CN103546732B - Image processing method and system - Google Patents

Abstract

The invention provides a kind of image processing method and system, comprise automatic explosion method, described automatic explosion method following steps: brightness statistics step: image is divided into multiple range statistics mean flow rates; Adjust exposure determining step: judge whether to adjust exposure according to described reference brightness and object brightness; Obtain the overall gain step of setting: ratio and current overall gain draw the overall gain of setting with exposing; Obtaining step: use the overall gain and the exposure Policy Table that arrange to draw time for exposure and gain; Step of exposure: the time for exposure in obtaining step and gain are set in imageing sensor and are exposed. The invention has the beneficial effects as follows that automatic explosion method of the present invention and unit have improved the unsettled problem of exposure well, by the different exposure strategy of customization, the image effect under varying environment is reached to optimum simultaneously.

Description

Image processing method and system

Technical field

The present invention relates to camera technique field, relate in particular to image processing method and system.

Background technology

Exposure is the process of the light line formed image that receives into of imageing sensor, and the time for exposure is that imageing sensor receivesThe time of light. In shooting process, extraneous light can change, and easily causes over-exposedly when light grow, causes imageCross brightly, when light dies down, easily cause under-exposure, cause image excessively dark. Therefore just need in shooting process, carry out exposureControl, wherein automatic exposure is exactly without manual operations, by automatic adjustment brightness, time for exposure, gain, image is obtainedSuitable exposure and brightness.

In current technology, expose unstable, and be subject to the impact of external environment condition.

Summary of the invention

In order to solve the problems of the prior art, the invention provides a kind of image processing method and system.

The invention provides a kind of image processing method, comprise automatic explosion method, described automatic explosion method walks as followsRapid:

Brightness statistics step: image is divided into multiple range statistics mean flow rates, by each region according to weight table middle fingerFixed weight calculation goes out the mean flow rate of whole two field picture, and the mean flow rate of adding up nearest some two field pictures is as with reference to brightness;

Adjust exposure determining step: judge whether to adjust exposure according to described reference brightness and object brightness, if will adjustWhole exposure, carries out the overall gain step of obtaining setting;

Obtain the overall gain step of setting: draw the overall gain of setting by exposure ratio and current overall gain, arrangeThe current overall gain of overall gain=exposure ratio *;

Obtaining step: use the overall gain and the exposure Policy Table that arrange to draw time for exposure and gain;

Step of exposure: the time for exposure in obtaining step and gain are set in imageing sensor and are exposed.

As a further improvement on the present invention, in described adjustment exposure determining step: if reference brightness is bright in targetIn the higher limit of degree and the scope of lower limit, do not need to adjust exposure; If reference brightness exceeds the higher limit of object brightnessWith the scope of lower limit, need to adjust exposure.

As a further improvement on the present invention, in the described overall gain step of obtaining setting:

In described exposure ratio, adjust exposure ratio according to reference brightness and object brightness phase extent, exposureRatio=F (abs(object brightness-reference brightness));

In current overall gain, be first exposure gain by the time for exposure divided by exposure gain Conversion of measurement unit, then willExposure gain * gain obtains current overall gain.

As a further improvement on the present invention, in described obtaining step, the project bag of described exposure Policy Table's list itemDraw together:

Maximum exposure gain, the maximum exposure that can reach for this section of exposure gain;

Maximum gain, the maximum gain that can reach for this section of exposure;

Maximum overall gain, the maximum overall gain that can reach for this section of exposure, maximum overall gain=maximum exposure increasesBenefit * maximum gain;

Exposure mode, for this section of exposure mode that exposure is used, is divided into two kinds of exposure gain mode and gain modes;Exposure gain mode, gain keeps maximum gain constant, and exposure gain changes between the gain of 0～maximum exposure; Gain mode,Time for exposure gain keeps maximum exposure gain constant, and gain changes between 1～maximum gain;

In described obtaining step, comprise execution following steps:

First from exposure Policy Table, search the list item that first maximum overall gain is more than or equal to the overall gain that will arrange;

Judge again the exposure mode of this list item searching: the gain mode that exposes if, gain=maximum gain,Gain=overall gain/gain exposes completely; Gain mode if, exposure gain=maximum gain, gain=overall gain/exposure increaseBenefit;

Then, the gain conversions of exposing is the time for exposure, time for exposure=exposure gain * exposure gain unit.

As a further improvement on the present invention, this image processing method also comprises auto white balance method, described automatically whiteBalance method comprises execution following steps:

A.R, G, B white balance statistics, for counting green red ratio G/R and the turquoise ratio G/R of present image;

B. estimate current colour temperature according to statistical information, for according to statistics green red ratio G/R and turquoise ratio G/R and lookTemperature estimation function, estimates current colour temperature;

C. gain and color correction matrix according to the colour temperature of estimating and the red indigo plant of white balance corresponding to high, medium and low three colour temperatures,Interpolation goes out the red blue gain of current white balance and color correction matrix;

D., the red blue gain of white balance and color correction matrix are set.

The present invention also provides a kind of image processing system, comprises automatic exposure unit, and described automatic exposure unit comprises:

Brightness statistic: for image being divided into multiple range statistics mean flow rates, by each region according to weight tableThe weight calculation of middle appointment goes out the mean flow rate of whole two field picture, and the mean flow rate of adding up nearest some two field pictures is bright as a referenceDegree;

Adjust exposure judge module: for judging whether to adjust exposure according to described reference brightness and object brightness, ifAdjust exposure, carry out the overall gain module of obtaining setting;

Obtain the overall gain module of setting: draw the overall gain of setting by exposure ratio and current overall gain, arrangeThe current overall gain of overall gain=exposure ratio *;

Acquisition module: draw time for exposure and gain for the overall gain and the exposure Policy Table that use setting;

Exposure module: for the time for exposure of acquisition module and gain are set in imageing sensor and are exposed.

As a further improvement on the present invention, in described adjustment exposure judge module: if reference brightness is bright in targetIn the higher limit of degree and the scope of lower limit, do not need to adjust exposure; If reference brightness exceeds the higher limit of object brightnessWith the scope of lower limit, need to adjust exposure.

As a further improvement on the present invention, in the described overall gain module of obtaining setting:

In described exposure ratio, adjust exposure ratio according to reference brightness and object brightness phase extent, exposureRatio=F (abs(object brightness-reference brightness));

In current overall gain, be first exposure gain by the time for exposure divided by exposure gain Conversion of measurement unit, then willExposure gain * gain obtains current overall gain.

As a further improvement on the present invention, in described acquisition module, the project bag of described exposure Policy Table's list itemDraw together:

Maximum exposure gain, the maximum exposure that can reach for this section of exposure gain;

Maximum gain, the maximum gain that can reach for this section of exposure;

Maximum overall gain, the maximum overall gain that can reach for this section of exposure, maximum overall gain=maximum exposure increasesBenefit * maximum gain;

Exposure mode, for this section of exposure mode that exposure is used, is divided into two kinds of exposure gain mode and gain modes;Exposure gain mode, gain keeps maximum gain constant, and exposure gain changes between the gain of 0～maximum exposure; Gain mode,Time for exposure gain keeps maximum exposure gain constant, and gain changes between 1～maximum gain;

In described acquisition module, comprising:

Search module, is more than or equal to total increasing that will arrange for searching first maximum overall gain from exposure Policy TableThe list item of benefit;

Judge module, for judging the exposure mode of this list item searching: the gain mode that exposes if, gain=Maximum gain, maximum exposure gain=overall gain/gain; Gain mode if, exposure gain=maximum gain, gain=totalGain/exposure gain;

Modular converter, is the time for exposure for the gain conversions of exposing, time for exposure=exposure gain * exposure gain unit.

As a further improvement on the present invention, this image processing system also comprises AWB unit, described automatically whiteBalancing unit comprises:

White balance statistics module, for R, G, B white balance statistics, counts the green red ratio G/R of present image and turquoiseRatio G/R;

Estimate current colour temperature module, for estimate current colour temperature according to statistical information, for according to statistics green red ratioG/R and turquoise ratio G/R and color temperature estimation function, estimate current colour temperature;

Interpolating module, for gaining and face according to the colour temperature of estimating and the red indigo plant of white balance corresponding to high, medium and low three colour temperaturesLook correction matrix, interpolation goes out the red blue gain of current white balance and color correction matrix;

Module is set, for the red blue gain of white balance and color correction matrix are set.

The invention has the beneficial effects as follows: it is unsettled that automatic explosion method of the present invention and unit have improved exposure wellProblem reaches optimum by the different exposure strategy of customization the image effect under varying environment simultaneously.

Brief description of the drawings

Fig. 1 is the schematic diagram of automatic explosion method and auto white balance method in the present invention.

Fig. 2 is auto white balance method flow chart of the present invention.

Detailed description of the invention

The invention discloses a kind of image processing method, comprise automatic explosion method and auto white balance method. As Fig. 1 instituteShow, the schematic diagram of automatic explosion method and auto white balance method in the present invention, automatic exposure module obtains from imageing sensorGraph data carries out automatic exposure, and then the time for exposure calculating and gain are established into imageing sensor and exposed, fromMoving white balance module obtains graph data from imageing sensor and carries out AWB.

Described automatic explosion method following steps:

Brightness statistics step: image is divided into multiple range statistics mean flow rates, by each region according to weight table middle fingerFixed weight calculation goes out the mean flow rate of whole two field picture, and the mean flow rate of adding up nearest some two field pictures is as with reference to brightness;

Adjust exposure determining step: judge whether to adjust exposure according to described reference brightness and object brightness, if will adjustWhole exposure, carries out the overall gain step of obtaining setting;

Obtain the overall gain step of setting: draw the overall gain of setting by exposure ratio and current overall gain, arrangeThe current overall gain of overall gain=exposure ratio *;

Obtaining step: use the overall gain and the exposure Policy Table that arrange to draw time for exposure and gain;

Step of exposure: the time for exposure in obtaining step and gain are set in imageing sensor and are exposed.

In described adjustment exposure determining step: if reference brightness is in the higher limit of object brightness and the scope of lower limitIn, do not need to adjust exposure; If reference brightness exceeds the higher limit of object brightness and the scope of lower limit, need to adjustExposure.

In the described overall gain step of obtaining setting:

In described exposure ratio, adjust exposure ratio according to reference brightness and object brightness phase extent, exposureRatio=F (abs(object brightness-reference brightness));

In current overall gain, be first exposure gain by the time for exposure divided by exposure gain Conversion of measurement unit, then willExposure gain * gain obtains current overall gain.

In described obtaining step, the project of described exposure Policy Table's list item comprises:

Maximum exposure gain, the maximum exposure that can reach for this section of exposure gain;

Maximum gain, the maximum gain that can reach for this section of exposure;

Maximum overall gain, the maximum overall gain that can reach for this section of exposure, maximum overall gain=maximum exposure increasesBenefit * maximum gain;

Exposure mode, for this section of exposure mode that exposure is used, is divided into two kinds of exposure gain mode and gain modes;Exposure gain mode, gain keeps maximum gain constant, and exposure gain changes between the gain of 0～maximum exposure; Gain mode,Time for exposure gain keeps maximum exposure gain constant, and gain changes between 1～maximum gain;

In described obtaining step, comprise execution following steps:

First from exposure Policy Table, search the list item that first maximum overall gain is more than or equal to the overall gain that will arrange;

Judge again the exposure mode of this list item searching: the gain mode that exposes if, gain=maximum gain,Gain=overall gain/gain exposes completely; Gain mode if, exposure gain=maximum gain, gain=overall gain/exposure increaseBenefit;

Then, the gain conversions of exposing is the time for exposure, time for exposure=exposure gain * exposure gain unit.

In automatic explosion method of the present invention, be a variable time for exposure, gain unification, be called overall gain; In advanceFirst customize the exposure Policy Table under varying environment; Image is divided into multiple range statistics mean flow rates; By each region according to powerThe weight calculation of heavily specifying in table the mean flow rate of whole two field picture; Add up the mean flow rate of nearest some two field pictures as a referenceBrightness; Judge whether to adjust exposure according to reference brightness and object brightness; What differ according to reference brightness and object brightness is largeThe little exposure ratio of adjusting; Draw the overall gain that will arrange with described exposure ratio and current overall gain; Use will be establishedThe exposure Policy Table of the overall gain of putting and appointment draws time for exposure, the gain that will arrange; The time for exposure obtaining, gain are establishedEntering imageing sensor exposes.

Be an overall gain time for exposure, gain unification, method is:

First be exposure gain: ExposureGain=by the time for exposure divided by exposure gain Conversion of measurement unitExposureTime/ExposureGainUnit;

Then the gain that will gain and expose is unified is overall gain: TotalGain=ExposureGain*Gain.

The exposure gain ExposureGainUnit of unit can arbitrarily specify a time for exposure, such as, 10 milliseconds.

Customize in advance the exposure Policy Table under varying environment, each list item of exposure Policy Table has represented one section of exposure placeReason, is divided into multistage exposure processing with an exposure Policy Table an exposure strategy and stores.

Customize different exposure Policy Tables according to different environment in advance, such as, for indoor environment, due to lighting sourcePeriodically changed by the light intensity that affects of 50Hz/60Hz alternating current, for fear of flicker, need to be by time for exposure furnishing light sourceThe integral multiple of flicker cycle, thereby the exposure strategy of customization indoor mode, for outdoor environment, owing to there not being lighting source,Substantially be all the natural daylight of stabilized intensity, to be that outdoor environment brightness of image is excessive cause over-exposedly more problem, should for this reasonThis allows the time for exposure can be short as much as possible, thereby customizes the exposure strategy of outdoor pattern.

The project of exposure Policy Table's list item, comprising:

Maximum exposure gain (MaxExposureGain): the maximum exposure gain that can reach in this section of exposure;

Maximum gain (MaxGain): the maximum gain that can reach in this section of exposure;

Maximum overall gain (MaxTotalGain): the maximum overall gain that can reach in this section of exposure, it equalsMaxExposureGain*MaxGain; List item in table is arranged from small to large according to the size of maximum overall gain;

Exposure mode (ExposureMode): exposure mode used in this section of exposure, is divided into exposure gain mode and increasingTwo kinds of benefit patterns;

Exposure gain mode, gain keeps MaxGain constant, and exposure gain becomes between 0～MaxExposureGainChange;

Gain mode, time for exposure gain keeps MaxExposureGain constant, and gain becomes between 1～MaxGainChange.

Image is divided into multiple range statistics mean flow rates, and region number can arbitrarily be divided, and is generally no less than 5x5. WillEach region the mean flow rate of whole two field picture according to the weight calculation of specifying in weight table. Add up the flat of nearest some two field picturesAll brightness is as with reference to brightness. In weight table, can want to see the larger weight in more image detail backlight region to some.

Judge whether will adjust exposure, determination methods according to reference brightness and object brightness:

If reference brightness in the higher limit of object brightness and the scope of lower limit, does not need to adjust exposure;

If reference brightness exceeds the higher limit of object brightness and the scope of lower limit, need to adjust exposure.

For example, judge that reference brightness refY is whether in the higher limit of object brightness tarY and the scope of lower limit, as, underLimit value is tarY-5, and higher limit tarY+5 judges that refY is whether within the scope of tarY-5～tarY+5, not within the scope of thisCarry out overall gain calculating and exposure adjustment.

Adjust exposure ratio ExpRate, phase according to reference brightness (refY) and object brightness (tarY) phase extentThe exposure ratio of poor larger adjustment is larger, and the exposure ratio that differs less adjustment is less, draws by exposure ratio function:

ExpRate=F(abs(tarY-refY))。

Draw the overall gain that will arrange: TotalGain=with described exposure ratio and current overall gainTotalGain*ExpRate。

The exposure Policy Table of the overall gain that use will arrange and appointment draws time for exposure, the gain that will arrange, specifically bagDraw together:

First searching first maximum overall gain MaxTotalGain Policy Table from exposure is more than or equal to and will arrangesThe list item of overall gain TotalGain;

Judge again the exposure mode of this list item searching:

The gain mode that exposes if,

Gain=MaxGain；

ExposureGain=TotalGain/Gain；

Gain mode if,

ExposureGain=MaxGain；

Gain=TotalGain/ExposureGain；

Then, the gain conversions of exposing is the time for exposure,

ExposureTime=ExposureGain*ExposureGainUnit。

The time for exposure obtaining, gain are established into imageing sensor and exposed.

This image processing method also comprises auto white balance method, use in advance colour temperature lamp box and standard color card measure high,In, the corresponding red blue gain of white balance and color correction matrix under low three colour temperatures, three colour temperatures of selected this need contain substantiallyColour temperature in surroundings, is determined as follows:

Under high color temperature, colour temperature is HColorTemp, and the red gain of white balance is HRedGain, and the blue gain of white balance isHBlueGain, color correction matrix is HColorMatrix;

Under middle colour temperature, colour temperature is MColorTemp, and the red gain of white balance is MRedGain, white balance

Blue gain is MBlueGain, and color correction matrix is MColorMatrix;

Under low colour temperature, colour temperature is LColorTemp, and the red gain of white balance is LRedGain, and the blue gain of white balance isLBlueGain, color correction matrix is LColorMatrix;

Had the red blue gain of these three groups of white balances of measuring in advance and color correction matrix, just can be according to Fig. 2 automaticWhite balance handling process is carried out white balance.

As shown in Figure 2, described auto white balance method comprises execution following steps:

In step S1, R, G, B white balance statistics, for counting green red ratio G/R and the turquoise ratio of present imageG/R；

In step S2, estimate current colour temperature according to statistical information, for according to statistics green red ratio G/R and turquoise ratioRate G/R and color temperature estimation function, estimate current colour temperature;

In step S3, according to the colour temperature of estimating and the corresponding red blue gain of white balance and the color of high, medium and low three colour temperaturesCorrection matrix, interpolation goes out the red blue gain of current white balance and color correction matrix;

In step S4, the red blue gain of white balance and color correction matrix are set.

In step S3, interpolation method can change flexibly, and the most easy method is linear interpolation, specific as follows:

If current colour temperature CColorTemp is between high color temperature and middle colour temperature,

The red gain of current white balance:

$\mathrm{CRedGain}=\mathrm{MRedgain}+\frac{(\mathrm{CColorTemp}-\mathrm{MColorTemp})}{(\mathrm{HColorTemp}-\mathrm{MColorTemp})}*(\mathrm{HRedGain}-\mathrm{MRedGain})$

The blue gain of current white balance:

$\mathrm{CBlueGain}=\mathrm{MBlueGain}+\frac{(\mathrm{CColorTemp}-\mathrm{MColorTemp})}{(\mathrm{HColorTemp}-\mathrm{MColorTemp})}*(\mathrm{HBlueGain}-\mathrm{MBlueGain})$

Current color correction matrix:

$\mathrm{CColorMatrix}=\mathrm{MColorMatrix}+\frac{(\mathrm{CColorTemp}-\mathrm{MColorTemp})}{(\mathrm{HColorTemp}-\mathrm{MColorTemp})}*(\mathrm{HColorMatrix}-\mathrm{MColorMatrix})$

If current colour temperature is between middle colour temperature and low colour temperature,

The red gain of current white balance:

$\mathrm{CRedGain}=\mathrm{LRedGain}+\frac{(\mathrm{CColorTemp}-\mathrm{LColorTemp})}{(\mathrm{MColorTemp}-\mathrm{LColoTemp})}*(\mathrm{MRedGain}-\mathrm{LRedGain})$

The blue gain of current white balance

$\mathrm{CBlueGain}=\mathrm{LBlueGain}+\frac{(\mathrm{CColorTemp}-\mathrm{LColorTemp})}{(\mathrm{MColorTemp}-\mathrm{LColorTemp})}*(\mathrm{MBlueGain}-\mathrm{LBlue}\mathrm{Gain})$

Current color correction matrix:

$\mathrm{CColorMatrix}=\mathrm{LColorMatrix}+\frac{(\mathrm{CColorTemp}-\mathrm{MColorTemp})}{(\mathrm{LColorTemp}-\mathrm{MColorTemp})}*(\mathrm{MColorMatrix}-\mathrm{LColorMatrix})$

Color correction matrix is the coefficient matrix of 3x3, concrete:

$\left(\begin{array}{ccc}\mathrm{RR}& \mathrm{RG}& \mathrm{RB}\\ \mathrm{BR}& \mathrm{BG}& \mathrm{BB}\end{array}\right)$

The method of color correction:

New color value=old color value × color correction matrix after correction.

Carry out white balance R, G, B statistics, count green red ratio G/R and the turquoise ratio G/R of present image.

Estimate current colour temperature according to statistical information, be specially according to green red ratio G/R and turquoise ratio G/R and the look of statisticsTemperature estimation function, estimates current colour temperature.

The invention also discloses a kind of image processing system, comprise automatic exposure unit, described automatic exposure unit comprises:

Brightness statistic: for image being divided into multiple range statistics mean flow rates, by each region according to weight tableThe weight calculation of middle appointment goes out the mean flow rate of whole two field picture, and the mean flow rate of adding up nearest some two field pictures is bright as a referenceDegree;

Adjust exposure judge module: for judging whether to adjust exposure according to described reference brightness and object brightness, ifAdjust exposure, carry out the overall gain module of obtaining setting;

Obtain the overall gain module of setting: draw the overall gain of setting by exposure ratio and current overall gain, arrangeThe current overall gain of overall gain=exposure ratio *;

Acquisition module: draw time for exposure and gain for the overall gain and the exposure Policy Table that use setting;

Exposure module: for the time for exposure of acquisition module and gain are set in imageing sensor and are exposed.

In described adjustment exposure judge module: if reference brightness is in the higher limit of object brightness and the scope of lower limitIn, do not need to adjust exposure; If reference brightness exceeds the higher limit of object brightness and the scope of lower limit, need to adjustExposure.

In the described overall gain module of obtaining setting:

In described exposure ratio, adjust exposure ratio according to reference brightness and object brightness phase extent, exposureRatio=F (abs(object brightness-reference brightness));

In current overall gain, be first exposure gain by the time for exposure divided by exposure gain Conversion of measurement unit, then willExposure gain * gain obtains current overall gain.

In described acquisition module, the project of described exposure Policy Table's list item comprises:

Maximum exposure gain, the maximum exposure that can reach for this section of exposure gain;

Maximum gain, the maximum gain that can reach for this section of exposure;

Maximum overall gain, the maximum overall gain that can reach for this section of exposure, maximum overall gain=maximum exposure increasesBenefit * maximum gain;

Exposure mode, for this section of exposure mode that exposure is used, is divided into two kinds of exposure gain mode and gain modes;Exposure gain mode, gain keeps maximum gain constant, and exposure gain changes between the gain of 0～maximum exposure; Gain mode,Time for exposure gain keeps maximum exposure gain constant, and gain changes between 1～maximum gain;

In described acquisition module, comprising:

Search module, is more than or equal to total increasing that will arrange for searching first maximum overall gain from exposure Policy TableThe list item of benefit;

Judge module, for judging the exposure mode of this list item searching: the gain mode that exposes if, gain=Maximum gain, maximum exposure gain=overall gain/gain; Gain mode if, exposure gain=maximum gain, gain=totalGain/exposure gain;

Modular converter, is the time for exposure for the gain conversions of exposing, time for exposure=exposure gain * exposure gain unit.

This image processing system also comprises AWB unit, and described AWB unit comprises:

White balance statistics module, for R, G, B white balance statistics, counts the green red ratio G/R of present image and turquoiseRatio G/R;

Estimate current colour temperature module, for estimate current colour temperature according to statistical information, for according to statistics green red ratioG/R and turquoise ratio G/R and color temperature estimation function, estimate current colour temperature;

Interpolating module, for gaining and face according to the colour temperature of estimating and the red indigo plant of white balance corresponding to high, medium and low three colour temperaturesLook correction matrix, interpolation goes out the red blue gain of current white balance and color correction matrix;

Module is set, for the red blue gain of white balance and color correction matrix are set.

Automatic explosion method of the present invention and unit have improved the unsettled problem of exposure well, pass through to customize simultaneouslyDifferent exposure strategies reaches optimum the image effect under varying environment; Auto white balance method of the present invention and unit according toWhen colour temperature has been carried out the red blue gain of white balance, also introduce color correction matrix, under different-colour, use different color correctionsMatrix, has so greatly improved the white balance colour cast problem producing due to the impact of imageing sensor and optical filter.

Above content is in conjunction with concrete preferred embodiment further description made for the present invention, can not assertSpecific embodiment of the invention is confined to these explanations. For general technical staff of the technical field of the invention,Do not depart under the prerequisite of the present invention's design, can also make some simple deduction or replace, all should be considered as belonging to of the present inventionProtection domain.

Claims (8)

1. an image processing method, is characterized in that, comprises automatic explosion method, described automatic explosion method following steps:

Brightness statistics step: image is divided into multiple range statistics mean flow rates, by each region according to specifying in weight tableWeight calculation goes out the mean flow rate of whole two field picture, and the mean flow rate of adding up nearest some two field pictures is as with reference to brightness;

Adjust exposure determining step: judge whether to adjust exposure according to described reference brightness and object brightness, expose to the sun if will adjustLight, carries out the overall gain step of obtaining setting;

Obtain the overall gain step of setting: draw the overall gain of setting by exposure ratio and current overall gain, what arrange is totalThe current overall gain of gain=exposure ratio *;

Obtaining step: use the overall gain and the exposure Policy Table that arrange to draw time for exposure and gain;

Step of exposure: the time for exposure in obtaining step and gain are set in imageing sensor and are exposed;

This image processing method also comprises auto white balance method, and described auto white balance method comprises execution following steps:

A.R, G, B white balance statistics, for counting green red ratio G/R and the turquoise ratio G/R of present image;

B. estimate current colour temperature according to statistical information, for according to statistics green red ratio G/R and turquoise ratio G/R and colour temperatureEstimation function, estimates current colour temperature;

C. according to the colour temperature of estimating and the corresponding red blue gain of white balance and color correction matrix of high, medium and low three colour temperatures, insertBe worth the red blue gain of current white balance and color correction matrix;

D., the red blue gain of white balance and color correction matrix are set.

2. image processing method according to claim 1, is characterized in that, in described adjustment exposure determining step: asFruit reference brightness, in the higher limit of object brightness and the scope of lower limit, does not need to adjust exposure; If reference brightness is superGo out the higher limit of object brightness and the scope of lower limit, need to adjust exposure.

3. image processing method according to claim 1, is characterized in that, in the described overall gain step of obtaining settingIn:

In described exposure ratio, adjust exposure ratio according to reference brightness and object brightness phase extent, exposure ratio=F (abs(object brightness-reference brightness));

In current overall gain, first the time for exposure is gained for exposing divided by exposure gain Conversion of measurement unit, then will exposeGain * gain obtains current overall gain.

4. image processing method according to claim 3, is characterized in that, in described obtaining step, and described exposure planThe project of the list item of sketch form comprises:

Maximum exposure gain, the maximum exposure that can reach for this section of exposure gain;

Maximum gain, the maximum gain that can reach for this section of exposure;

Maximum overall gain, the maximum overall gain that can reach for this section of exposure, maximum overall gain=maximum exposure gains *Large gain;

Exposure mode, for this section of exposure mode that exposure is used, is divided into two kinds of exposure gain mode and gain modes; ExposureGain mode, gain keeps maximum gain constant, and exposure gain changes between the gain of 0～maximum exposure; Gain mode, exposureGain keeps maximum exposure gain constant, and gain changes between 1～maximum gain;

In described obtaining step, comprise execution following steps:

First from exposure Policy Table, search the list item that first maximum overall gain is more than or equal to the overall gain that will arrange;

Judge again the exposure mode of this list item searching: the gain mode that exposes if, gain=maximum gain, exposure increaseBenefit=overall gain/gain; Gain mode if, exposure gain=maximum gain, gain=overall gain/exposure gain;

Then, the gain conversions of exposing is the time for exposure, time for exposure=exposure gain * exposure gain unit.

5. an image processing system, is characterized in that, comprises automatic exposure unit, and described automatic exposure unit comprises:

Brightness statistic: for image being divided into multiple range statistics mean flow rates, by each region according to weight table middle fingerFixed weight calculation goes out the mean flow rate of whole two field picture, and the mean flow rate of adding up nearest some two field pictures is as with reference to brightness;

Adjust exposure judge module: for judging whether to adjust exposure according to described reference brightness and object brightness, if will adjustWhole exposure, carries out the overall gain module of obtaining setting;

Obtain the overall gain module of setting: draw the overall gain of setting by exposure ratio and current overall gain, what arrange is totalThe current overall gain of gain=exposure ratio *;

Acquisition module: draw time for exposure and gain for the overall gain and the exposure Policy Table that use setting;

Exposure module: for the time for exposure of acquisition module and gain are set in imageing sensor and are exposed;

This image processing system also comprises AWB unit, and described AWB unit comprises:

White balance statistics module, for R, G, B white balance statistics, counts green red ratio G/R and the turquoise ratio of present imageG/R；

Estimate current colour temperature module, for estimate current colour temperature according to statistical information, for according to statistics green red ratio G/R andTurquoise ratio G/R and color temperature estimation function, estimate current colour temperature;

Interpolating module, for gaining and color school according to the colour temperature of estimating and the red indigo plant of white balance corresponding to high, medium and low three colour temperaturesPositive matrices, interpolation goes out the red blue gain of current white balance and color correction matrix;

Module is set, for the red blue gain of white balance and color correction matrix are set.

6. image processing system according to claim 5, is characterized in that, in described adjustment exposure judge module: asFruit reference brightness, in the higher limit of object brightness and the scope of lower limit, does not need to adjust exposure; If reference brightness is superGo out the higher limit of object brightness and the scope of lower limit, need to adjust exposure.

7. image processing system according to claim 5, is characterized in that, in the described overall gain module of obtaining settingIn:

In described exposure ratio, adjust exposure ratio according to reference brightness and object brightness phase extent, exposure ratio=F (abs(object brightness-reference brightness));

In current overall gain, first the time for exposure is gained for exposing divided by exposure gain Conversion of measurement unit, then will exposeGain * gain obtains current overall gain.

8. image processing system according to claim 5, is characterized in that, in described acquisition module, and described exposure planThe project of the list item of sketch form comprises:

Maximum exposure gain, the maximum exposure that can reach for this section of exposure gain;

Maximum gain, the maximum gain that can reach for this section of exposure;

Maximum overall gain, the maximum overall gain that can reach for this section of exposure, maximum overall gain=maximum exposure gains *Large gain;

Exposure mode, for this section of exposure mode that exposure is used, is divided into two kinds of exposure gain mode and gain modes; ExposureGain mode, gain keeps maximum gain constant, and exposure gain changes between the gain of 0～maximum exposure; Gain mode, exposureGain keeps maximum exposure gain constant, and gain changes between 1～maximum gain;

In described acquisition module, comprising:

Search module, is more than or equal to the overall gain that will arrange for search first maximum overall gain from exposure Policy TableList item;

Judge module, for judging the exposure mode of this list item searching: the gain mode that exposes if, gain=maximumGain, exposure gain=overall gain/gain; Gain mode if, exposure gain=maximum gain, gain=overall gain/Exposure gain;

Modular converter, is the time for exposure for the gain conversions of exposing, time for exposure=exposure gain * exposure gain unit.