CN104077759A - Multi-exposure image fusion method based on color perception and local quality factors - Google Patents

Abstract

The invention belongs to the technical field of image and video processing and particularly relates to a multi-exposure image fusion method based on color perception and local quality factors. The multi-exposure image fusion method mainly solves the problem of image color distortion caused by improper selection of fusion evaluation criterions during fusion processing of a multi-exposure image and can be used for improving quality of a final image under the premise that a dynamic range of the image is not changed. Remarkable details of the image under different exposures are fused in the same image with a low dynamic range, a proper reference image is selected by means of the local quality factors to perform color correction, and the image with the low dynamic range comparing favorably with the image with a high dynamic range in the result can be obtained under the premise that the dynamic range of the image is not changed. By means of the multi-exposure image fusion method based on color perception and local quality factors, the nature and clear multi-exposure image fusion result can be obtained conveniently. Therefore, the multi-exposure image fusion method based on color perception and local quality factors can be widely applied to the fields related to image and video processing.

Description

A kind of many exposures image interfusion method based on colour vision perception and the global quality factor

Technical field

The invention belongs to image, technical field of video processing, be specifically related to a kind of many exposures image interfusion method based on colour vision perception and the global quality factor.

Background technology

High dynamic range HDR image is one of problem in computer vision and Digital Image Processing, is also one of the hot issue in current image field.The dynamic range of image refers to the ratio between the brightness value of pixel the brightest in piece image and the brightness value of the darkest pixel.Due to the limitation of dynamic range, imaging device often can not reconstruct the scene of true nature.At first, this class image all completes with optical simulation.Nowadays, high dynamic range images is easy to generate, and only needs the different photo of a series of depth of exposures, just can make the image of high dynamic range.Practice shows, utilizes the different time shutter to take identical scene, because the every width image of difference of exposure all can show not details out of other picture showings.For instance, directly by the true colors of sunlit part, can only in the minimum image of exposure, be shown, and in other photos of these a series of different exposures the color of this part overexposure all.In image, the details of shaded side could complete presenting in the high image of exposure, and be fuzzy a slice in the lower image of exposure.Due to above problem, ordinary camera cannot catch simultaneously and represent all details in piece image; And for the mankind, because people's eyes can automatically adjust according to the brightness difference of object, the details under these different exposures of perception simultaneously, so photo always can not represent true and natural scene completely aspect visual experience.

The object that high dynamic range images is processed is: utilize existing imaging technique that observed scene is shown as far as possible really, make people by ordinary camera, with regard to restructural, go out the image that can compare favourably with reality scene validity, this processing procedure is exactly the synthetic and reconstruct of high dynamic range images.Aspect synthetic, utilize prior art by the different synthetic panel height dynamic image of picture of exposure, the dynamic range of this image can be up to 25,000:1; Yet the dynamic range of general display device is usually less than 100:1, therefore, compare with low dynamic range echograms, the details in the seizure image that high dynamic range images obviously can be sharper.

Compare with low dynamic range echograms, high dynamic range images has lot of advantages, and at medical image, in some application such as video monitor, high dynamic range images is particularly important.Yet, high dynamic range images is also for imaging technique has brought challenge: picture reproducer of today, as display, the dynamic range of printer etc. is all far smaller than the dynamic range in real scene, and urgent problem is exactly how under the prerequisite that as far as possible retains image detail and vision content, to utilize the picture reproducer of low-dynamic range to show the image of high dynamic range.

The integration technology of many exposures image can the remarkable details under different exposures merge image among the image of same low-dynamic range, obtains the low dynamic range echograms that can compare favourably with high dynamic range images result under the prerequisite that does not change dynamic range of images.Therefore can be widely applied to as digital camera, smart mobile phone etc. need to obtain in the electronic product of high dynamic range images effect fast.

In recent ten years, occurred the fusion method of the image that many kinds are different, generally speaking these fusion methods can be divided into two classes, that is: for the image co-registration of moving target with for the image interfusion method of static target.

For the image interfusion method of moving target, because the object in scene under different exposures moves, simple fused images can cause " ghost " effect.So the image interfusion method for moving target is different according to going the method for " ghost ", is subdivided into again two classes: directly remove " ghost " after merging the image that several exposures are different; Or the image that utilizes a width normal exposure degree obtains the under-exposure of same position through local histogram's stretching and overexposure image merges the two kinds of methods of result figure that generate again.

For the image interfusion method of static target, because the object in scene under different exposures is not mobile, so merge relatively simple.For integration technology difference, can be subdivided into three classes, that is: Multi-sensor Image Fusion, merges the detailed information of the same image that a plurality of different sensors take; Multiple focal length images merges, and merges the detailed information that a plurality of focal lengths are taken lower same image; Many exposures image co-registration, merges the detailed information of the same image of taking under a plurality of exposures.

, still there are a lot of problems in existing many exposures image fusion technology, image height dynamic image is synthetic easily produces colour vision distortion, and the color of fused images is difficult to correct, and in fused images, eigenvalue is improper easily brings distortion.

Summary of the invention

The object of this invention is to provide a kind of many exposures image interfusion method based on colour vision perception and the global quality factor, overcome existing many exposures image fusion technology and in high dynamic range images is synthetic, easily produce the problem of colour vision distortion.

For this reason, technical scheme of the present invention is: a kind of many exposures image interfusion method based on colour vision perception and the global quality factor is provided, has it is characterized in that: comprise the steps:

Step 1: the different image I of input n width exposure _n, utilize criteria of quality evaluation, i.e. contrast C _n, saturation degree S _nwith brightness B _ncalculate respectively the weight of different exposure images;

Wherein: n is the quantity of image, C _nrefer to contrast, S _nrefer to saturation degree, B _nthe brightness referring to, I _nrefer to n width image;

Step 2: according to the weighted value calculating in step 1, for the image I of different exposures _k, with linear method, obtain scalar weight map W respectively _{ij, k};

Wherein: W _{ij, k}corresponding scalar weight, subscript i, j, k refers to respectively pixel (i, j) in k width image;

Step 3: obtain original many exposures fused images according to original input picture and the fusion of scalar weight map:

The first step, the image I of utilizing Laplacian pyramid to input _k, obtain every layer corresponding utilize the weight map of gaussian pyramid decomposition input as W _{ij, k}, obtain every layer corresponding

Second step carries out image co-registration on every tomographic image, utilizes respectively n width image and its corresponding n standardized weight map as Alpha's mask;

The 3rd step, every layer of corresponding coefficient of hierarchical fusion;

The 4th step, decomposes the result figure L{R} obtaining ^lobtain final fused images R;

Wherein: l refers to the number of plies of the laplacian pyramid of image or the figure layer on gaussian pyramid;

Step 4: utilize global quality to evaluate the factor and select the width that meets colour vision perception most as with reference to image I in the image of the different exposures of n width of input _in;

Step 5: utilize with reference to image I _inbrightness value L _inwith chromatic value C _ininitial fused images R is carried out to many exposures image C that colour correction obtains final rectification _out.

In the criteria of quality evaluation that utilizes described in step 1, i.e. contrast C _n, saturation degree S _nwith brightness B _nthe weight that calculates respectively different exposure images, is calculated as follows:

C _n(i,j)=Y _n(i,j)*h(i,j)

Wherein: * is convolution symbol, Y _nit is the different image I of n width exposure of input _ngray level image, $Y_n=0.299\×I_n^r+0.587\×I_n^g+0.114\×I_n^b,$ N is the quantity of input picture, represent respectively the R on corresponding diagram layer, G, B value, Hi-pass filter h is defined as follows:

$h=\left(\begin{array}{ccc}0& -1& 0\\ -1& 4& -1\\ 0& -1& 0\end{array}\right)$

After having calculated for the local contrast of each pixel (i, j) in image, adopt the method for " winner complete " to draw the contrast of view picture figure as follows:

${\hat{C}}_n(i,j)=\left\{\begin{array}{c}1,\mathrm{if}{C}_n(i,j)=\max \{C_n(i,j),n=\mathrm{1,2},......n\}\\ 0,\mathrm{otherwise}\end{array}\right.;$

Saturation degree S _nby calculating respectively the R of each pixel (i, j), G, the standard deviation of B draws;

Brightness B _nbe to take 0.5 as standard judges, not have luminance difference distance in width image, the Gaussian curvature of trying to achieve is calculated, as follows:

$B_n(I,\mathrm{\σ})=\exp (-\frac{{(b-0.5)}^2}{{2\mathrm{\σ}}^2})$

Wherein: σ=0.2, b refers to the brightness value of pixel (i, j).

At the scalar weight map W described in step 2 _{ij, k}, by following formula, undertaken:

$W_{\mathrm{ij},k}={\left(C_{\mathrm{ij},k}\right)}^{W_c}\×{\left(S_{\mathrm{ij},k}\right)}^{W_s}\×{\left(B_{\mathrm{ij},k}\right)}^{W_B}$

Wherein: C, S, what B referred to respectively is contrast, saturation degree and brightness, its corresponding weight is respectively W _c, W _s, W _b; If ω=0, its corresponding measurement index is not counted in the calculating of final weight map.

As follows at every layer of corresponding coefficient method of the hierarchical fusion described in step 3:

${L\left\{R\right\}}_{\mathrm{ij}}^l=\underset{k=1}{\overset{n}{\mathrm{\Σ}}}{G\left\{\hat{W}\right\}}_{\mathrm{ij},k}^l{L\left\{I\right\}}_{\mathrm{ij},k}^l$

Wherein: the image I of Laplacian pyramid input _k, obtain every layer corresponding utilize the weight map of gaussian pyramid decomposition input as W _{ij, k}, obtain every layer corresponding n is the quantity of input picture, and l represents the corresponding diagram number of plies on laplacian pyramid or gaussian pyramid.

Described in step 4, utilizing global quality to evaluate the factor in the image of the different exposures of n width of input, to select the width that meets colour vision perception most as with reference to image I _inmethod as follows:

Col=0.10×Bri+0.50×Con+0.23×Det+0.12×Art

Wherein: Bri is the brightness of image, Con is the contrast of image, and Det is the details of image, and Art is the distorted portion of image, and the image of Col maximum is the reference picture of final selection.

In the utilization described in step 5 with reference to image I _inbrightness value L _inwith chromatic value C _ininitial fused images R is carried out to many exposures image C that colour correction obtains final rectification _outmethod as follows:

$C_{\mathrm{out}}=((\frac{C_{\mathrm{in}}}{L_{\mathrm{in}}}-1)r+1)L_{\mathrm{out}}$

Wherein: C _outthe many exposures fused images C=R through colour correction of output, G, B1; L _outthe brightness of original many exposures fused images; C _inthe colourity with reference to image of selecting according to the global quality factor; L _inbe the brightness with reference to image of selecting according to the global quality factor, r is that its span of coefficient is (to comprise 0.4 and 0.6) between 0.4～0.6.

Advantage of the present invention is: image, the remarkable details under different exposures merges among the image of same low-dynamic range method of the present invention, and utilize the global quality factor to select suitable with reference to figure, to carry out colour correction, under the prerequisite that does not change dynamic range of images, obtain the low dynamic range echograms that can compare favourably with high dynamic range images result.Mainly solve the synthetic middle colour vision problem of dtmf distortion DTMF producing of high dynamic range images, have the following advantages compared with the conventional method:

The colour vision cognitive method that optimal characteristics while 1, having proposed many exposures image co-registration is chosen.First utilize the overall image quality factor to judge the most important characteristics of image, finally utilize quality factor to select the optimal characteristics that needs fusion;

2, the color antidote of fused images has been proposed.The improper distortion bringing of eigenvalue in fused images that utilized colour correction correction.

Below with reference to accompanying drawing, the present invention is described in further details.

Accompanying drawing explanation

Fig. 1 is realization flow schematic diagram of the present invention.

The visual effect comparison diagram that Fig. 2 is the present invention and existing method on outdoor night image.

The visual effect comparison diagram that Fig. 3 is the present invention and existing method on outdoor daytime image.

The visual effect comparison diagram that Fig. 4 is the present invention and existing method on indoor night image.

The visual effect comparison diagram that Fig. 5 is the present invention and existing method on indoor daytime image.

Embodiment

Embodiment 1:

The invention provides as shown in Figure 1 a kind of many exposures image interfusion method based on colour vision perception and the global quality factor, remarkable details image under different exposures merges among the image of same low-dynamic range, and utilize the global quality factor to select suitable with reference to figure, to carry out colour correction, under the prerequisite that does not change dynamic range of images, obtain the low dynamic range echograms that can compare favourably with high dynamic range images result, its specific implementation step comprises as follows:

Step 1: the different image I of input n width exposure _n, utilize criteria of quality evaluation, i.e. contrast C _n, saturation degree S _nwith brightness B _ncalculate respectively the weight of different exposure images;

Wherein: n is the quantity of image, C _nrefer to contrast, S _nrefer to saturation degree, B _nthe brightness referring to, I _nrefer to n width image;

Step 2: according to the weighted value calculating in step 1, for the image I of different exposures _k, with linear method, obtain scalar weight map W respectively _{ij, k};

Wherein: W _{ij, k}corresponding scalar weight, subscript i, j, k refers to respectively pixel (i, j) in k width image;

Step 3: obtain original many exposures fused images according to original input picture and the fusion of scalar weight map:

The first step, the image I of utilizing Laplacian pyramid to input _k, obtain every layer corresponding utilize the weight map of gaussian pyramid decomposition input as W _{ij, k}, obtain every layer corresponding

Second step carries out image co-registration on every tomographic image, utilizes respectively n width image and its

N corresponding standardized weight map is as Alpha's mask;

The 3rd step, every layer of corresponding coefficient of hierarchical fusion;

The 4th step, decomposes the result figure L{R} obtaining ^lobtain final fused images R;

Wherein: l refer on the laplacian pyramid of image or or gaussian pyramid on the number of plies of figure layer; Described Alpha's mask is Alpha's mask, is with selected image, figure or object, and pending image (whole or local) is blocked to region or the processing procedure of coming control chart picture to process.Alpha is called mask or template for specific image or the object covering.In Digital Image Processing, mask is two-dimensional matrix array, sometimes also uses multivalue image.Be mainly used in extracting region of interest, with region of interest mask and the pending image made in advance, multiply each other, obtain Image with Region of Interest, in region of interest, image value remains unchanged, and outside district, image value is all 0.

Step 4: utilize global quality to evaluate the factor and select the width that meets colour vision perception most as with reference to image I in the image of the different exposures of n width of input _in;

Step 5: utilize with reference to image I _inbrightness value L _inwith chromatic value C _ininitial fused images R is carried out to many exposures image C that colour correction obtains final rectification _out.

In the criteria of quality evaluation that utilizes described in step 1, i.e. contrast C _n, saturation degree S _nwith brightness B _nthe weight that calculates respectively different exposure images, is calculated as follows:

C _n(i,j)=Y _n(i,j)*h(i,j)

Wherein: * is convolution symbol, Y _nit is the different image I of n width exposure of input _ngray level image, $Y_n=0.299\×I_n^r+0.587\×I_n^g+0.114\×I_n^b,$ N is the quantity of input picture, represent respectively the R on corresponding diagram layer, G, B value, Hi-pass filter h is defined as follows:

$h=\left(\begin{array}{ccc}0& -1& 0\\ -1& 4& -1\\ 0& -1& 0\end{array}\right)$

After having calculated for the local contrast of each pixel (i, j) in image, adopt the method for " winner complete " to draw the contrast of view picture figure as follows:

${\hat{C}}_n(i,j)=\left\{\begin{array}{c}1,\mathrm{if}{C}_n(i,j)=\max \{C_n(i,j),n=\mathrm{1,2},......n\}\\ 0,\mathrm{otherwise}\end{array}\right.;$

Saturation degree S _nby calculating respectively the R of each pixel (i, j), G, the standard deviation of B draws;

Brightness B _nbe to take 0.5 as standard judges, not have luminance difference distance in width image, the Gaussian curvature of trying to achieve is calculated, as follows:

$B_n(I,\mathrm{\σ})=\exp (-\frac{{(b-0.5)}^2}{{2\mathrm{\σ}}^2})$

Wherein: σ=0.2, b refers to the brightness value of pixel (i, j).

At the scalar weight map W described in step 2 _{ij, k}, by following formula, undertaken:

$W_{\mathrm{ij},k}={\left(C_{\mathrm{ij},k}\right)}^{W_c}\×{\left(S_{\mathrm{ij},k}\right)}^{W_s}\×{\left(B_{\mathrm{ij},k}\right)}^{W_B}$

Wherein: C, S, what B referred to respectively is contrast, saturation degree and brightness, its corresponding weight is respectively W _c, W _s, W _b; If ω=0, its corresponding measurement index is not counted in the calculating of final weight map.

As follows at every layer of corresponding coefficient method of the hierarchical fusion described in step 3:

${L\left\{R\right\}}_{\mathrm{ij}}^l=\underset{k=1}{\overset{n}{\mathrm{\Σ}}}{G\left\{\hat{W}\right\}}_{\mathrm{ij},k}^l{L\left\{I\right\}}_{\mathrm{ij},k}^l$

Wherein: the image I of Laplacian pyramid input _k, obtain every layer corresponding utilize the weight map of gaussian pyramid decomposition input as W _{ij, k}, obtain every layer corresponding n is the quantity of input picture, and l represents the corresponding diagram number of plies on laplacian pyramid or gaussian pyramid.

Described in step 4, utilizing global quality to evaluate the factor in the image of the different exposures of n width of input, to select the width that meets colour vision perception most as with reference to image I _inmethod as follows:

Col=0.10×Bri+0.50×Con+0.23×Det+0.12×Art

Wherein: Bri is the brightness of image, Con is the contrast of image, and Det is the details of image, and Art is the distorted portion of image, and the image of Col maximum is the reference picture of final selection.

In the utilization described in step 5 with reference to image I _inbrightness value L _inwith chromatic value C _ininitial fused images R is carried out to many exposures image C that colour correction obtains final rectification _outmethod as follows:

$C_{\mathrm{out}}=((\frac{C_{\mathrm{in}}}{L_{\mathrm{in}}}-1)r+1)L_{\mathrm{out}}$

Wherein: C _outthe many exposures fused images C=R through colour correction of output, G, B1, R refers to redness herein, and G refers to green, and B1 refers to blueness; L _outthe brightness of original many exposures fused images; C _inthe colourity with reference to image of selecting according to the global quality factor; L _inbe the brightness with reference to image of selecting according to the global quality factor, r is that its span of coefficient is (to comprise 0.4 and 0.6) between 0.4～0.6.

Embodiment 2:

Result of the present invention can further illustrate by following experiment:

1. experiment condition:

At central processing unit, be Intel (R) Pentium (R) Dual CPU, internal memory 16G, video card is NVIDIA Quadro NVS140M, operating system: on the platform of Windows Vista Home Basic x32Edition, carry out.

2. experiment content:

For the validity of checking this method, select four class images under 4 width different illumination conditions to test.Wherein, Fig. 2 is Monument image, outdoor night image, Fig. 3 is Plateau image, outdoor daytime image, Fig. 4 is Memorial image, indoor night image, Fig. 5 is Church image, indoor daytime image.

Below by 4 experiments, describe.

Experiment 1, carries out the comparison of visual effect by the present invention and existing additive method, and result is as Fig. 2, wherein:

Shown in Fig. 2 (a) is the result of original many exposures image co-registration,

Shown in Fig. 2 (b) is experimental result of the present invention,

Shown in Fig. 2 (c) is the result of histogram equalization,

Shown in Fig. 2 (d) is the result of the function of the imadjust that proposes in matlab software.

As seen from Figure 2, the present invention can effectively retain tone sense organ and the detailed information of image, and it is more naturally clear that Output rusults image is compared with other existing methods.And original many exposures Image Fusion cannot be complete the color detail of reservation image, the function that matlab provides cannot retain the hue information of image, algorithm of histogram equalization can cause the image fault of bringing due to a large amount of artificial information.In sum, the image nature that these several existing methods produce not as the present invention.

Experiment 2, carries out the comparison of visual effect by the present invention and existing additive method, and result is as Fig. 3, wherein:

Shown in Fig. 3 (a) is the result of original many exposures image co-registration,

Shown in Fig. 3 (b) is experimental result of the present invention,

Shown in Fig. 3 (c) is the result of histogram equalization,

Shown in Fig. 3 (d) is the result of the function of the imadjust that proposes in matlab software.

As seen from Figure 3, the present invention can effectively retain tone sense organ and the detailed information of image, and it is more naturally clear that Output rusults image is compared with other existing methods.And original many exposures Image Fusion cannot be complete the color detail of reservation image, the function that matlab provides cannot retain the hue information of image, algorithm of histogram equalization can cause the image fault of bringing due to a large amount of artificial information.In sum, the image nature that these several existing methods produce not as the present invention.

Experiment 3, carries out the comparison of visual effect by the present invention and existing additive method, and result is as Fig. 4, wherein:

Shown in Fig. 4 (a) is the result of original many exposures image co-registration,

Shown in Fig. 4 (b) is experimental result of the present invention,

Shown in Fig. 4 (c) is the result of histogram equalization,

Shown in Fig. 4 (d) is the result of the function of the imadjust that proposes in matlab software.

As seen from Figure 4, the present invention can effectively retain tone sense organ and the detailed information of image, and it is more naturally clear that Output rusults image is compared with other existing methods.And original many exposures Image Fusion cannot be complete the color detail of reservation image, the function that matlab provides cannot retain the hue information of image, algorithm of histogram equalization can cause the image fault of bringing due to a large amount of artificial information.In sum, the image nature that these several existing methods produce not as the present invention.

Experiment 4, carries out the comparison of visual effect by the present invention and existing additive method, and result is as Fig. 5, wherein:

Shown in Fig. 5 (a) is the result of original many exposures image co-registration,

Shown in Fig. 5 (b) is experimental result of the present invention,

Shown in Fig. 5 (c) is the result of histogram equalization,

Shown in Fig. 5 (d) is the result of the function of the imadjust that proposes in matlab software.

The present invention can effectively retain tone sense organ and the detailed information of image, and it is more naturally clear that Output rusults image is compared with other existing methods.And original many exposures Image Fusion cannot be complete the color detail of reservation image, the function that matlab provides cannot retain the hue information of image, algorithm of histogram equalization can cause the image fault of bringing due to a large amount of artificial information.In sum, the image nature that these several existing methods produce not as the present invention.

For further checking validity of the present invention, used respectively two kinds of methods of subjective evaluation and objective evaluation, the results are shown in Table 1 and table 2.

The average suggestion value of table 1 test pattern (Mean Opinion Score)

The objective evaluation result of table 2 test pattern

By form, can be found out, the method that the present invention proposes can obtain the low dynamic range echograms that can compare favourably with high dynamic range images result under the prerequisite that does not change dynamic range of images.

More than exemplifying is only to illustrate of the present invention, does not form the restriction to protection scope of the present invention, within the every and same or analogous design of the present invention all belongs to protection scope of the present invention.

Claims (6)

1. the many exposures image interfusion method based on colour vision perception and the global quality factor, is characterized in that: comprise the steps:

Step 1: the different image I of input n width exposure _n, utilize criteria of quality evaluation, i.e. contrast C _n, saturation degree S _nwith brightness B _ncalculate respectively the weight of different exposure images;

Wherein: n is the quantity of image, C _nrefer to contrast, S _nrefer to saturation degree, B _nthe brightness referring to, I _nrefer to n width image;

Step 2: according to the weighted value calculating in step 1, for the image I of different exposures _k, with linear method, obtain scalar weight map W respectively _{ij, k};

Wherein: W _{ij, k}corresponding scalar weight, subscript i, j, k refers to respectively pixel (i, j) in k width image;

Step 3: obtain original many exposures fused images according to original input picture and the fusion of scalar weight map:

The first step, the image I of utilizing Laplacian pyramid to input _k, obtain every layer of correspondence

's utilize the weight map of gaussian pyramid decomposition input as W _{ij, k}, obtain every layer corresponding

Second step carries out image co-registration on every tomographic image, utilizes respectively n width image and its corresponding n standardized weight map as Alpha's mask;

The 3rd step, every layer of corresponding coefficient of hierarchical fusion;

The 4th step, decomposes the result figure L{R} obtaining ^lobtain final fused images R;

Wherein: l refers on the laplacian pyramid of image or the number of plies of the figure layer on gaussian pyramid;

Step 4: utilize global quality to evaluate the factor and select the width that meets colour vision perception most as with reference to image I in the image of the different exposures of n width of input _in;

Step 5: utilize with reference to image I _inbrightness value L _inwith chromatic value C _ininitial fused images R is carried out to many exposures image C that colour correction obtains final rectification _out.

2. a kind of many exposures image interfusion method based on colour vision perception and the global quality factor as claimed in claim 1, is characterized in that: described in step 1, utilize criteria of quality evaluation, i.e. contrast C _n, saturation degree S _nwith brightness B _nthe weight that calculates respectively different exposure images, is calculated as follows:

C _n(i,j)=Y _n(i,j)*h(i,j)

Wherein: * is convolution symbol, Y _nit is the different image I of n width exposure of input _ngray level image, $Y_n=0.299\×I_n^r+0.587\×I_n^g+0.114\×I_n^b,$ N is the quantity of input picture, represent respectively the R on corresponding diagram layer, G, B value, Hi-pass filter h is defined as follows:

$h=\left(\begin{array}{ccc}0& -1& 0\\ -1& 4& -1\\ 0& -1& 0\end{array}\right)$

After having calculated for the local contrast of each pixel (i, j) in image, adopt the method for " winner complete " to draw the contrast of view picture figure as follows:

${\hat{C}}_n(i,j)=\left\{\begin{array}{c}1,\mathrm{if}{C}_n(i,j)=\max \{C_n(i,j),n=\mathrm{1,2},......n\}\\ 0,\mathrm{otherwise}\end{array}\right.;$

Saturation degree S _nby calculating respectively the R of each pixel (i, j), G, the standard deviation of B draws;

Brightness B _nbe to take 0.5 as standard judges, not have luminance difference distance in width image, the Gaussian curvature of trying to achieve is calculated, as follows:

$B_n(I,\mathrm{\σ})=\exp (-\frac{{(b-0.5)}^2}{{2\mathrm{\σ}}^2})$

Wherein: σ=0.2, b refers to the brightness value of pixel (i, j).

3. a kind of many exposures image interfusion method based on colour vision perception and the global quality factor as claimed in claim 1, is characterized in that: the scalar weight map W described in step 2 _{ij, k}, by following formula, undertaken:

$W_{\mathrm{ij},k}={\left(C_{\mathrm{ij},k}\right)}^{W_c}\×{\left(S_{\mathrm{ij},k}\right)}^{W_s}\×{\left(B_{\mathrm{ij},k}\right)}^{W_B}$

Wherein: C, S, what B referred to respectively is contrast, saturation degree and brightness, its corresponding weight is respectively W _c, W _s, W _b; If ω=0, its corresponding measurement index is not counted in the calculating of final weight map.

4. a kind of many exposures image interfusion method based on colour vision perception and the global quality factor as claimed in claim 1, is characterized in that: every layer of corresponding coefficient method of the hierarchical fusion described in step 3 is as follows:

${L\left\{R\right\}}_{\mathrm{ij}}^1=\underset{k=1}{\overset{n}{\mathrm{\Σ}}}{G\left\{\hat{W}\right\}}_{\mathrm{ij},k}^1{L\left\{I\right\}}_{\mathrm{ij},k}^1$

Wherein: the image I of Laplacian pyramid input _k, obtain every layer corresponding utilize the weight map of gaussian pyramid decomposition input as W _{ij, k}, obtain every layer corresponding n is the quantity of input picture, and l represents the corresponding diagram number of plies on laplacian pyramid or gaussian pyramid.

5. a kind of many exposures image interfusion method based on colour vision perception and the global quality factor as claimed in claim 1, is characterized in that: described in step 4, utilize global quality to evaluate the factor in the image of the different exposures of n width of input, to select the width that meets colour vision perception most as with reference to image I _inmethod as follows:

Col=0.10×Bri+0.50×Con+0.23×Det+0.12×Art

Wherein: Bri is the brightness of image, Con is the contrast of image, and Det is the details of image, and Art is the distorted portion of image, and the image of Col maximum is the reference picture of final selection.

6. a kind of many exposures image interfusion method based on colour vision perception and the global quality factor as claimed in claim 1, is characterized in that: the utilization described in step 5 is with reference to image I _inbrightness value L _inwith chromatic value C _ininitial fused images R is carried out to many exposures image C that colour correction obtains final rectification _outmethod as follows:

$C_{\mathrm{out}}=((\frac{C_{\mathrm{in}}}{L_{\mathrm{in}}}-1)r+1)L_{\mathrm{out}}$

Wherein: C _outthe many exposures fused images C=R through colour correction of output, G, B1; L _outthe brightness of original many exposures fused images; C _inthe colourity with reference to image of selecting according to the global quality factor; L _inbe the brightness with reference to image of selecting according to the global quality factor, r is that its span of coefficient is (to comprise 0.4 and 0.6) between 0.4～0.6.