CN1293515C - Apparatus and method for determining exposure settings for images with human face regions - Google Patents

Abstract

Translated fromChinese

本发明提供一种确定将被提取的影像的曝光设定，该影像关于一呈现人脸区域的景象。多个似人脸轮廓图案事先提供。基于一准则判断出大致上指向该景象的人脸区域。判断关于该影像的一曝光测光矩阵中指向该景象的人脸区域的特定单元，并且加重该特定单元的权值，进而调整该曝光测光矩阵。根据该经调整过的曝光测光矩阵，确定该影像的曝光设定。

The present invention provides a method for determining exposure settings of an image to be extracted, the image being about a scene showing a human face region. A plurality of human face contour patterns are provided in advance. A human face region roughly pointing to the scene is determined based on a criterion. A specific unit pointing to the human face region of the scene in an exposure metering matrix for the image is determined, and the weight of the specific unit is increased, thereby adjusting the exposure metering matrix. The exposure settings of the image are determined based on the adjusted exposure metering matrix.

Description

Determine to have the device and method of exposure settings of the image of human face region

Technical field

The present invention relates to a kind of device and method of exposure settings of definite image, particularly about the device and method of the exposure settings of the image of the scene that presents human face region.

Background technology

The digitized video extraction element will make image definition distinguish when extracting image, just correct exposure settings must be arranged.The exposure settings of digitized video extraction element is decided according to the result of metering mode (metering mode).Matrix photometry (matrix metering) then is one of the most normal metering mode that is used.The matrix photometry is that the video that the digitized video extraction element is extracted is divided into a plurality of unit and forms a photometry matrix, analyzes the light power of a plurality of different units in the photometry matrix then.The digitized video extraction element carries out the distribution of photometry weights according to analysis result, obtains a photometry weight matrix.

Though the digitized video extraction element can be so that the scene of being extracted obtains the image of correct exposure by metering mode.But, if have the existence of people's face in the scene of extracting, still can be excessive because of the contrast of scene light and shade, or the photometry weights of people's face position distribute lower, often cause in the image incorrect for people's face exposure of main body, and can't obtain one make us the image photo of subjective satisfaction.Therefore, be necessary to do a suitable adjustment, make people's face obtain correct exposure at the exposure weights of people's face position in the scene.

Summary of the invention

Fundamental purpose of the present invention provides a kind of device and method that is used for the exposure settings in the digitized video extraction element, image for the scene that presents human face region (human face area), dynamically adjust the exposure settings of this digitized video extraction element, make people's face obtain correct exposure to improve the shortcoming of prior art.

A kind of exposure settings device that is used in the Equipment for forming image (Image forming apparatus) according to the present invention, this exposure settings device is in order to determine the exposure settings of first image that will be extracted.This first image that will be extracted is about a scene (Scene) that presents human face region.This device comprises, storage module, image extraction module, exposure control module, first processing module, first analysis module, second processing module and second analysis module.

This storage module prestores a plurality of like facial contour pattern (Human-face-contour-likepattern).This image extraction module is in order to extracting the scene that this presents human face region, and then produces one second image.This exposure control module produces an exposure photometry matrix (Exposure metering matrix) corresponding to second image of this extraction based on second image of this extraction.This first processing module produces second image that at least one dwindles sampling (Down-sampled) in order to second image according to this extraction, and at least one dwindles second image of sampling and takes out a first information from second image of this extraction and this respectively.This first analysis module is a plurality of like the facial contour pattern based on this, at least one dwindles the first information of second image of sampling to analyze second image of this extraction and this respectively, and then judges the first area that may point to the human face region of this scene in second image of this extraction.This second processing module is taken out at least one second information from the first area of second image of this extraction.This second analysis module is analyzed this at least one second information based at least one criterion, and then judges second area from the first area of second image of this extraction.This second area points to the human face region of this scene haply.This second analysis module and according to the second area of second image of this extraction determines to point in this exposure photometry matrix the discrete cell of the human face region of this scene.

This exposure control module increases the weight of the weights of this discrete cell, and then adjust this exposure photometry matrix, this exposure control module and according to the exposure photometry matrix of this adjustment, with the exposure settings of definite this first image that will be extracted, and control this image extraction module and extract this first image according to this exposure settings.

So exposure settings device and method of the present invention, can make the image of digitized video extraction element for the scene that presents human face region (human face area), dynamically adjust the exposure settings of this digitized video extraction element, make that people's face position obtains correct exposure in the scene.

Can be further understood by following detailed Description Of The Invention and appended accompanying drawing about the advantages and spirit of the present invention.

Description of drawings

Fig. 1 is used in the functional block diagram of the exposure settings device of an Equipment for forming image for the present invention.

A plurality of synoptic diagram that Fig. 2 prestores for storage module among Fig. 1 like the facial contour pattern.

Two synoptic diagram that dwindle second image of sampling that Fig. 3 is produced for first processing module among Fig. 1.

Fig. 4 is the second area synoptic diagram of second image among Fig. 3.

Fig. 5 determines in the scene process flow diagram about the exposure settings method of the image that presents human face region for the present invention.

Embodiment

At first the shown by reference numeral among the present invention is described, wherein, 10: the exposure settings device; 11: scene; 12: storage module; 13: the first images; 14: the image extraction module; 16: the exposure control module; 18: the first processing modules; 20: the first analysis modules; 22: the second processing modules; 24: the second analysis modules; 26,26a, 26b, 26c, 26d: like the facial contour pattern; 28: the second images; 30: exposure photometry matrix; 32,32a, 32b: second image that dwindles sampling; 34,34a, 34b, 34c: the first information; 36,36a, 36b, 36c, 36d: first area; 38: the second information; 40,40a, 40b: second area; 42: the exposure photometry matrix of adjustment; 44: control signal.

See also Fig. 1 and Fig. 2, Fig. 1 is used in the functional block diagram of the exposure settings device 10 of an Equipment for forming image for the present invention.A plurality of synoptic diagram that Fig. 2 prestores for storage module 12 among Fig. 1 like facial contour pattern 26.Exposure settings device 10 of the present invention is in order to determining the exposure settings of first image 13 that will be extracted, and this first image 13 that will be extracted is about a scene 11 (Scene) that presents human face region.Exposure settings device 10 comprises a storage module 12, an image extraction module 14, an exposure control module 16, one first processing module 18, one first analysis module 20, one second processing module 22 and one second analysis module 24.

Storage module 12 prestores a plurality of like facial contour pattern 26, and wherein the saving format like facial contour pattern 26 is binary two-dimensional matrix, and " 1 " represents facial contour (i.e. black line part shown in the figure two), and " 0 " then represents non-face profile.And will be like facial contour pattern 26 inputs first analysis module 20.In example shown in Figure 2, then prestore four likefacial contour pattern 26a, 26b, 26c, 26d.Image extraction module 14 is in order to extracting this scene that presents human face region 11, and then produces one second image 28.Exposure control module 16 produces an exposure photometry matrix 30 (Exposure metering matrix) corresponding tosecond image 28 that extracts based onsecond image 28 that extracts.

First processing module 18 is dwindled second image 32 of sampling in order to produce at least one according tosecond image 28 that extracts, and takes out a first information 34 respectively fromsecond image 28 that extracts and second image 32 that at least one dwindles sampling.

First analysis module 20 is based on a plurality of seeminglyfacial contour pattern 26a, 26b, 26c, 26d,second image 28 that analyze to extract respectively and the first information 34 of second image 32 that at least one dwindles sampling, and then judge the first area 36 that may point to human face region in the scene 11 insecond image 28 of extraction.

Second processing module 22 is taken out at least one second information 38 from the first area 36 ofsecond image 28 of extraction.

Second analysis module 24 is based at least one criterion, analyze at least one second information 38, and then from the first area 36 ofsecond image 28 that extracts, judge second area 40, second area 40 points to human face region in the scene 11 haply, second analysis module 24 and according to the second area 40 ofsecond image 28 that extracts determines to point in the exposure photometry matrix 30 discrete cell of human face region in the scene 11.

Exposure control module 16 increases the weight of the weights of this discrete cell, and then adjusts exposure photometry matrix 30.Exposure control module 16 and according to the exposure photometry matrix of adjusting 42 is with the exposure settings of first image 13 determining to be extracted.And exposure control module 16 outputs one control signal 44 is extracted this first image 13 with control image extraction module 14 according to this exposure settings.

See also Fig. 3, Fig. 3 is two that first processing module 18 is produced among Fig. 1 and dwindles the second image 32a of sampling, the synoptic diagram of 32b.In example shown in Figure 3,second image 28 produces two the second image 32a, 32b that dwindle sampling after 18 computings of first processing module.

See also Fig. 3, first processing module 18 is taken out Y (brightness) data fromsecond image 28 and two the second image 32a, 32b that dwindle sampling, and after high frequency filter (High-pass Filter) extraction image contour data and dualization processing, can produce first information 34a, 34b, 34c that as shown in Figure 3 " 0 " and " 1 " distributes, " 1 " represents the position of contour of object in the image, and " 0 " then represents the position of non-contour of object.

As shown in Figures 2 and 3, first analysis module 20 likefacial contour pattern 26a, 26b, 26c, 26d, is analyzedsecond image 28 and two the second image 32a, first information 34a, the 34b of 32b, the 34c that dwindle sampling that extract according to a plurality of respectively.Because the comparatively strong cause of borderline region brightness contrast of object and background, therefore the borderline region of object and background can produce the juncture area of as shown in Figure 3 " 0 " and " 1 " in first information 34a, 34b, 34c, according to the shape of this juncture area and a plurality of making comparisons, and judgefirst area 36a, 36b, 36c, the 36d that may point to human face region in the scene 11 insecond image 28 of extraction likefacial contour pattern 26a, 26b, 26c, 26d by storage module 12 inputs.

In addition, take out the data of Cb (blue poor) and Cr (red poor) second processing module 22first area 36a, 36b, 36c, the 36d, and produce second information 38 fromsecond image 28 of extraction.

See also Fig. 4, Fig. 4 is second area 40 synoptic diagram ofsecond image 28 among Fig. 3.After second information 38 produced, second analysis module 24 was analyzed second information 38 based on this criterion, and then judgedsecond area 40a, 40b fromfirst area 36a, the 36b ofsecond image 28 that extracts, 36c, 36d.This criterion is to judge whetherfirst area 36a, the 36b insecond image 28, the color of 36c, 36d approach the colour of skin of people's face.This criterion is defined as follows:

-33≤Cb≤-13；

19≤Cr≤39

Among the embodiment of Fig. 3 and Fig. 4, comprise two human face regions and two balloon zones about the scene 11 (Scene) that presents human face region.Second analysis module 24 is based on above-mentioned criterion, judgessecond area 40a, 40b fromfirst area 36a, the 36b ofsecond image 28 that extracts, 36c, 36d.Second area 40a, 40b point to human face region in the scene 11 haply.And twofirst area 36c, 36d that point to the balloon of scene 11 haply in left side among Fig. 4 because it does not meet the criterion of judgement, therefore can not be judged as second area 40.Second analysis module 24 and according tosecond area 40a, the 40b ofsecond image 28 that extracts determines to point in the exposure photometry matrix 30 discrete cell of human face region in the scene 11.

Therefore, exposure control module 16 is by increasing the weight of by the weights that point to the discrete cell of human face region in the scene 11 in second analysis module, the 24 determined exposure photometry matrixes 30, and then adjusts exposure photometry matrix 30.Exposure control module 16 and according to the exposure photometry matrix of adjusting 42, with the exposure settings of definite this first image 13 that will be extracted, and the exposure settings that output control signal 44 is exported according to exposure control modules 16 with control image extraction module 14 is extracted first image 13.

See also Fig. 5, Fig. 5 determines in the scene 11 process flow diagram about the exposure settings method of the image that presents human face region for the present invention.Below detailed description is applied to the establishing method of the exposure settings device 10 of the present invention of Fig. 1.Exposure settings method of the present invention comprises the following step:

S50:, determine the exposure photometry matrix 30 of first image 13 that this will be extracted based on a predetermined logic.

S52: extract onesecond image 28, and at least one dwindles second image 32 of sampling based on second image, 28 generations of extracting about scene 11.

S54: fromsecond image 28 that extracts and second image 32 that at least one dwindles sampling, take out Y (brightness) data respectively, and extract image contour data and dualization through high frequency filter and handle the back and produce a first information 34.

S56: based on a plurality of seemingly facial contour patterns 26,second image 28 that analyze to extract respectively and the first information 34 of second image 32 that at least one dwindles sampling, and then judge the first area 36 that may point to human face region in the scene 11 insecond image 28 of extraction.

S58: take out the data of Cb (blue poor) and Cr (redness is poor) from the first area 36 ofsecond image 28 that extracts, and produce one second information 38.

S60: based on the criterion shown in the following equation, analyze at least one second information 38 ofsecond image 28 that extracts, and then from the first area 36 ofsecond image 28 that extracts, judge second area 40, second area 40 is to point to human face region in the scene 11 haply.The equation of relevant this criterion is as follows:

-33≤Cb≤-13；

19≤Cr≤39

S62:, determine to point in the exposure photometry matrix 30 discrete cell of human face region in the scene 11 according to the second area 40 ofsecond image 28 that extracts.

S64: increase the weight of the weights of this discrete cell, and then adjust exposure photometry matrix 30.

S66:, determine the exposure settings of first image 13 that this will be extracted according to the exposure photometry matrix of adjusting 42.

Therefore, according to exposure settings device 10 of the present invention and method, the digitized video extraction element can dynamically be adjusted the exposure settings of this digitized video extraction element for the image of the scene 11 that presents human face region (human face area), makes that people's face position obtains correct exposure in the scene 11.

By the above detailed description of preferred embodiments, hope can be known description feature of the present invention and spirit more, and is not to come category of the present invention is limited with above-mentioned disclosed preferred embodiment.On the contrary, its objective is that hope can contain in the category of claim of being arranged in of various changes and tool equality institute of the present invention desire application.

Claims (12)

Translated fromChinese

1.一种使用于影像形成装置内的曝光设定装置，该曝光设定装置用以确定将被提取的第一影像的曝光设定，该第一影像关于一呈现人脸区域的景象，该曝光设定装置包含：1. An exposure setting device used in an image forming device, the exposure setting device is used to determine the exposure setting of a first image to be extracted, the first image is about a scene showing a human face area, the The exposure setting device consists of:一储存模块，预存多个似人脸轮廓图案；A storage module, which pre-stores a plurality of face-like contour patterns;一影像提取模块，用以提取该呈现人脸区域的景象，进而产生一第二影像；An image extraction module, used to extract the scene showing the human face area, and then generate a second image;一曝光控制模块，基于该提取的第二影像产生一相应于该提取的第二影像的曝光测光矩阵；an exposure control module for generating an exposure metering matrix corresponding to the extracted second image based on the extracted second image;一第一处理模块，用以根据该提取的第二影像产生至少一个缩小采样的第二影像，并且分别从该提取的第二影像以及该至少一个缩小采样的第二影像中取出一第一信息；A first processing module, configured to generate at least one reduced-sampled second image based on the extracted second image, and extract a first information from the extracted second image and the at least one reduced-sampled second image ;一第一分析模块，基于该多个似人脸轮廓图案，分别分析该提取的第二影像以及该至少一个缩小采样的第二影像的第一信息，进而判断出该提取的第二影像中可能指向该景象的人脸区域的第一区域；A first analysis module, based on the plurality of human-like contour patterns, respectively analyze the first information of the extracted second image and the at least one downsampled second image, and then determine that the extracted second image may Pointing to the first area of the face area of the scene;一第二处理模块，从该提取的第二影像的第一区域取出至少一个第二信息；以及a second processing module, extracting at least one second information from the first region of the extracted second image; and一第二分析模块，基于至少一个准则，分析该至少一个第二信息，进而从该提取的第二影像的第一区域中判断出第二区域，该第二区域大致上指向该景象的人脸区域，该第二分析模块并且根据该提取的第二影像的第二区域，确定该曝光测光矩阵中指向该景象的人脸区域的特定单元，A second analyzing module, analyzing the at least one second information based on at least one criterion, and then judging a second area from the extracted first area of the second image, the second area generally pointing to the face of the scene area, the second analyzing module and according to the extracted second area of the second image, determine a specific cell in the exposure metering matrix pointing to the face area of the scene,其中，该曝光控制模块加重该特定单元的权值，进而调整该曝光测光矩阵，该曝光控制模块并且根据该调整的曝光测光矩阵，以确定该将被提取的第一影像的曝光设定，并且控制该影像提取模块依据该曝光设定提取该第一影像。Wherein, the exposure control module increases the weight of the specific unit, and then adjusts the exposure metering matrix, and the exposure control module determines the exposure setting of the first image to be extracted according to the adjusted exposure metering matrix , and controlling the image capture module to capture the first image according to the exposure setting.2.如权利要求1所述的装置，其中该提取的第二影像以及该至少一个缩小采样的第二影像的第一信息是从中取出的亮度Y数据并且经过高频滤波器提取影像轮廓数据及二元化处理。2. The apparatus as claimed in claim 1, wherein the first information of the extracted second image and the at least one downsampled second image is luminance Y data extracted therefrom and subjected to a high-frequency filter to extract image contour data and Binary processing.3.如权利要求2所述的装置，其中该提取的第二影像的至少一个第二信息包含从中取出的蓝色差Cb数据。3. The apparatus of claim 2, wherein the extracted at least one second information of the second image includes blue color difference Cb data extracted therefrom.4.如权利要求3所述的装置，其中该至少一个准则包含一准则，其定义如下：4. The apparatus of claim 3, wherein the at least one criterion comprises a criterion defined as follows:-33≤Cb≤-13。-33≤Cb≤-13.5.如权利要求2所述的装置，其中该提取的第二影像的至少一个第二信息包含从中取出的红色差Cr数据。5. The apparatus of claim 2, wherein the extracted at least one second information of the second image includes red color difference Cr data extracted therefrom.6.如权利要求5所述的装置，其中该至少一个准则包含一准则，其定义如下：6. The apparatus of claim 5, wherein the at least one criterion comprises a criterion defined as follows:19≤Cr≤39。19≤Cr≤39.7.一种用以确定将被提取的第一影像的曝光设定的方法，该将被提取的第一影像关于一呈现人脸区域的景象，该方法包含下列步骤：7. A method for determining an exposure setting for a first image to be extracted, the first image to be extracted with respect to a scene showing a region of a human face, the method comprising the steps of:提供多个似人脸轮廓图案；Provide multiple human-like contour patterns;基于一预定逻辑，确定该将被提取的第一影像的一曝光测光矩阵；determining an exposure metering matrix of the first image to be extracted based on a predetermined logic;提取关于该景象的一第二影像，并且基于该提取的第二影像产生至少一个缩小采样的第二影像；extracting a second image of the scene, and generating at least one downsampled second image based on the extracted second image;分别从该提取的第二影像以及该至少一个缩小采样的第二影像中取出一第一信息；extracting a first information from the extracted second image and the at least one downsampled second image respectively;基于该多个似人脸轮廓图案，分别分析该提取的第二影像以及该至少一个缩小采样的第二影像的第一信息，进而判断出该提取的第二影像中可能指向该景象的人脸区域的第一区域；Based on the plurality of human-face-like contour patterns, respectively analyzing the first information of the extracted second image and the at least one downsampled second image, and then judging the human face that may point to the scene in the extracted second image the first zone of the zone;从该提取的第二影像的第一区域取出至少一个第二信息；fetching at least one second information from the first region of the extracted second image;基于至少一个准则，分析该提取的第二影像的至少一个第二信息，进而从该提取的第二影像的第一区域中判断出第二区域，该第二区域大致上指向该景象的人脸区域；analyzing at least one second information of the extracted second image based on at least one criterion, and then determining a second area from the first area of the extracted second image, the second area generally pointing to a human face of the scene area;根据该提取的第二影像的第二区域，确定该曝光测光矩阵中指向该景象的人脸区域的特定单元；determining, based on the extracted second region of the second image, a specific cell in the exposure metering matrix directed to the face region of the scene;加重该特定单元的权值，进而调整该曝光测光矩阵；以及weighting the particular cell, thereby adjusting the exposure metering matrix; and根据该经调整过的曝光测光矩阵，确定该将被提取的第一影像的曝光设定。An exposure setting of the first image to be extracted is determined according to the adjusted exposure metering matrix.8.如权利要求7所述的方法，其中该提取的第二影像以及该至少一个缩小采样的第二影像的第一信息是从中取出的亮度Y数据并且经过高频滤波器提取影像轮廓数据及二元化处理。8. The method as claimed in claim 7, wherein the first information of the extracted second image and the at least one downsampled second image is luminance Y data extracted therefrom and subjected to a high frequency filter to extract image contour data and Binary processing.9.如权利要求8所述的方法，其中该提取的第二影像的至少一个第二信息包含从中取出的蓝色差Cb数据。9. The method of claim 8, wherein the extracted at least one second information of the second image includes blue color difference Cb data extracted therefrom.10.如权利要求9所述的方法，其中该至少一个准则包含一准则，其定义如下：10. The method of claim 9, wherein the at least one criterion comprises a criterion defined as follows:-33≤Cb≤-13。-33≤Cb≤-13.11.如权利要求8所述的方法，其中该提取的第二影像的至少一个第二信息包含从中取出的红色差Cr数据。11. The method of claim 8, wherein the extracted at least one second information of the second image includes red color difference Cr data extracted therefrom.12.如权利要求11所述的方法，其中该至少一个准则包含一准则，定义如下：12. The method of claim 11 , wherein the at least one criterion comprises a criterion defined as follows:19≤Cr≤39。19≤Cr≤39.

Images