) the luminance gain threshold value G that allows when calculating different exposure time of ÷ (n+1)+offset _Thd.n, and result of calculation is stored in the luminance gain threshold value table of comparisons, wherein n is the multiple of time shutter to intensity of illumination period of change time T, offset is an arbitrary integer, G _Thd.1=G _Thd

3, method as claimed in claim 1 or 2 is characterized in that: the exposure control basic parameter of setting in step (1) comprises object brightness standard value Y _Target, luminance threshold Y _Thd, described definite target brightness value scope is: [Y _Target-Y _Thd, Y _Target+ Y _Thd].

4, method as claimed in claim 3 is characterized in that,

In step (3), the described concrete grammar that utilizes the adjustment of luminance gain value brightness value can be adjusted in the target brightness value scope that judges whether is:

(3.1) utilize formula $G_2=\frac{\ln (M-Y_{t\arg \mathrm{et}})-\ln \left(M\right)}{\ln (M-Y_1)-\ln \left(M\right)}\&times;G_1$ Calculate required luminance gain value G ₂, M is the brightness range maximal value in the formula;

(3.2) judge whether G _Target-G _Thd.n≤ G ₂≤ G _Target+ G _Thd.n

5, method as claimed in claim 3 is characterized in that, in step (4), and the optimum time shutter T of described calculating ₂Be to utilize formula $f\left(Y_1\right)=\frac{\ln (M-Y_1)-\ln \left(M\right)}{\ln (M-Y_{t\arg \mathrm{et}})-\ln \left(M\right)}$ Calculate f (Y ₁); Utilize formula then $T_{t\arg \mathrm{et}}=\frac{G_1{T}_1}{f\left(Y_1\right)G_{t\arg \mathrm{et}}}$ Calculate target time shutter T _Target, get T _TargetBe optimum time shutter T ₂

6, method as claimed in claim 3 is characterized in that,

Also different brightness and logarithm value thereof are stored in the table of comparisons in step (1), brightness range maximal value M is decided to be constant;

Calculating target time shutter T described in the step (4) _TargetMay further comprise the steps:

(4.1) read Y ₁, M-Y ₁, M-Y _TargetAnd the logarithm value of M in the table of comparisons;

(4.2) utilize formula $f\left(Y_1\right)=\frac{\ln (M-Y_1)-\ln \left(M\right)}{\ln (M-Y_{t\arg \mathrm{et}})-\ln \left(M\right)}$ Calculate f (Y ₁);

(4.3) utilize formula $T_{t\arg \mathrm{et}}=\frac{G_1{T}_1}{f\left(Y_1\right)G_{t\arg \mathrm{et}}}$ Calculate T _Target

7, method as claimed in claim 6 is characterized in that,

In step (1), also be provided with intensity of illumination period of change time T;

In step (4), utilize formula to calculate target time shutter T _TargetAfter also carry out following processing: if the target time shutter T that calculates _TargetFor intensity of illumination period of change time T integral multiple, then get time T _TargetBe T ₂, if the target time shutter T that calculates _TargetBe not the integral multiple of intensity of illumination period of change time T, the time of then getting is near T _TargetThe integer multiple of T be T ₂

8, method as claimed in claim 7 is characterized in that, is calculating optimum time shutter T ₂After also carry out following steps:

(5) use formula $G_2=\frac{G_1{T}_1}{f\left(Y_1\right)T_2}$ Calculate optimum luminance gain value G ₂, and adjust luminance gain value G with result of calculation ₁

9, method as claimed in claim 8 is characterized in that,

In step (5), calculate optimum luminance gain value G ₂After, according to time shutter T ₂Read corresponding luminance gain value threshold value G in the luminance gain threshold value table of comparisons _Thd.n, judge whether G earlier ₂At G _TargetIn the scope that allows: if G ₂＞G _Target+ G _Thd.n, then get G ₂=G _Target+ G _Thd.nIf G ₂＜G _Target-G _Thd.n, then get G ₂=G _Target-G _Thd.nIf G _Target-G _Thd.n≤ G ₂≤ G _Target+ G _Thd.n, then do not revise G ₂And then use G ₂Adjust luminance gain value G ₁

10, method as claimed in claim 3 is characterized in that, also is set with the luminance gain value adjustment of determining at the different model image inductor delay time G that comes into force in the step (1) _Delay

11, method as claimed in claim 3 is characterized in that, following parameter is set:

Maximum exposure time Maxet is used to prevent time shutter T ₂Be increased to certain value and can cause frame per second slack-off;

Minimum exposure time Minet;

High-high brightness yield value Maxgain is used for when the time shutter has reached maximal value Maxet and still can not meet the demands, and the setting of yield value can exceed the scope of luminance gain threshold value, and maximal value can be made as high-high brightness yield value Maxgain;

Minimum brightness yield value Mingain, be used at first will being reduced to a T time shutter, reduce the scope that the luminance gain value is provided with to minimum gain value Mingain then when time shutter that calculates during less than a T, if still can not satisfy, reduce the time shutter again to less than a T.

12, the described method of claim 11, it is characterized in that, a speed control parameter speed also is set in step (1), and when time shutter that needs are adjusted and current time shutter gap during greater than 4 times of intensity of illumination period of change time T, the time shutter is progressively adjusted in segmentation.

13, the method for the adjusting of the automatic exposure in the claim 11: increase model selection, determine the current environment and the effect of exposure by the user, and finish pattern by the parameter value that configuration or modification are listed above and select, model selection is by any matched combined of automatic mode, outdoor pattern, indoor mode or Night and this several modes;

Described automatic mode, its parameter setting values is: Minet=0, Maxet=0xe0, Y _Top=0xe0; When time shutter that calculates during less than a T, if this moment y ₁Value is less than y _TopValue is not then adjusted the time shutter setting, otherwise adjusts the time shutter setting;

Described outdoor pattern, its parameter setting values is: Minet=0, Maxet=0xe0, Y _Top=y _Target+ y _Thd,, be set to calculated value T2 as the time shutter T2 that calculates during less than T.

Described indoor mode, its parameter setting values is: Minet=1, Maxet=0xe0, y _Top=0xe0 avoids occurring the flicker phenomenon under indoor environment;

Described Night, its parameter setting values is: Minet=5, Maxet=255, the minimum exposure time is 5T under this pattern.

14, a kind of automatic exposure control system of using above-mentioned automatic exposure control method, comprise view data load module (sensor), control device, view data output module (LCD), it is characterized in that, is furnished with the register that to revise view data load module time shutter and/or Flame Image Process luminance gain value in the control device, and be provided with register that target brightness value and automatic exposure produce the brightness change threshold that interrupts and the formula of realizing the calculation exposure time $f\left(Y_1\right)=\frac{\ln (M-Y_1)-\ln \left(M\right)}{\ln (M-Y_{t\arg \mathrm{et}})-\ln \left(M\right)}$ And $T_2=\frac{G_1{T}_1}{f\left(Y_1\right)G_{t\arg \mathrm{et}}}$ The device of algorithm, wherein M is a maximum brightness value, Y ₁Be current brightness value, Y _TargetBe object brightness standard value, T ₂Be optimum time shutter, T ₁Be current time shutter, G ₁Be luminance gain value, G _TargetBe the object brightness yield value.

15, the described automatic exposure control system of claim 14, it is characterized in that, include view data load module control interface module (SIF), image processing module (ISP) and processor (HOST) in the control device, in the processor (HOST) device that can realize described computing formula algorithm is installed, control interface module (SIF) reads image data from the view data load module sends image processing module (ISP) to, and read controlled variable and send processor to, processor is adjusted the controlled variable of view data load module by the control interface module.

16, claim 14 or 15 described automatic control Exposure Control Modules, it is characterized in that, dispose the register of storing the luminance gain threshold value table of comparisons, the register of storing the table of comparisons of brightness logarithm value in the control device, image processing module statistical picture processing parameter also sends it in the control device processor, processor will calculate adjusted Flame Image Process parameter and send image processing module back to, adjust the setting of Flame Image Process parameter.

Images