CN1905687A - Apparatus and method capable of correcting human skin pixels in picture - Google Patents
Apparatus and method capable of correcting human skin pixels in picture Download PDFInfo
- Publication number
- CN1905687A CN1905687A CN 200610029652 CN200610029652A CN1905687A CN 1905687 A CN1905687 A CN 1905687A CN 200610029652 CN200610029652 CN 200610029652 CN 200610029652 A CN200610029652 A CN 200610029652A CN 1905687 A CN1905687 A CN 1905687A
- Authority
- CN
- China
- Prior art keywords
- pixel
- image
- skin pixels
- unit
- human body
- 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
Images
Landscapes
- Processing Of Color Television Signals (AREA)
Abstract
The invention relates to a device and method for calibrating the color of human skin, in which,the device consists of a judging unit, a controlling unit and a calibrating unit. The input-signals of image in the YUV domain are output from the judging unit and calibrating unit. The controlling unit is connected with the judging unit and calibrating unit, which controls the calibrating unit to calibrate the color of pixel of human skin according to the results determined by the judging unit.
Description
Technical field
The present invention relates to a kind of chrominance processing arrangement and method of picture signal, especially a kind of flesh correction apparatus and method of carrying out the colourity processing at the human skin pixels in the image.
Background technology
Because human interest to self is any other things head and shoulders above, is human body at the object of taking pictures, choosing when making a video recording very major part therefore, promptly in the image applications of reality, human body complexion is the important component part of image.But because take pictures, have interference such as many light, electricity, heat when making a video recording, make often effect nature of the human body complexion that presents in the image, thereby weakened the expressive force of image greatly.For the skin pixels in the image is carried out chromaticity correction, must judge that at first which pixel is the pixel that belongs to human body skin.
The method of available technology adopting is to utilize look-up table to determine skin area in the image, remake corresponding chromaticity correction, promptly the information in the brightness of each pixel in the picture signal, chrominance information and the table of comparisons is compared, if in the table of comparisons, find corresponding value, think that then this pixel is a skin pixels, otherwise think and be not skin pixels.This look-up table method must be opened up a storage area in advance and be stored this table of comparisons, and the table of comparisons quality of making directly has influence on the accuracy of judgement, if table of comparisons precision height then accuracy of judgement degree height, but the memory space that takies is also inevitable big; If table of comparisons precision is low, though can save storage resources, judge that precision is low, can influence final calibration result.
Summary of the invention
The objective of the invention is for the apparatus and method of human skin pixels in a kind of recoverable image are provided, it not only can more accurately judge the pixel of expression human body skin in the image, it is carried out chromaticity correction, and can save memory space, reduce hardware resource consumption.
In order to reach described purpose, solution of the present invention is: the device that human skin pixels in a kind of recoverable image is provided, it comprises a judging unit, a control unit and a correcting unit, the image input signal in YUV territory is successively via judging unit and correcting unit output, described control unit is connected with correcting unit with judging unit respectively, and according to the result of determination control correcting unit of judging unit the pixel that belongs to human body skin is carried out chromaticity correction.
In above-mentioned recoverable image in the device of human skin pixels, judging unit is according to two oblique lines and the colourity span that the rectangular area surrounds, and minimum brightness, colourity thresholding judge whether pixel belongs to human body skin, and correcting unit then utilizes gain coefficient and offset correction formula that the pixel that belongs to human body skin is carried out chromaticity correction.
In the device of human skin pixels, also can comprise a format conversion unit in above-mentioned recoverable image, the picture signal in RGB territory inputs to this judging unit via the picture signal that this format conversion unit converts the YUV territory to.
Another program of the present invention provides the method for human skin pixels in a kind of recoverable image, and it comprises the following steps: that (1) judge whether the pixel in the image input signal in YUV territory belongs to human body skin; (2) according to the result of determination of step (1), the pixel that belongs to human body skin in the image is carried out chromaticity correction.
In the method for human skin pixels, step (1) is according to two oblique lines and the colourity span that the rectangular area surrounds in above-mentioned recoverable image, and minimum brightness, colourity thresholding judge whether pixel belongs to human body skin.Step (2) utilizes gain coefficient and offset correction formula that the pixel that belongs to human body skin is carried out chromaticity correction.
In the method for human skin pixels, also can comprise a format conversion step in above-mentioned recoverable image, the picture signal that the picture signal in RGB territory is converted to the YUV territory is carried out skin pixels again and is judged and chromaticity correction.
The apparatus and method of human skin pixels utilize linear equation to determine the polygonal region of expression skin pixels at color space in the recoverable image provided by the invention, by simple algorithm the picture signal in YUV territory being carried out skin pixels judges, simplified hardware configuration, avoid unnecessary storage, saved hardware resource.Simultaneously, under the prerequisite that reduces erroneous judgement as far as possible, the pixel that is judged to be human body complexion is carried out chromaticity correction, make human body complexion after the correction more ruddy, healthy, naturally.
Description of drawings
Apparatus structure of the present invention and method flow are provided by following embodiment and accompanying drawing.
Fig. 1 is the schematic diagram of expression skin pixels chromaticity range in the U-V chrominance space;
Fig. 2 is the structured flowchart of the device of human skin pixels in the recoverable image of the present invention;
Fig. 3 is the flow chart of the method for human skin pixels in the recoverable image of the present invention.
Embodiment
Be described in further detail below with reference to the apparatus and method of accompanying drawing human skin pixels in the recoverable image of the present invention.
As shown in Figure 1, carry out sampling analysis, find that the chromatic value of human skin pixels generally meets the chromaticity range of the irregular area that dots among the figure through U, V chromatic component to the skin pixels in the great amount of images.On the basis of a large amount of statisticss, can think: when the colourity of a certain pixel falls into dashed region shown in Figure 1 in the U-V coordinate system, what judge then that this pixel represents is human body skin.The function of this dashed region can be represented with conic section, in a preferred embodiment of the present invention, in order to simplify hardware configuration, to save calculation resources, adopts linear function to represent this dashed region.As shown in Figure 1, the polygonal shape that two oblique line L1, L2 and rectangle frame ABCD surround is similar to this dashed region, because people's vision is lower to the sensitivity of colourity, therefore can't influence final flesh correction effect by this approximate error that is caused.
Referring to Fig. 2, the device of human skin pixels comprises judging unit 1, control unit 2 and correcting unit 3 in the recoverable image of the present invention, the image input signal in YUV territory is successively via judging unit 1 and correcting unit 3 outputs, described control unit 2 is connected with correcting unit 3 with judging unit 1 respectively, and carries out chromaticity correction according to the pixel that 3 pairs of correcting units of the result of determination of judging unit 1 control belong to human body skin.If input picture is a rgb format, then also need through a format conversion unit, the conversion of signals in RGB territory is inputed to judging unit 1 again to the YUV territory.
Referring to Fig. 1, Fig. 2 and Fig. 3, describe the workflow of the method and apparatus of human skin pixels in the recoverable image of the present invention in detail.After the YUV picture signal is imported device of the present invention, luminance component Y and chromatic component U, V by each pixel in 1 pair of image of judging unit judge respectively, to determine whether this pixel belongs to human body skin (step S10), if the brightness value and the chromatic value of a certain pixel satisfy following 3 conditions simultaneously, judge that then this pixel is a skin pixels.
1) luminance threshold
Cross dark pixel and be not considered to human body skin, even if perhaps, the effect of not proofreading and correct also meets the demands, and therefore, the brightness value of skin pixels should satisfy:
Y_in＞Y_low_threshold (1)
The brightness value of pixel in the Y_in presentation video input signal wherein, Y_low_threshold represents the minimum brightness threshold value.
2) colourity thresholding
Similar to luminance threshold, for the shallow excessively pixel of color, do not process equally, so the chromatic value of skin pixels should satisfy:
|U_in|+|V_in|＞C_low_threshold (2)
Wherein the part of the chrominance information of pixel in U_in and the V_in difference presentation video input signal is promptly represented complete chromatic value altogether, and C_low_threshold represents minimum colourity threshold value.
3) chroma areas restriction
As shown in Figure 1, the shadow region in the U-V coordinate system is that it is surrounded by two oblique line L1, L2 and a rectangular area ABCD by the polygonal region of the colourity span of the approximate expression human skin pixels that obtains.If the function of oblique line L1, L2 is expressed as V=k_up*U+b_up and V=k_low*U+b_low respectively, then the chromatic value of skin pixels should satisfy:
k_low*U_in+b_low＜V_in＜k_up*U_in+b_up (3)
Wherein, k_up and k_low represent the slope of oblique line L1, L2 respectively, and b_up and b_low represent L1, the L2 intercept on the V axle respectively.If threshold value determined that then the chromatic value of skin pixels also should satisfy about the scope of rectangular area ABCD adopted two groups:
U_low_threshold＜U_in＜U_up_threshold (4) and V_low_threshold＜V_in＜V_up_threshold (5)
If a certain pixel can not satisfy above-mentioned 3 conditions simultaneously, then judging unit 1 judges that this pixel is not a skin pixels, and to control unit 2 transmission information, does not deal with and directly output with 3 pairs of these pixels of control correcting unit; If a certain pixel has satisfied above-mentioned 3 conditions simultaneously, then send one to control unit 2 and judge that this pixel is the information of skin pixels by judging unit 1, after control unit 2 is received this confirmation signal, 3 pairs of these skin pixels of control correcting unit carry out chromaticity correction (step S20), and concrete chromaticity correction formula is:
U_out＝U_in*K_U_gain+128*(K_U_gain-1)
V_out＝V_in*K_V_gain+128*(K_V_gain-1)
Wherein U_out, V_out distinguish the chromatic value of presentation video output signal, and K_U_gain and K_V_gain represent the gain coefficient of U component and V component respectively, and doing addition is offset correction.
In a preferred embodiment of the present invention, Y_low_threshold is made as 40, and C_low_threshold is made as 16, and then inequality (1), (2) are expressed as: Y_in＞40 Hes | U_in|+|V_in|＞16; K_up gets-0.7, and k_low gets-2, and b_up gets 10, and b_low gets-42, and then inequality (3) is expressed as: (2) * U_in-42＜V_in＜(0.7) * U_in+10; Threshold value is set to 0 ,-50 and 40,0 respectively about two groups of rectangular area ABCD, and then inequality (3), (4) are expressed as :-50＜U_in＜0 and 0＜V_in＜40.Each pixel in the YUV image input signal is through after judging, if satisfy above-mentioned 5 inequality, then the chrominance information of this pixel is handled by gain coefficient and offset correction, wherein, gain coefficient K_U_gain can be provided with according to different requirements with K_V_gain, to reach required calibration result, in present embodiment, K_U_gain and K_V_gain value respectively are 1.2 and 1.3, so final updating formula is expressed as: U_out=U_in*1.2+128* (1.2-1); V_out=V_in*1.3+128* (1.3-1), the colour of skin after the correction seems more ruddy, healthy before proofreading and correct.
Adopt the apparatus and method of human skin pixels in the recoverable image of the present invention, can determine the skin area of polygonal shape by linear function in the U-V space, and the pixel in the picture signal is judged in conjunction with minimum brightness, colourity thresholding, simplified hardware configuration, avoid unnecessary storage, saved hardware resource.Simultaneously, under the prerequisite that reduces erroneous judgement as far as possible, the pixel that is judged to be human body complexion is carried out chromaticity correction, make the human body complexion after the correction more natural, improved the expressive force of image.
Claims (8)
1, the device of human skin pixels in a kind of recoverable image, it is characterized in that, described device comprises: a judging unit (1), a control unit (2) and a correcting unit (3), the image input signal in YUV territory is successively via judging unit (1) and correcting unit (3) output, described control unit (2) is connected with correcting unit (3) with judging unit (1) respectively, and according to the result of determination control correcting unit (3) of judging unit (1) pixel that belongs to human body skin is carried out chromaticity correction.
2, the device of human skin pixels in the recoverable image as claimed in claim 1, it is characterized in that: described judging unit (1) is according to two oblique lines and the colourity span that the rectangular area surrounds, and minimum brightness, colourity thresholding judge whether pixel belongs to human body skin.
3, the device of human skin pixels in the recoverable image as claimed in claim 1 is characterized in that: described correcting unit (3) utilizes gain coefficient and offset correction formula that the pixel that belongs to human body skin is carried out chromaticity correction.
4, the device of human skin pixels in the recoverable image as claimed in claim 1, it is characterized in that: described device also comprises a format conversion unit, and the picture signal in RGB territory inputs to this judging unit (1) via the picture signal that this format conversion unit converts the YUV territory to.
5, the method for human skin pixels in a kind of recoverable image is characterized in that described method comprises the following steps:
(1) judges whether the pixel in the image input signal in YUV territory belongs to human body skin;
(2) according to the result of determination of step (1), the pixel that belongs to human body skin in the image is carried out chromaticity correction.
6, the method for human skin pixels in the recoverable image as claimed in claim 5, it is characterized in that, step (1) is according to two oblique lines and the colourity span that the rectangular area surrounds, and minimum brightness, colourity thresholding judge whether pixel belongs to human body skin.
7, the method for human skin pixels in the recoverable image as claimed in claim 5 is characterized in that: step (2) utilizes gain coefficient and offset correction formula that the pixel that belongs to human body skin is carried out chromaticity correction.
8, the method for human skin pixels in the recoverable image as claimed in claim 5, it is characterized in that: described method also comprises a format conversion step, and the picture signal that the picture signal in RGB territory is converted to the YUV territory is carried out skin pixels again and judged and chromaticity correction.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200610029652 CN1905687A (en) | 2006-08-02 | 2006-08-02 | Apparatus and method capable of correcting human skin pixels in picture |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200610029652 CN1905687A (en) | 2006-08-02 | 2006-08-02 | Apparatus and method capable of correcting human skin pixels in picture |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN1905687A true CN1905687A (en) | 2007-01-31 |
Family
ID=37674786
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 200610029652 Pending CN1905687A (en) | 2006-08-02 | 2006-08-02 | Apparatus and method capable of correcting human skin pixels in picture |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1905687A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101320477B (en) * | 2008-07-10 | 2011-04-27 | 北京中星微电子有限公司 | Human body tracing method and equipment thereof |
| CN101079952B (en) * | 2006-05-26 | 2011-10-05 | 佳能株式会社 | Image processing method and image processing apparatus |
| CN103065290A (en) * | 2013-01-23 | 2013-04-24 | 广东欧珀移动通信有限公司 | Device and method for correcting skin color in photo |
| CN105096238A (en) * | 2015-03-27 | 2015-11-25 | 浙江慧谷信息技术有限公司 | Visual lossless CMYK dynamic conversion algorithm |
| CN108604293A (en) * | 2016-01-30 | 2018-09-28 | 三星电子株式会社 | The device and method for improving picture quality |
-
2006
- 2006-08-02 CN CN 200610029652 patent/CN1905687A/en active Pending
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101079952B (en) * | 2006-05-26 | 2011-10-05 | 佳能株式会社 | Image processing method and image processing apparatus |
| CN101320477B (en) * | 2008-07-10 | 2011-04-27 | 北京中星微电子有限公司 | Human body tracing method and equipment thereof |
| CN103065290A (en) * | 2013-01-23 | 2013-04-24 | 广东欧珀移动通信有限公司 | Device and method for correcting skin color in photo |
| CN105096238A (en) * | 2015-03-27 | 2015-11-25 | 浙江慧谷信息技术有限公司 | Visual lossless CMYK dynamic conversion algorithm |
| CN108604293A (en) * | 2016-01-30 | 2018-09-28 | 三星电子株式会社 | The device and method for improving picture quality |
| CN108604293B (en) * | 2016-01-30 | 2022-05-27 | 三星电子株式会社 | Apparatus and method for improving image quality |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20230217116A1 (en) | Merging multiple exposures to generate a high dynamic range image | |
| US11074882B2 (en) | Method and device for adjusting grayscale of display panel | |
| TWI399100B (en) | Image processing method | |
| CN1890692A (en) | Image adjustment with tone rendering curve | |
| US9071761B2 (en) | Information processing apparatus, information processing method, program, and imaging apparatus including optical microscope | |
| KR20180032750A (en) | Method of processing image and display apparatus performing the same | |
| US20110134328A1 (en) | Video processing apparatus and video processing method | |
| CN101340510A (en) | Method for video enhancement and apparatus thereof | |
| KR102522423B1 (en) | Display apparatus and method for processing image of the display apparatus | |
| US20070211176A1 (en) | Apparatus and method for adjusting saturation image | |
| CN1905687A (en) | Apparatus and method capable of correcting human skin pixels in picture | |
| US20140348428A1 (en) | Dynamic range-adjustment apparatuses and methods | |
| US8269898B2 (en) | Image processing apparatus | |
| KR20160130005A (en) | Optical compensation system and Optical compensation method thereof | |
| CN111885312A (en) | HDR image imaging method, system, electronic device and storage medium | |
| CN111899690B (en) | Method and device for improving low-brightness Mura | |
| US20180084180A1 (en) | Display apparatus and method of processing image thereof | |
| WO2010089836A1 (en) | Image processing device | |
| JP2012235465A (en) | Apparatus and method for processing image in digital camera | |
| KR102423052B1 (en) | Display apparatus for dog and method of driving the same | |
| EP2802139B1 (en) | Image color adjusting method and electronic device using the same | |
| JP2005318499A (en) | Image processing apparatus | |
| CN101047866A (en) | Color index searching method for color picture | |
| CN112419217B (en) | Image processing method, device, computer equipment and medium | |
| CN114822444B (en) | Compatible display method, device, equipment and readable 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 | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |