CN104182721A - Image processing system and image processing method capable of improving face identification rate - Google Patents
Image processing system and image processing method capable of improving face identification rate Download PDFInfo
- Publication number
- CN104182721A CN104182721A CN201310192437.5A CN201310192437A CN104182721A CN 104182721 A CN104182721 A CN 104182721A CN 201310192437 A CN201310192437 A CN 201310192437A CN 104182721 A CN104182721 A CN 104182721A
- Authority
- CN
- China
- Prior art keywords
- image
- face
- image processing
- mentioned
- default zone
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Abstract
The invention discloses an image processing system and an image processing method, which are capable of improving a face identification rate. The method includes the following steps: firstly, acquiring an image which includes a face; next, carrying out face detection on a default zone in the image; when face detection cannot be completed, adjusting an exposure value of the default zone so as to enable the exposure value of the default zone to conform to an expectation value; when the luminance of the default zone conforms to the expectation value and the face detection is completed, analyzing image information in a face zone of the image; and next, according to the image information, selecting a model parameter and an image adjustment parameter corresponding to the model parameter and carrying out image processing on the face zone in the image and then outputting the processed image to the face identification system. Through the technical scheme, an image which is acquired under a backlight or low-illuminance environment can be enhanced to a bright and clear output image so as to be used by a rear-end face identification system so that the successful detection and identification rate can be improved significantly.
Description
Technical field
The present invention relates to a kind of image processing system, relate in particular to a kind of image processing system and image processing method that promotes recognition of face rate.
Background technology
Portrait face detection can be applied to man-machine interface (human computer interface), domestic video monitoring (home video surveillance) or the face recognition of biological detection and image database management technology thereof, visible portrait face detection occupies the role who becomes more and more important in image recognition technology, is also that one of technology of research and development is endeavoured in each large scientific and technological research centre at present.But, recognition of face have higher difficulty and many changes because of, similarly be under the environment of backlight or low-light (level), face identification system often cannot gather enough biological attribute datas.
As can be seen here, not yet full maturity of portrait face detection at present, and be further improved.In order to address the above problem, association area is sought solution one after another painstakingly, has developed but have no for a long time applicable mode always.
Summary of the invention
The invention provides a kind of image processing system and method that promotes recognition of face success ratio.
First image processing method provided by the present invention comprises the following steps:, gathers the image that contains face; Then, the default zone in image carries out face detection; In the time cannot completing face detection, adjust the exposure value of default zone, make the exposure value of default zone meet expectation value; In the time that the brightness of default zone meets expectation value and completes face detection, the image information of the human face region in analysis image; Then, according to image information, the corresponding image adjustment parameter of Selection Model parameter and model parameter, and the human face region in image is carried out to image processing, then export image after treatment to face identification system.
Image processing system provided by the present invention comprises camera, face detection unit, exposure adjustment unit, image analyzing unit, graphics processing unit and image output unit.Camera is in order to gather the image that contains face, and face detection unit carries out face detection in order to the default zone in image.In the time cannot completing face detection, exposure adjustment unit, in order to adjust the exposure value of default zone, makes the exposure value of default zone meet expectation value.In the time that the exposure value of default zone meets expectation value and completes face detection, image analyzing unit is in order to the image information in the human face region of analysis image.Graphics processing unit is in order to according to image information, the corresponding image adjustment parameter of Selection Model parameter and model parameter, and the human face region in image is carried out to image processing.Image output unit is in order to export image after treatment to face identification system.
In sum, technical scheme of the present invention compared with prior art has obvious advantage and beneficial effect.By technique scheme, can be by the image gathering under the environment in backlight or low-light (level), be enhanced to bright and output image clearly, use for rear end face identification system, so can significantly improve and detect and the probalility of success of identification.
To above-mentioned explanation be explained in detail with embodiment below, and technical scheme of the present invention is provided further and explained.
Brief description of the drawings
Fig. 1 is the process flow diagram according to a kind of image processing method that promotes recognition of face rate of one embodiment of the invention;
Fig. 2 is the schematic diagram according to the default zone in the image of one embodiment of the invention;
Fig. 3 is the schematic according to the image information of one embodiment of the invention;
Fig. 4 is a kind of image processing system block diagram that promotes recognition of face rate according to one embodiment of the invention;
Fig. 5 A, Fig. 5 B, Fig. 5 C are the image acquisition pictures according to one embodiment of the invention;
Fig. 6 A, Fig. 6 B, Fig. 6 C are the image acquisition pictures according to another embodiment of the present invention.
Embodiment
In order to make narration of the present invention more detailed and complete, can be with reference to accompanying drawing and the various embodiment of the following stated, in accompanying drawing, identical number represents same or analogous element.On the other hand, well-known element and step are not described in an embodiment, to avoid that the present invention is caused to unnecessary restriction.
" approximately " used herein, " approximately " or " roughly " are in order to modify any quantity that can change a little, but this variation a little can't change its essence.In embodiment, unless otherwise noted, the error range that represents the numerical value of being modified with " approximately ", " approximately " or " roughly " is generally to allow in 20 percent, is preferably in 10, is better in percentage five.
Fig. 1 is the process flow diagram according to a kind of image processing method 100 that promotes recognition of face rate of one embodiment of the invention.As shown in Figure 1, image processing method 100 methods comprise step 110～180(and should be appreciated that, mentioned step in the present embodiment except its order of special instruction, all can be adjusted order before and after it according to actual needs, even can be simultaneously or part carry out simultaneously).
The image that contains face step 110 collection; For instance, can utilize camera to carry out image acquisition to user's face, camera can be built-in camera or the circumscribed video camera on electronic product, the image that camera gathers can be presented in the preview screen of electronic product screen, therefore under normal condition, user can aim at face the middle position of preview screen.When the image that camera collection contains face, under normal circumstances, can shoot image clearly, if but take under the environment of backlight or low-light (level), automatic exposure (Auto Exposure) cannot be adjusted to correct exposure value, therefore, cannot carry out recognition of face according to image, please refer to shown in Fig. 5 A～Fig. 6 C, under environment backlight, image as shown in Figure 5A, because background luminance is high, the average exposure value of picture central range is on the low side, is approximately 16; In addition, under the environment of low-light (level), image as shown in Figure 6A, whole picture is excessively dark, and the average exposure value of picture central range is only about 43, therefore, cannot carry out recognition of face to image.
In Fig. 1, the default zone of step 120 in image carries out face detection.In the present embodiment, step 120 is mainly to judge in image, whether there is face, for instance, the method of face detection can adopt the cascade classifier algorithm of Viola – Jones whether the default zone in image (middle section) to be had to the detection of face, to find fast and effectively the facial image of many attitude and size.In the present embodiment, the face identification system that the mechanism of face detection adopts with step 180 is not identical, the method of face detection can only judge in image, whether there is face, and the face identification system adopting in step 180 can further carry out discriminating and the identification of biological characteristic for the face characteristic in image.
In one embodiment, default zone 210 is as shown in Figure 2 less than the whole image 220 of image 200, default zone 210 is positioned at the middle position scope of the whole image 220 of image 200, the custom that face can be aimed to picture central authorities to meet user, therefore human face region 230 will fall within default zone 210.Occupy the actual ratio of whole image 220 as for default zone 210, the visual different electronic product of those skilled in the art and elasticity adjustment.
Get back to Fig. 1, if cannot identify face by the default zone from image in step 120, that is cannot complete face detection, while representing the image that contains face gathering in step 110, may be under the environment of backlight or low-light (level), therefore, need be in the adjustment that looks like to expose of step 130 comparison chart, first, can first detect the average exposure value of the default zone in this image, judge the average exposure value size of current default zone, compare with expectation value according to the average exposure value size detecting, if average exposure value is lower than expectation value, representative is taken under the environment of backlight or low-light (level), therefore, can be to the image adjustment that exposes in step 130, that is exposure value in adjustment default zone.
Particularly, under the environment of backlight or low-light (level), can be according to the gradual adjustment threshold value of its look-up table (Lookup Table) to increase the average exposure value of middle position scope (default zone) in step 130, in the process increasing at average exposure value, can continue revised image to carry out recognition of face in step 120, reach expectation value until step 130 is adjusted exposure value for default zone, can find the human face region in image in step 120.
For instance, under environment backlight, the image shown in Fig. 5 A can be adjusted to the more suitable image of the exposure value shown in Fig. 5 B in step 130 originally, and the exposure value of its picture central range is about 84; In addition, under the environment of low-light (level), the image shown in Fig. 6 A can be adjusted to the more suitable image of the exposure value shown in Fig. 6 B in step 130 originally, and the exposure value of its picture central range is about 79.In practical operation, if the exposure value of image is on the low side, step 120 can can't detect face; But the in the situation that of low light source, if only adjust exposure in step 130 merely, although the exposure value of image is very high, but can cause the image frame speed (frame rate) of image to reduce, easily make image have residual elephant or ambiguous situation, therefore, the face identification system that step 180 adopts also goes out None-identified the biological characteristic of face, also cannot carry out identification.But expectation value also can not be too high, in order to avoid image frame rate reduction.For instance, expectation value is about the exposure value between 70～90, and as 85 left and right, but this numerical value is only for illustrating, and does not limit the present invention, and those skilled in the art should look the height of the expectation value of elasticity of demand adjustment at that time.
Therefore, in the time that the exposure value of default zone meets expectation value, can obtain an image that exposure value is more suitable, but, in order to make image be carried out living things feature recognition by face identification system is clearer, also must carry out Image Adjusting for the human face region in image, can be identified by face identification system.In the time adjusting image, under restriction that can not be too low in image frame rate (frame rate), can carry out step 140～the 160 dark portion region details with human face region in real time enhancing image.
When the exposure value of default zone meets expectation value and completes after face detection, the image information of human face region in step 140 continues analysis and adjusts default zone; Particularly, step 140 comprises that sub-step 141,142 analyzes brightness histogram and two image informations of mean flow rate (exposure value) with the human face region from default zone respectively, please refer to shown in Fig. 3.According to the schematic of the image information of the embodiment of the present invention, chart comprises the brightness histogram 310(Luminance Histogram in default zone) and mean flow rate 320(Average Brightness), wherein brightness histogram 310 represents the number of pixels that each brightness is shared, and 320 of mean flow rates are by average the brightness of each pixel in the default zone rear brightness value obtaining.The present invention can analyze brightness histogram and mean flow rate from default zone, and by both in conjunction with the space vector that forms one group of 256 dimension using as image information.
Get back to Fig. 1, in step 150, according to the image information (brightness histogram 310 and mean flow rate 320) of previously having found out, that does in step 150 and step 160 chooses image adjustment parameter and carries out in the step of image processing, the method of utilizing is a kind of wide dynamic range (WDR, Wide Dynamic Range) image processing techniques, wherein, the method of wide dynamic range has built multiple hyperspace vector model parameters in advance, each model parameter all has its corresponding wide dynamic range parameter, and every group model parameter has the space vector of 256 dimensions.
In step 150 and step 160, wide dynamic range can be found out the immediate corresponding wide dynamic range parameter of the image information forming with the space vector of above-mentioned these group 256 dimensions from the model parameter of multiple hyperspace vectors, and then carry out the processing to image execution wide dynamic range based on this wide dynamic range parameter, and the processing to image execution wide dynamic range in step 160, its main target is the readability that strengthens dark portion region in image, make the increasing degree of the brightness in dark portion region in image higher than the increasing degree of the brightness in image middle light region.Moreover, about how to find out in multiple hyperspace vectors and the immediate space vector of this image, in the present embodiment, can adopt expansion Jaccard similarity coefficient method (extended Jaccard similarity coefficient) to measure the overlapping degree between wantonly two groups of space vector set, so as to find out a hyperspace vector nearest with this image from the hyperspace vector model parameter building in advance.
So, in step 170, face identification system can obtain bright and image clearly, and then the biological characteristic that from then on gathers more sufficient face in image carries out identity discriminating, significantly improves and detects and the probalility of success of identification.For instance, under environment backlight, originally the image shown in Fig. 5 B be becoming clear and image clearly shown in Fig. 5 C after step 150 and 160 is processed, and the brightness of its picture central range is about 127; In addition, under the environment of low-light (level), originally the image shown in Fig. 6 B be becoming clear and image clearly shown in Fig. 6 C after step 160 is processed, and the brightness of its picture central range is about 120.
On the other hand, in other embodiments, the present invention also can adopt high dynamic range (HDR, High Dynamic Range) image technique to replace the image technique of wide dynamic range.Particularly, scene can be built in advance to record multiple models and corresponding multiple high dynamic range parameters thereof, the immediate corresponding high dynamic range parameter of the image information forming with the space vector of above-mentioned these group 256 dimensions can be from the model parameter of multiple hyperspace vectors, found out in step 150; Then, based on this high dynamic range parameter, carry out image to carry out the processing of high dynamic range in step 160, its main target is the readability that strengthens dark portion region in image.
In practical operation, the embodiment of above-mentioned step 120～180 can adopt software, hardware and/or firmware.For instance, image processing method 100 is applicable to electronic product arbitrarily, as: notebook computer, desk-top computer, intelligent mobile phone, panel computer etc., and these steps can adopt software, hardware and firmware work compound simultaneously.
Fig. 4 is a kind of image processing system 400 block diagrams that promote recognition of face rate according to one embodiment of the invention.As shown in Figure 4, image processing system 400 comprises camera 410, face detection unit 420, exposure adjustment unit 430, image analyzing unit 440, graphics processing unit 450 and image output unit 460.
Camera 410 is in order to gather the image that contains face, and face detection unit 420 carries out face detection in order to the default zone in image.In the time cannot completing face detection, exposure adjustment unit 430, in order to adjust the exposure value of default zone, makes the exposure value of default zone meet expectation value.In the time that the brightness of default zone meets expectation value and completes face detection, image analyzing unit 440 is in order to the image information in the human face region of analysis image.Graphics processing unit 450 is in order to according to image information, the corresponding image adjustment parameter of Selection Model parameter and model parameter, and the human face region in image is carried out to image processing.Image output unit 460 is in order to export image after treatment to face identification system 490.
In use, camera 410 can be built-in camera or the circumscribed video camera of electronic product, the image that camera 410 gathers can be presented in the preview screen of electronic product screen, and therefore, under normal condition, user can aim at face the central authorities of preview screen to carry out recognition of face.When camera 410 gathers the image that contains face, if but under the environment of backlight or low-light (level), automatic exposure is not yet adjusted to suitable degree, and face identification system 490 cannot gather enough biological attribute datas from image.Please refer to shown in Fig. 5 A～Fig. 6 C, under environment backlight, image as shown in Figure 5A, because background luminance is high, the mean flow rate of picture central range is on the low side, is approximately 16; In addition, under the environment of low-light (level), image as shown in Figure 6A, whole picture is excessively dark, and picture central range mean flow rate is only about 43.
In the present embodiment, face detection unit 420 is mainly to judge in image, whether there is face, for instance, face detection unit 420 can adopt the cascade classifier algorithm of Viola – Jones the default zone in image to be carried out to the detection of face kenel, to find fast and effectively the facial image of many attitude and size.In the present embodiment, face detection unit 420 is not identical with face identification system 490, and face identification system 490 can further carry out identity discriminating for the biological characteristic of the face in image.
When face detection unit 420 cannot identify face from image, represent when camera 410 gathers the image that contains face, may be under the environment of backlight or low-light (level), therefore, need the comparison chart adjustment that looks like to expose, first, can carry out photometry to default zone, and importing the information of this photometry into exposure adjustment unit 430, exposure adjustment unit 430 can detect for the average exposure value in default zone, judges the average exposure value size of current default zone.In exposure adjustment unit 430, have and there is expectation value, to compare with expectation value according to the average exposure value size detecting, if average exposure value is lower than expectation value, represents and take under the environment of backlight or low-light (level), therefore, exposure adjustment unit 430 can be to the image adjustment that exposes.
Particularly, under the environment of backlight or low-light (level), exposure adjustment unit 430 can be according to the gradual adjustment threshold value of its look-up table to increase the mean flow rate of middle position scope, in the process increasing in brightness, face detection unit 420 can continue revised image to carry out recognition of face, reach expectation value until exposure adjustment unit 430 is adjusted exposure value for default zone, face detection unit 420 can find the human face region in image.
For instance, under environment backlight, the image shown in Fig. 5 A can be exposed adjustment unit 430 and be adjusted to the more suitable image of the brightness shown in Fig. 5 B originally, and the brightness of its picture central range is about 84; In addition, under the environment of low-light (level), the image shown in Fig. 6 A can be exposed adjustment unit 430 and be adjusted to the more suitable image of the brightness shown in Fig. 6 B originally, and the brightness of its picture central range is about 79.
In practical operation, if the brightness of image is on the low side, face detection unit 420 can can't detect face; But the in the situation that of low light source, if only the simple exposure adjustment unit 430 that uses is adjusted exposure, although the brightness value of image is very high, but can cause the image frame rate reduction of image, easily make image have residual elephant or ambiguous situation, therefore, face identification system 490 also goes out None-identified people's biological characteristic, also cannot carry out identification.Therefore, in the time that the brightness of default zone meets expectation value, can obtain an image that brightness is more suitable, under restriction that can not be too low in image frame speed, the image analyzing unit 440 of can arranging in pairs or groups, graphics processing unit 450 and the details of image output unit 460 with dark portion region in real time enhancing image.
Should be appreciated that, above-mentioned default expectation value can not be too low, in order to avoid face detection unit 420 lost efficacy; But expectation value also can not be too high, in order to avoid image frame rate reduction.For instance, expectation value is about the brightness value between 70～90, and as 85 left and right, but this numerical value is only for illustrating, and does not limit the present invention, and those skilled in the art should look the height of the expectation value of elasticity of demand adjustment at that time.
Particularly, in the time that the brightness of default zone meets expectation value, image analyzing unit 440 is analyzed the image information in default zone; Luminance analyzing unit 440 can analyze brightness histogram and mean flow rate from default zone, and by both in conjunction with the space vector that forms one group of 256 dimension using as image information.
Image processing system 400 can build scene in advance to record multiple model parameters and corresponding multiple image adjustment parameters thereof.In one embodiment, the present invention adopts the image processing techniques of wide dynamic range, builds in advance the model parameter of multiple hyperspace vectors, and each model parameter all has its corresponding wide dynamic range parameter.Graphics processing unit 450 can be found out the immediate corresponding wide dynamic range parameter of the image information forming with the space vector of above-mentioned these group 256 dimensions in the model parameter of multiple hyperspace vectors, and then carry out the processing to image execution wide dynamic range based on this wide dynamic range parameter, its main target is the readability that strengthens dark portion region in image, makes the increasing degree of the brightness in dark portion region in image higher than the increasing degree of the brightness in highlights region.
So, face identification system 490 can obtain bright and image clearly, and then the biological characteristic that from then on gathers more sufficient face in image carries out identity discriminating, significantly improves and detects and the probalility of success of identification.For instance, under environment backlight, originally the image shown in Fig. 5 B is bright shown in Fig. 5 C and image clearly after being processed by graphics processing unit 450, and the brightness of its picture central range is about 127; In addition, under the environment of low-light (level), originally the image shown in Fig. 6 B is bright shown in Fig. 6 C and image clearly after can being processed by graphics processing unit 450, and the brightness of its picture central range is about 120.
Moreover, about how to find out with the space vector of this group 256 dimensions immediate one in the model of multiple hyperspace vectors, in one embodiment, graphics processing unit 450 can adopt expansion Jaccard similarity coefficient method to measure the overlapping degree between any two groups of space vector set, so as to finding out from the scene building in advance with the space vector of this group 256 dimensions at a distance of nearest model parameter.
On the other hand, in other embodiments, can adopt high dynamic range (HDR) image technique to replace the image technique of wide dynamic range.Particularly, image processing system 400 can build scene in advance to record multiple model parameters and corresponding multiple high dynamic range parameters thereof; The processing of graphics processing unit 450 based on carrying out this image to carry out with the immediate corresponding high dynamic range parameter of this image information high dynamic range in multiple model parameters, its main target is the readability that strengthens dark portion region in image.
In practical operation, above-mentioned face detection unit 420, exposure adjustment unit 430, image analyzing unit 440, graphics processing unit 450 and image output unit 460 etc., its embodiment can be software, hardware and/or firmware.For instance, image processing system 400 is applicable to electronic product arbitrarily, as: notebook computer, desk-top computer, intelligent mobile phone, panel computer etc., these unit can adopt software, hardware and firmware work compound simultaneously.
In sum, the present invention can be by the image gathering under the environment in backlight or low-light (level), is enhanced to bright and output image clearly, uses for rear end face identification system, so can significantly improve and detect and the probalility of success of identification.
Although the present invention with embodiment openly as above; so it is not in order to limit the present invention; under any, in technical field, have and conventionally know the knowledgeable; without departing from the spirit and scope of the present invention; when doing a little variation and retouching, therefore protection scope of the present invention is when being as the criterion depending on claims person of defining.
Claims (13)
1. an image processing method that promotes recognition of face rate, is applicable to face identification system, it is characterized in that, this image processing method comprises:
The image that collection contains face;
Default zone in above-mentioned image carries out face detection;
In the time cannot completing above-mentioned face detection, adjust the exposure value of above-mentioned default zone, make the exposure value of above-mentioned default zone meet expectation value;
In the time that the exposure value of above-mentioned default zone meets above-mentioned expectation value, carry out above-mentioned face detection, and analyze the image information in the human face region of above-mentioned image;
According to above-mentioned image information, the corresponding image adjustment parameter of Selection Model parameter and model parameter, and the human face region in above-mentioned image is carried out to image processing; And
Export image after treatment to above-mentioned face identification system.
2. the image processing method that promotes as claimed in claim 1 recognition of face rate, is characterized in that, described image information comprises brightness histogram and the mean flow rate in described human face region.
3. the image processing method that promotes as claimed in claim 1 recognition of face rate, is characterized in that, described image adjustment parameter is wide dynamic range parameter.
4. the image processing method that promotes as claimed in claim 3 recognition of face rate, is characterized in that, described in carry out image processing step comprise:
Based on described wide dynamic range parameter, described image is carried out to the processing of wide dynamic range.
5. the image processing method that promotes as claimed in claim 1 recognition of face rate, is characterized in that, described default zone is less than the whole picture of described image, and is positioned at the middle position scope of the whole picture of described image.
6. the image processing method that promotes as claimed in claim 1 recognition of face rate, is characterized in that, in the step of the image that described collection contains face, acquisition mode utilization shooting or video recording mode complete.
7. promote as claimed in claim 1 the image processing method of recognition of face rate, it is characterized in that, too high or too low when completing the average exposure value of the default zone of image described in described face detection interval scale, therefore, to the adjustment that exposes of described default zone, until described image completes face detection.
8. the image processing method that promotes as claimed in claim 1 recognition of face rate, is characterized in that, carries out in the step of image processing at the human face region in described image, and described disposal route strengthens the readability in dark portion region in image.
9. an image processing system that promotes recognition of face rate, is applicable to face identification system, it is characterized in that, this image processing system comprises:
Device, it is in order to gather the image that contains face;
Face detection unit, it carries out face detection in order to the default zone in above-mentioned image;
Exposure adjustment unit, it,, in order in the time cannot completing above-mentioned face detection, adjusts the exposure value of above-mentioned default zone, makes the exposure value of above-mentioned default zone meet expectation value;
Image analyzing unit, when the exposure value of above-mentioned default zone meets above-mentioned expectation value, this image analyzing unit carries out above-mentioned face detection, and analyzes the image information in the human face region of above-mentioned image;
Graphics processing unit, it is in order to according to above-mentioned image information, the corresponding image adjustment parameter of Selection Model parameter and model parameter, and the human face region in above-mentioned image is carried out to image processing; And
Image output unit, it is in order to export image after treatment to above-mentioned face identification system.
10. the image processing system that promotes as claimed in claim 9 recognition of face rate, is characterized in that, described image information comprises brightness histogram and the mean flow rate in described human face region.
11. promote the image processing system of recognition of face rate as claimed in claim 9, it is characterized in that, described image adjustment parameter is wide dynamic range parameter.
12. promote the image processing system of recognition of face rate as claimed in claim 9, it is characterized in that, described graphics processing unit, based on described wide dynamic range parameter, is carried out the processing of wide dynamic range to described image.
13. promote the image processing system of recognition of face rate as claimed in claim 9, it is characterized in that, described default zone is less than the whole picture of described image, and are positioned at the middle position scope of the whole picture of described image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310192437.5A CN104182721A (en) | 2013-05-22 | 2013-05-22 | Image processing system and image processing method capable of improving face identification rate |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310192437.5A CN104182721A (en) | 2013-05-22 | 2013-05-22 | Image processing system and image processing method capable of improving face identification rate |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104182721A true CN104182721A (en) | 2014-12-03 |
Family
ID=51963748
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310192437.5A Pending CN104182721A (en) | 2013-05-22 | 2013-05-22 | Image processing system and image processing method capable of improving face identification rate |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104182721A (en) |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104504668A (en) * | 2014-12-30 | 2015-04-08 | 宇龙计算机通信科技(深圳)有限公司 | Face-contained image sharpening method and device |
| CN104834908A (en) * | 2015-05-07 | 2015-08-12 | 惠州Tcl移动通信有限公司 | Image exposure method for mobile terminal based on eye pattern recognition and exposure system |
| CN105844227A (en) * | 2016-03-21 | 2016-08-10 | 湖南君士德赛科技发展有限公司 | Driver identity authentication method for school bus safety |
| CN105898147A (en) * | 2016-05-24 | 2016-08-24 | 广东欧珀移动通信有限公司 | Photometry processing method and device applied to mobile terminal |
| CN105893994A (en) * | 2016-06-23 | 2016-08-24 | 安徽时旭智能科技有限公司 | Face recognition system |
| CN106454146A (en) * | 2016-10-20 | 2017-02-22 | 北京旷视科技有限公司 | Image processing method and device and electronic system |
| CN106446860A (en) * | 2016-10-10 | 2017-02-22 | 上海成业智能科技股份有限公司 | Method for clearly acquiring face recognition image under light interference condition |
| CN106470315A (en) * | 2015-08-20 | 2017-03-01 | 卡西欧计算机株式会社 | Image processing apparatus and image processing method |
| CN107862265A (en) * | 2017-10-30 | 2018-03-30 | 广东欧珀移动通信有限公司 | Image processing method and related product |
| CN108322664A (en) * | 2017-01-16 | 2018-07-24 | 上海聚虹光电科技有限公司 | A kind of exposal control method based on infrared eye gray level image |
| CN110415407A (en) * | 2019-07-24 | 2019-11-05 | 厦门立林科技有限公司 | A kind of intelligent entrance guard control method and device and equipment based on high definition recognition of face |
| CN110414408A (en) * | 2019-07-24 | 2019-11-05 | 厦门立林科技有限公司 | A kind of face identification method, device and equipment |
| CN110807735A (en) * | 2018-08-06 | 2020-02-18 | Tcl集团股份有限公司 | Image processing method, image processing device, terminal equipment and computer readable storage medium |
| CN113938597A (en) * | 2020-06-29 | 2022-01-14 | 腾讯科技（深圳）有限公司 | Face recognition method and device, computer equipment and storage medium |
| CN115953333A (en) * | 2023-03-15 | 2023-04-11 | 杭州魔点科技有限公司 | Dynamic backlight compensation method and system |
| CN115953267A (en) * | 2023-03-09 | 2023-04-11 | 中川建投集团有限公司 | Intelligent construction site management system |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1797166A (en) * | 2004-12-30 | 2006-07-05 | 矽统科技股份有限公司 | Method for controlling exposure system |
| CN101247480A (en) * | 2008-03-26 | 2008-08-20 | 北京中星微电子有限公司 | Automatic exposure method based on objective area in image |
| CN101272504A (en) * | 2007-03-19 | 2008-09-24 | 帆宣系统科技股份有限公司 | Image processing apparatus |
| CN101399922A (en) * | 2007-09-28 | 2009-04-01 | 索尼株式会社 | Image pickup apparatus, image pickup method, and program therefore |
| CN101478641A (en) * | 2008-01-04 | 2009-07-08 | 三星Techwin株式会社 | Digital photographing apparatus and method of controlling the same |
| CN101600056A (en) * | 2008-06-03 | 2009-12-09 | 索尼株式会社 | Image pick up equipment and image pickup method |
| CN102075688A (en) * | 2010-12-28 | 2011-05-25 | 青岛海信网络科技股份有限公司 | Wide dynamic processing method for single-frame double-exposure image |
| JP4957030B2 (en) * | 2006-03-15 | 2012-06-20 | オムロン株式会社 | Image processing apparatus, image processing method, and program |
-
2013
- 2013-05-22 CN CN201310192437.5A patent/CN104182721A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1797166A (en) * | 2004-12-30 | 2006-07-05 | 矽统科技股份有限公司 | Method for controlling exposure system |
| JP4957030B2 (en) * | 2006-03-15 | 2012-06-20 | オムロン株式会社 | Image processing apparatus, image processing method, and program |
| CN101272504A (en) * | 2007-03-19 | 2008-09-24 | 帆宣系统科技股份有限公司 | Image processing apparatus |
| CN101399922A (en) * | 2007-09-28 | 2009-04-01 | 索尼株式会社 | Image pickup apparatus, image pickup method, and program therefore |
| CN101478641A (en) * | 2008-01-04 | 2009-07-08 | 三星Techwin株式会社 | Digital photographing apparatus and method of controlling the same |
| CN101247480A (en) * | 2008-03-26 | 2008-08-20 | 北京中星微电子有限公司 | Automatic exposure method based on objective area in image |
| CN101600056A (en) * | 2008-06-03 | 2009-12-09 | 索尼株式会社 | Image pick up equipment and image pickup method |
| CN102075688A (en) * | 2010-12-28 | 2011-05-25 | 青岛海信网络科技股份有限公司 | Wide dynamic processing method for single-frame double-exposure image |
Non-Patent Citations (2)
| Title |
|---|
| LI TAO,MING-JUNG SEOW,VIJAYAN K. ASARI: "Nonlinear Image Enhancement to Improve Face Detection in Complex Lighting Environment", 《INTERNATIONAL JOURNAL OF COMPUTATIONAL INTELLIGENCE RESEARCH》 * |
| 王小明、方晓颖、刘锦高: "复杂光照下的自适应人脸图像增强", 《计算机工程与应用》 * |
Cited By (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104504668A (en) * | 2014-12-30 | 2015-04-08 | 宇龙计算机通信科技(深圳)有限公司 | Face-contained image sharpening method and device |
| CN104834908B (en) * | 2015-05-07 | 2018-09-07 | 惠州Tcl移动通信有限公司 | The image exposure method and exposure system that a kind of mobile terminal is identified based on eyeprint |
| CN104834908A (en) * | 2015-05-07 | 2015-08-12 | 惠州Tcl移动通信有限公司 | Image exposure method for mobile terminal based on eye pattern recognition and exposure system |
| CN106470315A (en) * | 2015-08-20 | 2017-03-01 | 卡西欧计算机株式会社 | Image processing apparatus and image processing method |
| CN105844227A (en) * | 2016-03-21 | 2016-08-10 | 湖南君士德赛科技发展有限公司 | Driver identity authentication method for school bus safety |
| CN105898147A (en) * | 2016-05-24 | 2016-08-24 | 广东欧珀移动通信有限公司 | Photometry processing method and device applied to mobile terminal |
| CN105898147B (en) * | 2016-05-24 | 2018-09-11 | 广东欧珀移动通信有限公司 | Survey light processing method and apparatus applied to mobile terminal |
| CN105893994A (en) * | 2016-06-23 | 2016-08-24 | 安徽时旭智能科技有限公司 | Face recognition system |
| CN106446860A (en) * | 2016-10-10 | 2017-02-22 | 上海成业智能科技股份有限公司 | Method for clearly acquiring face recognition image under light interference condition |
| CN106454146A (en) * | 2016-10-20 | 2017-02-22 | 北京旷视科技有限公司 | Image processing method and device and electronic system |
| CN108322664A (en) * | 2017-01-16 | 2018-07-24 | 上海聚虹光电科技有限公司 | A kind of exposal control method based on infrared eye gray level image |
| CN107862265A (en) * | 2017-10-30 | 2018-03-30 | 广东欧珀移动通信有限公司 | Image processing method and related product |
| CN110807735A (en) * | 2018-08-06 | 2020-02-18 | Tcl集团股份有限公司 | Image processing method, image processing device, terminal equipment and computer readable storage medium |
| CN110415407A (en) * | 2019-07-24 | 2019-11-05 | 厦门立林科技有限公司 | A kind of intelligent entrance guard control method and device and equipment based on high definition recognition of face |
| CN110414408A (en) * | 2019-07-24 | 2019-11-05 | 厦门立林科技有限公司 | A kind of face identification method, device and equipment |
| CN110414408B (en) * | 2019-07-24 | 2022-04-12 | 厦门立林科技有限公司 | Face recognition method, device and equipment |
| CN113938597A (en) * | 2020-06-29 | 2022-01-14 | 腾讯科技（深圳）有限公司 | Face recognition method and device, computer equipment and storage medium |
| CN113938597B (en) * | 2020-06-29 | 2023-10-10 | 腾讯科技（深圳）有限公司 | Face recognition method, device, computer equipment and storage medium |
| CN115953267A (en) * | 2023-03-09 | 2023-04-11 | 中川建投集团有限公司 | Intelligent construction site management system |
| CN115953267B (en) * | 2023-03-09 | 2023-06-16 | 中川建投集团有限公司 | Intelligent building site management system |
| CN115953333A (en) * | 2023-03-15 | 2023-04-11 | 杭州魔点科技有限公司 | Dynamic backlight compensation method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104182721A (en) | Image processing system and image processing method capable of improving face identification rate | |
| EP3937481A1 (en) | Image display method and device | |
| US10963993B2 (en) | Image noise intensity estimation method, image noise intensity estimation device, and image recognition device | |
| US9311526B2 (en) | Image processing system and method of improving human face recognition | |
| WO2022042135A1 (en) | Face image selection method and apparatus, device, and storage medium | |
| WO2018113523A1 (en) | Image processing method and device, and storage medium | |
| CN110580428A (en) | image processing method, image processing device, computer-readable storage medium and electronic equipment | |
| CN110287791B (en) | Screening method and system for face pictures | |
| CN102625030B (en) | video enhancement method and system | |
| CN102542552B (en) | Frontlighting and backlighting judgment method of video images and detection method of shooting time | |
| CN102932582A (en) | Method and device for realizing motion detection | |
| CN103646392B (en) | Backlighting detecting and equipment | |
| US20210082127A1 (en) | Image analysis apparatus, image analysis method, and storage medium to display information representing flow quantity | |
| CN110059666B (en) | Attention detection method and device | |
| CN107172354A (en) | Method for processing video frequency, device, electronic equipment and storage medium | |
| CN110572636B (en) | Camera contamination detection method and device, storage medium and electronic equipment | |
| CN110796600A (en) | Image super-resolution reconstruction method, image super-resolution reconstruction device and electronic equipment | |
| CN108924427A (en) | A kind of video camera focus method, device and video camera | |
| CN111654643B (en) | Exposure parameter determination method and device, unmanned aerial vehicle and computer readable storage medium | |
| WO2023045626A1 (en) | Image acquisition method and apparatus, terminal, computer-readable storage medium and computer program product | |
| CN113596344A (en) | Shooting processing method and device, electronic equipment and readable storage medium | |
| US9940543B2 (en) | Control of computer vision pre-processing based on image matching using structural similarity | |
| US20190304152A1 (en) | Method and device for processing image | |
| US10733706B2 (en) | Mobile device, and image processing method for mobile device | |
| WO2024051697A1 (en) | Image fusion method and apparatus, electronic device, and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20141203 |
|
| WD01 | Invention patent application deemed withdrawn after publication |