CN101287073B - Adaptive acquiring method of lightness stabilized image from machine vision system in variable irradiation surroundings - Google Patents

Abstract

The invention provides a method for automatically acquiring standard images with stable quality for a machine vision system in variable illumination environment, which relates to the image processing technical field, aiming at solving the technical difficulty of the prior art to guarantee that the machine vision system, more particularly the machine vision system without illumination equipment can acquire the standard images with the stable quality in the extensive variable illumination environment. The automatic adjusting method for acquiring image quality includes the following steps of: starting a digital camera according to preset initial parameters, calculating the pixel grayness average 'u' of an image after collecting a fame of the image, judging whether the image qualities (including the adjustment of a too bright area, a too dark area and an exact adjusting unit indicated by the digital camera) needs adjusting according to the pixel grayness average 'u' of the image.

Description

Become Vision Builder for Automated Inspection adaptive acquiring method of lightness stabilized image in the photoenvironment

Technical field

The present invention relates to technical field of image processing, relate in particular in the external environment of illumination variation, Vision Builder for Automated Inspection, the Vision Builder for Automated Inspection that lighting apparatus particularly is not installed is obtained the method for stay-in-grade standard picture.

Background technology

Vision Builder for Automated Inspection is widely used in the links such as operating condition monitoring, product inspection and quality control of industries such as semiconductor, machinery, automobile, medicine, food.In recent years, at other non-industrial circle such as safety inspection, Vision Builder for Automated Inspection also has been subjected to attention day by day.

In most engineering practical application, Vision Builder for Automated Inspection need be worked in various photoenvironments complicated and changeable.The image definition that the illumination condition that changes causes image capture device to grasp is poor, contrast is on the low side, quality is unstable, compares with the common required standard images acquired of image processing section, and the certain quality gap is arranged.This phenomenon lacks site of deployment such as the contactless dimensional measurement, literal identification, recognition of face of lighting apparatus only relying on lamp comparatively common.The illumination that changes in the external environment has brought great challenge to Vision Builder for Automated Inspection, has influenced the performance of Vision Builder for Automated Inspection.There is data to show, the one of the main reasons that illumination variation breaks down when being the Vision Builder for Automated Inspection running.

Can both operate as normal under most of illumination conditions in order to ensure Vision Builder for Automated Inspection, in design process, improve its to external world in the environment adaptive capacity of variation illumination have great importance.Especially lack initiatively light source for those, rely on the Vision Builder for Automated Inspection of lamp merely, the illumination variation of how handling in the external environment is a stern challenge.If the illumination condition that changes in the ignorance external environment, because the picture quality contrast that collects is on the low side, the Vision Builder for Automated Inspection that has in some cases can't operate as normal.

In order to guarantee the stability and the uniformity of destination object surface light photograph, the designer usually introduces extra lighting apparatus and weakens and even remove natural daylight and change the influence for the treatment of process object surface light photograph, for example lighting source initiatively such as LED lamp.Though adopt extra lighting source can remove the influence that natural daylight changes to a certain extent, the method has following limitation.

(1) lighting apparatus has increased the overall cost of Vision Builder for Automated Inspection.

(2) need purchase bracing or strutting arrangement, fixture, electric power system, and take certain installing space, increase the cost and the complexity of Vision Builder for Automated Inspection.Because the space is limited, in some occasion lighting apparatus can't be installed.

(3) running of lighting apparatus has increased the energy loss of Vision Builder for Automated Inspection.Because the restriction of energy storage devices such as battery capacity, in tasks such as deep-sea detecting, space development, the energy that offers lighting apparatus is limited even can't offers the lighting apparatus energy.At this moment, can't set up the intensity of illumination of predefined object surface.

(4) some part in the lighting apparatus is a consumptive material, needs regular maintenance and replacing, part such as bulb for example, and this needs extra maintenance cost.The operating state of these parts will influence the work of system, increase the risk that system breaks down.Its fault even may cause Vision Builder for Automated Inspection normally to turn round.

(5), need configuration certain limit inner glow intensity can be in harmonious proportion the continuously adjustable lighting apparatus of irradiating angle continuously in order to guarantee the continuous scalable of object surface intensity of illumination.

(6) in some cases, the use of lighting apparatus is comparatively complicated.Even disposed all continuously adjustable lighting apparatus of luminous intensity and irradiating angle, in order to set up the satisfactory intensity of illumination of object surface, the user must manually set the luminous intensity and the irradiating angle of lighting apparatus according to the illumination condition in the external world.In some cases, adjustment process is comparatively loaded down with trivial details.The intensity of illumination of final object surface is subjected to the influence of many-sided factor.

Because be subjected to the restriction of conditions such as cost, operational environment, energy resource consumption, some Vision Builder for Automated Inspection lacks lighting device, can only work in the natural lighting environment fully.The image processing section of most of Vision Builder for Automated Inspection all designs at standard picture.Yet when illumination variation was big, image that image collecting device grabs and predefined standard picture had bigger gap on picture quality.May cause follow-up image processing bigger error to occur thus even make mistakes.For this type of Vision Builder for Automated Inspection, how (intensity that requires the ambient light photograph is greater than certain threshold value) gets access to stay-in-grade standard picture in the photoenvironment that changes, and is the major issue of guaranteeing the machine system operate as normal.The Vision Builder for Automated Inspection to lighting apparatus is installed that addresses this problem also has positive meaning.If under the prerequisite of not introducing lighting apparatus, solved and become the problem of obtaining standard picture in the photoenvironment, just can in the design of some Vision Builder for Automated Inspection, remove lighting device, guarantee Vision Builder for Automated Inspection to external world the robustness of illumination variation reduce the overall cost of system simultaneously.

At present, weakening extraneous illumination variation is the image preliminary treatment to the main method that picture quality influences.Pretreated is standard picture or quality and the approximate image of standard picture with the image transitions that collects.The image preliminary treatment can weaken the influence of ambient light photograph to a certain extent.Most of image pre-processing methods improve the quality of image by adjusting the contrast that grasps image, and for example histogram equalization and the various of other are improved one's methods.Preliminary treatment has following shortcoming.

(1) with respect to original image, strengthen operation and brought extra noise, caused the loss of original image partial information.

(2) in most cases, the image preliminary treatment relies on software processes to realize.This means that Vision Builder for Automated Inspection need provide extra processing time and memory space for preliminary treatment.When big or memory space is big when the image resolution ratio that grasps, long processing time that the image preliminary treatment needs and bigger memory space.This can't allow in some high speed processing system.

(3) effects of pretreatment and some parameter has chosen direct relation.Choose and the illumination of parameter have direct relation, usually need manual intervention so that obtain treatment effect preferably.So when the preliminary treatment automatic operating, be difficult to guarantee the effect after the extracting figure image intensifying.Under situations such as the contrast that grasps image is lower, preprocess method lost efficacy probably.

Except above-mentioned image pre-processing method, extraction has the feature of illumination consistency or introduces certain methods such as illumination model and also can weaken the influence of object surface illumination variation to collection image quality.Yet these methods only are confined to some specific application places such as recognition of face, target following, are not of universal significance, and are difficult to use widely in other Vision Builder for Automated Inspection.

Existing technology major part is all taked the mode of operation of " grasping image preliminary treatment more earlier " in obtaining the process of image.This mode of operation has been ignored the variation of ambient light photograph, each all start collecting device with intrinsic parameter and grasps image, relies on software processing technology means such as image preliminary treatment to improve the quality of image more passively.Under some light conditions, for example in stronger illumination or darker photoenvironment, the above-mentioned image that only depends on after the collection again the method for simple software processes to be difficult to the standard that obtains sometimes.

In sum, existing technology is difficult to guarantee becoming in the photoenvironment on a large scale, and the Vision Builder for Automated Inspection that Vision Builder for Automated Inspection, particularly without a licence exposed installation are equipped with is obtained stay-in-grade standard picture.

Technology contents

The invention provides in a kind of change photoenvironment Vision Builder for Automated Inspection and obtain stay-in-grade standard picture method automatically, its purpose is to solve existing technology and is difficult to guarantee becoming on a large scale in the photoenvironment, Vision Builder for Automated Inspection, particularly the Vision Builder for Automated Inspection that is equipped with of without a licence exposed installation is obtained stay-in-grade standard picture, to improve the robustness of Vision Builder for Automated Inspection to illumination variation, particularly under the situation of not introducing lighting apparatus (intensity that requires the ambient light photograph is greater than certain threshold value) helps Vision Builder for Automated Inspection promptly to obtain the standard picture that satisfies quality requirement (contrast etc.) effectively.

Above-mentioned purpose of the present invention is achieved in that accompanying drawings is as follows:

Vision Builder for Automated Inspection is obtained stay-in-grade standard picture method automatically in a kind of change photoenvironment, it is characterized in that adopting the automatic adjusting method that grasps picture quality, specifically comprises the steps:

1) start digital camera according to predefined initial parameter, gather a two field picture after, the pixel gray average u of computed image, computing formula is:

Wherein, f (i, j) be the picture element coordinate (i, gray average j), u are the gray average of this width of cloth image pixel, M and N are the parameters of token image size;

2) the pixel gray average u according to image judges whether that needs carry out the adjustment of picture quality, and by the requirement of quality such as picture contrast, the expectation interval of setting the gray average of image pixel is [u _Min, u _Max], if the pixel gray average u of the image that collects is in the expectation interval, do not need to adjust the parameter of Vision Builder for Automated Inspection image capture device, then withdraw from the automatic adjustment program of picture quality;

3) cross the clear zone and regulate, if the pixel gray average u of image is greater than u _Max, then key diagram looks like to be in the clear zone, the beginning image cross clear zone adjustment, the clear zone is regulated the pixel gray average that comprises the index of selecting digital camera in a certain order, index change, image and is upgraded, judges whether to be in the clear zone.If the adjustment through the digital camera index had broken away from the clear zone, then enter next step and cross the dark space adjusting; If extraneous illumination can't break away from the clear zone too by force all the time, withdraw from whole automatic adjustment program;

4) cross the dark space and regulate, if the pixel gray average u of image is less than u _MinIllustrate that then the image pixel gray average was in the dark space, the dark space excessively of beginning image is adjusted, the dark space is regulated and is comprised that index, the index of selecting digital camera are in a certain order changed, the pixel gray average of image upgrades, judges whether to be in the dark space, if cross the dark space through adjusting to have broken away from, then enter accurate adjustment unit; If can't break away from the dark space all the time a little less than the extraneous illumination too, withdraw from whole automatic adjustment program;

5) the accurate adjustment unit of digital camera index, this element comprise digital camera index selection, parametric solution, judge whether to satisfy the picture quality requirement, at first adopt the mode of On-line Estimation to obtain the parameter sets I of intensity of illumination in the reflection current environment; According to predefined picture quality requirement, select the index of number of digital video camera, corresponding digital camera parameter set U; The parameter sets U functional relation between the two of the pixel gray average u of image and digital camera be u=f (U, I).Pixel gray average u according to predefined image _d, find the solution digital camera parameter sets U _d=f ^-1(u _d, I); With the digital camera parameter sets of trying to achieve, as final image acquisition parameter; If do not satisfy the requirement of pixel gray average with the image of digital camera parameter acquisition, then reselect the index of number of digital video camera, find the solution new digital camera parameter sets U _n

The good effect of technical scheme provided by the present invention is: all to grasp the mode that image handles again different at every turn with most existing, the present invention changes parameter as the digital camera of image collecting device in real time according to the light conditions in the external world, save the software processes time that " grasping image preliminary treatment more earlier " needs, improved the efficient of figure image intensifying.Under the situation of the intensity of illumination approximate constant of the present invention in external environment, only need once to adjust, and help collecting device to grab satisfactory stay-in-grade standard picture just at the relevant parameter of Vision Builder for Automated Inspection of can automatically adjusting at short notice.

Description of drawings

The present invention will be by example, and the following accompanying drawing of reference is further to set forth:

Fig. 1 for the present invention in becoming photoenvironment, automatically obtain the system block diagram of stay-in-grade standard picture.

Fig. 2 is the present invention in becoming photoenvironment, automatically obtains the program flow diagram of stay-in-grade standard picture.

Fig. 3 is the function curve diagram between the gain of the pixel gray average of image and digital camera.

Fig. 4 is the function curve diagram of the time for exposure of the pixel gray average of image and digital camera.

Fig. 5 crosses under the bright photograph image grey level histogram when adjusting without picture quality.

Fig. 6 crosses under the bright photograph image grey level histogram when picture quality is adjusted automatically.

Fig. 7 shines down the image grey level histogram when adjusting without picture quality than half-light.

Fig. 8 shines down the image grey level histogram when picture quality is adjusted automatically than half-light.

Embodiment

The invention provides under a kind of change illumination condition, Vision Builder for Automated Inspection does not particularly have the Vision Builder for Automated Inspection of lighting apparatus automatically to obtain the method for stay-in-grade standard picture, in order to solve the problem that the integral image gray scale changes along with extraneous illumination variation in the Vision Builder for Automated Inspection.Below in conjunction with the accompanying drawing illustrated embodiment, technical scheme of the present invention is described in further detail.

In the present invention, the automatic adjusting method of extracting picture quality comprises following step.

Start digital camera according to predefined initial parameter.After gathering a two field picture, the pixel gray average u of computed image.

Pixel gray average u according to image judges whether that needs carry out the adjustment of picture quality.If the image that collects meets the demands, then withdraw from the automatic adjustment program of picture quality.Otherwise, begin following picture quality and adjust automatically.

Crossing the clear zone regulates.This step mainly is that to guarantee to grasp the gray value of image not excessive, prevents that image was in bright state.Judge whether that according to the pixel gray average u of image needs cross the clear zone and regulate.If needed the clear zone to regulate,, select corresponding digital video camera index to adjust then according to the intensity of current light; Otherwise forwarding the dark space to regulates.Cross the clear zone adjustment and comprise that index, the index of selecting digital camera are in a certain order changed, the pixel gray average of image upgrades, judges whether to be in steps such as clear zone.If the adjustment through the digital camera index had broken away from the clear zone, then entered the dark space and regulated.If can't break away from the clear zone all the time, to such an extent as to it is bigger than normal to illustrate that then strong gradation of image value is crossed in extraneous illumination, the contrast of image does not meet the demands, and withdraws from whole automatic adjustment program.Crossing the clear zone adjusting leaves it at that.

Crossing the dark space regulates.Pixel gray average u by image needed to judge whether the dark space to regulate.Judge whether to start the adjusting of dark space excessively of picture quality.If needed the dark space to regulate, then adjusted the corresponding index of digital camera; Otherwise forward accurate adjustment unit to.Cross the dark space adjustment and comprise that index, the index of selecting digital camera are in a certain order changed, the pixel gray average of image upgrades, judges whether to be in steps such as dark space.If cross the dark space through adjusting to have broken away from, then enter accurate adjustment unit.If can't break away from the dark space all the time, then a little less than the illumination too in the explanation external world, withdraw from whole automatic adjustment program.The dark space adjusting of crossing of digital camera index leaves it at that.

The accurate adjustment unit of digital camera index.Through the operation of this element, can determine the accurate numerical value of the some internal indicator of digital camera.This element comprise digital camera index selection, parametric solution, judge whether to satisfy steps such as picture quality requirement.At first adopt the mode of On-line Estimation to obtain the parameter sets I that reflects intensity of illumination in the current environment.According to predefined picture quality requirement, select the index of number of digital video camera, corresponding digital camera parameter set U.G, time for exposure t for example gain _eEtc. index.The parameter sets U functional relation between the two of the pixel gray average u of image and digital camera be u=f (U, I).Pixel gray average u according to predefined image _d, find the solution digital camera parameter sets U _d=f ^-1(u _d, I).With the digital camera parameter sets of trying to achieve, as final image acquisition parameter.If do not satisfy the requirement of pixel gray average with the image of digital camera parameter acquisition, then reselect the index of number of digital video camera, find the solution new digital camera parameter sets U _n

Fig. 1 for the present invention in becoming photoenvironment, automatically obtain the system block diagram of stay-in-grade standard picture.Consult accompanying drawing: according to initial parameter the startup digital camera is set and gathers a two field picture, calculate the pixel gray average of this image.

$u=\frac{\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\underset{j=1}{\overset{N}{\mathrm{\Σ}}}f(i,j)}{\mathrm{MN}}---\left(1\right)$

Wherein f (i, j) be the picture element coordinate (i, gray average j), u are the gray average of this width of cloth image pixel, M and N are the parameters of token image size.

Consider the requirement of quality such as picture contrast, the expectation interval of setting the gray average of image pixel is [u _Min, u _Max].

If the pixel gray average u of image is in the expectation interval, illustrate that the quality of present image meets the requirement of subsequent processes, the picture quality that promptly collects and the quality of standard picture differ less.The parameter that do not need to adjust the Vision Builder for Automated Inspection image capture device this moment withdraws from self-regulating process.Otherwise the clear zone of crossing that enters is as shown in Figure 2 regulated.

Crossing the clear zone regulates.If the pixel gray average u of image is greater than u _Max, then key diagram looks like to be in the clear zone, and the clear zone excessively of beginning image is adjusted; Otherwise forwarding the dark space to regulates.Cross the clear zone adjustment and comprise that index, the index of selecting digital camera are in a certain order changed, the pixel gray average of image upgrades, judges whether to be in steps such as clear zone.The digital camera index that can regulate comprises a plurality of indexs such as time for exposure, white balance, black level, gain and Gamma correction.

With time for exposure t _eBe example, the index adjustment process of digital camera is described.

(a) suitably reduce the time for exposure t of digital camera _e

t _e＝t _e-t _d (2)

T wherein _dFor reducing the step-length of time for exposure.

(b) judge whether the time for exposure reaches predefined minimum value t _MaxIf then the explanation adjusting time for exposure can't make the pixel gray average of image break away from the clear zone, withdraws from the adjusting of time for exposure, select other digital camera index to continue to regulate.Otherwise, continue next step.

(c) the pixel gray average of computed image.

(d) whether test u is in the interval [u of expectation _Min, u _Max].If it is interval that u is in expectation, then the pixel gray average of key diagram picture had broken away from the clear zone, rotated the dark space and regulated; Otherwise forward (a) to, continue to regulate the time for exposure.

If after regulating the index of digital camera in a certain order, still can't break away from the clear zone, cross the strong gray average that can't adjust the extracting image to the expectation interval to such an extent as to the illumination in the external environment then is described, withdraw from the automatic adjustment program of entire image quality.Crossing the clear zone adjusting leaves it at that.

Regulate the index of digital camera and should follow certain sequence.This wherein will consider the characteristic of digital camera, can regulate carelessly regularly.Rationally arrange the adjusting order of digital camera index, can reach the effect of getting twice the result with half the effort; Otherwise,,, waste the unnecessary time with prolonging the adjustment process of entire image quality if adopt the index of inappropriate sequential adjustment digital camera.Can break away under the situation in clear zone fully, and adopt inappropriate index adjusting order possibly can't reduce the pixel gray average of image, cause occurring asserting that extraneous light crossed the mistake that can't adjust picture quality by force.

Crossing the dark space regulates.Pixel gray average u by image needed to judge whether the dark space to regulate.If the pixel gray average u of image is less than u _Min, illustrating that then the image pixel gray average was in the dark space, the dark space excessively of beginning image is adjusted; Otherwise forward accurate adjustment unit to.Cross the dark space adjustment and comprise that index, the index of selecting digital camera are in a certain order changed, the pixel gray average of image upgrades, judges whether the quality of image was in steps such as dark space.Digital camera to be selected is regulated index and is comprised a plurality of indexs such as time for exposure, white balance, black level, gain and Gamma correction.

If after regulating the index of digital camera in a certain order, still can't make the gray average of image pixel break away from the dark space, can't grab the image that meets the demands by the parameter of adjusting image capture device to such an extent as to illustrate then that illumination in the external environment is too weak.Withdraw from the automatic adjustment program of entire image quality.Crossing the dark space adjusting leaves it at that.

Similar with the adjusting of crossing the clear zone, regulate the index of digital camera and should follow certain sequence.If adopt the index of inappropriate sequential adjustment digital camera,, waste the unnecessary time with prolonging the whole adjustment process of crossing the dark space.Adopt inappropriate adjusting order, even cause to break away under the situation of dark space fully, to such an extent as to the too weak mistake that can't adjust picture quality of extraneous intensity of illumination occurs asserting.

Accurately adjust.Through the operation of this element, can determine the definite numerical value of the some internal indicator of digital camera.Adopt the mode of On-line Estimation to ask for the parameter sets I that reflects intensity of illumination in the current environment.Expectation interval [u according to the gray average of intensity of illumination in the current environment and image pixel _Min, u _Max], the index of selection number of digital video camera constitutes corresponding digital camera parameter set U.Parameter sets comprises gain g, time for exposure t _e, indexs such as white balance w, black level b.Know that through experiment the pixel gray average u of image and the parameter sets U of digital camera satisfy certain functional relation, (U I) describes with u=f.Fig. 3 is under the different light situation, and the functional relation between the pixel gray average u of image and the gain g of digital camera is wherein all kept constant except that other indexs of the digital camera of gain.The pixel gray average u of Fig. 4 image and digital camera time for exposure t _eFunctional relation, wherein except that the time for exposure, other indexs of digital camera are all kept constant.

Digital camera parameter sets U is found the solution in expectation interval by the gray average of image pixel _d=f ^-1(u _d, I).With the parameter sets of trying to achieve, digital camera is set as final acquisition parameter.If the image of digital camera parameter acquisition does not still satisfy the requirement of pixel gray average, then reselect the index of number of digital video camera, find the solution new digital camera parameter sets U _n

Gain g with digital camera is the process that the example explanation is accurately adjusted below.

As shown in Figure 3, under different light, curve-equipartition between the two becomes two zones.When gain surpassed certain amplitude, the pixel gray average of image was approximate constant, and the pixel gray average of image and the curve of gain enter the saturation region; In remaining zone of curve, linear relationship is satisfied in the pixel gray average of image and gain approx, can think that in this zone the pixel gray average u of image and the gain g of digital camera satisfy following relation between the two approx

u＝kg+c (3)

The method of employing On-line Estimation is asked for the unknown parameter k under the current intensity of illumination.

In certain interval, get gain sequence { g _j(j=1,2 ... N).Adopt these gains to start digital camera respectively, obtain the pixel gray average sequence { u of image after the images acquired _j.Following function T to be optimized is set.

$T(k,g)=\underset{j=1}{\overset{N}{\mathrm{\Σ}}}{[({\mathrm{kg}}_j+c)-u_j]}^2---\left(4\right)$

The necessary condition of function T minimalization to be optimized is

$\frac{\∂T}{\∂k}=\mathrm{Nc}+\underset{j=1}{\overset{N}{\mathrm{\Σ}}}{g}_{j}k-\underset{j=1}{\overset{N}{\mathrm{\Σ}}}{u}_j=0---\left(5\right)$

$\frac{\∂T}{\∂c}=\underset{j=1}{\overset{N}{\mathrm{\Σ}}}{g}_{j}c+\underset{j=1}{\overset{N}{\mathrm{\Σ}}}{g^2}_{j}k-\underset{j=1}{\overset{N}{\mathrm{\Σ}}}{g}_j{u}_j=0---\left(6\right)$

So obtain following equation group

$\left\{\begin{array}{c}\mathrm{Nc}+\underset{j=1}{\overset{N}{\mathrm{\Σ}}}{g}_{j}k=\underset{j=1}{\overset{N}{\mathrm{\Σ}}}{u}_j\\ \underset{j=1}{\overset{N}{\mathrm{\Σ}}}{g}_{j}c+\underset{j=1}{\overset{N}{\mathrm{\Σ}}}{g^2}_{j}k=\underset{j=1}{\overset{N}{\mathrm{\Σ}}}{g}_j{u}_j\end{array}---\left(7\right)\right.$

Separate the least-squares estimation value k that above-mentioned equation group obtains unknown parameter k, c _e, c _e

If setting the pixel gray average of images acquired is u _d, then the gain g of digital camera is

$g=\frac{u_d-c_e}{k_e}---\left(8\right)$

For the gray average that guarantees image pixel is in the interval [u of expectation _Min, u _Max], select the gain g of digital camera to be

$g=\left\{\begin{array}{c}\left\{g_{\min }\right\}+\left[\frac{\left[g_{\max }\right]-\left\{g_{\min }\right\}}{2}\right]\mathrm{if}\left[g_{\max }\right]-\left\{g_{\min }\right\}\&GreaterEqual;0\\ \left[g_{\max }\right]\mathrm{if}\left[g_{\max }\right]-\left\{g_{\min }\right\}<0\end{array}\right.---\left(9\right)$

Wherein [] represents downward rounding operation, and { } represents to round up computing, g _MinAnd g _MaxFor

$g_{\max }=\frac{u_{\max }-c_e}{k_e}---\left(10\right)$

$g_{\min }=\frac{u_{\min }-c_e}{k_e}---\left(11\right)$

With the digital camera parameter sets of trying to achieve, as final image acquisition parameter.If do not satisfy the requirement of pixel gray average with the image of digital camera parameter acquisition, then reselect the index of number of digital video camera, find the solution new digital camera parameter sets U _n

Under than the intense light irradiation condition, the grey level histogram that adopts preset parameter to start digital camera extracting image is seen Fig. 5.After the automatic adjustment of picture quality, digital camera grasps the grey level histogram of image and sees Fig. 6.Compare two grey level histograms as can be known, through after the automatic adjustment of picture quality, the contrast of image is improved, and the intensity profile of pixel is more even in the image.

Under more weak illumination condition, the grey level histogram that adopts preset parameter to start digital camera extracting image is seen Fig. 7.Through after the automatic adjustment of picture quality, digital camera grasps the grey level histogram of image and sees Fig. 8.Relatively the grey level histogram of two width of cloth images is adjusted and the gray scale major part of the image pixel that grasps concentrates on the interval of [0,100] automatically without picture quality as can be known, intensity profile and inhomogeneous, and contrast is on the low side, is unfavorable for the carrying out of subsequent processes.Through after the automatic adjustment of picture quality, evenly how digital camera grasps gray distribution of image, has effectively changed the contrast of image.

Above-mentioned experimental result shows that the present invention can help Vision Builder for Automated Inspection particularly not have the Vision Builder for Automated Inspection of lighting apparatus automatically to obtain stay-in-grade standard picture becoming under the illumination condition, has improved the adaptive capacity of Vision Builder for Automated Inspection to illumination.

Obviously, those skilled in the art can carry out various changes and modification to the present invention and not break away from the spirit and scope of the present invention.Like this, if of the present invention these are revised and modification belongs within the scope of claim of the present invention and equivalent technologies thereof, then the present invention also is intended to comprise these changes and modification interior.

Claims (1)

1. one kind becomes in the photoenvironment Vision Builder for Automated Inspection and obtains stay-in-grade standard picture method automatically, it is characterized in that adopting the automatic adjusting method that grasps picture quality, specifically comprises the steps:

1) start digital camera according to predefined initial parameter, gather a two field picture after, the pixel gray average u of computed image, computing formula is: $u=\frac{\underset{i=1}{\overset{M}{\&Sum;}}\underset{j=1}{\overset{N}{\&Sum;}}f(i,j)}{\mathrm{MN}}$

Wherein, f (i, j) be the picture element coordinate (i, gray average j), u are the gray average of this width of cloth image pixel, M and N are the parameters of token image size;

2) the pixel gray average u according to image judges whether that needs carry out the adjustment of picture quality, and by the requirement of picture contrast quality, the expectation interval of setting the gray average of image pixel is [u _Min, u _Max], if the pixel gray average u of the image that collects is in the expectation interval, do not need to adjust the parameter of Vision Builder for Automated Inspection digital camera, then withdraw from the automatic adjustment program of picture quality;

3) cross the clear zone and regulate, if the pixel gray average u of image is greater than u _MaxThen key diagram looks like to be in the clear zone, the clear zone excessively of beginning image is adjusted, the clear zone is regulated and is comprised that index, the index of selecting digital camera are in a certain order changed, the pixel gray average of image upgrades, judges whether to be in the clear zone, if the adjustment through the digital camera index had broken away from the clear zone, then enter next step and cross the dark space adjusting,, withdraw from whole automatic adjustment program if extraneous illumination can't break away from the clear zone too by force all the time;

4) cross the dark space and regulate, if the pixel gray average u of image is less than u _MinIllustrate that then the image pixel gray average was in the dark space, the dark space excessively of beginning image is adjusted, the dark space is regulated and is comprised that index, the index of selecting digital camera are in a certain order changed, the pixel gray average of image upgrades, judges whether to be in the dark space, if cross the dark space through adjusting to have broken away from, then enter accurate set-up procedure,, withdraw from whole automatic adjustment program if can't break away from the dark space all the time a little less than the extraneous illumination too;

5) the accurate adjustment of digital camera index, this step comprise digital camera index selection, parametric solution, judge whether to satisfy the picture quality requirement; At first adopt the mode of On-line Estimation to obtain the parameter sets I that reflects intensity of illumination in the current environment; According to predefined picture quality requirement, select the index of number of digital video camera, constitute corresponding digital camera parameter set U; The parameter sets U functional relation between the two of the pixel gray average u of image and digital camera be u=f (U, I); Pixel gray average u according to predefined image _d, find the solution digital camera parameter sets U _d=f ^-1(u _d, I); With the digital camera parameter sets of trying to achieve, as final image acquisition parameter; If with digital camera parameter sets U _dThe image of gathering does not satisfy the requirement of pixel gray average, then reselects the index of number of digital video camera, finds the solution new digital camera parameter sets U _n

Images