CN102184401B - Facial feature extraction method - Google Patents
Facial feature extraction method Download PDFInfo
- Publication number
- CN102184401B CN102184401B CN201110110982.6A CN201110110982A CN102184401B CN 102184401 B CN102184401 B CN 102184401B CN 201110110982 A CN201110110982 A CN 201110110982A CN 102184401 B CN102184401 B CN 102184401B
- Authority
- CN
- China
- Prior art keywords
- integral projection
- image
- horizontal
- face
- vertical
- 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.)
- Expired - Fee Related
Links
- 238000000605 extraction Methods 0.000 title claims abstract description 17
- 230000001815 facial effect Effects 0.000 title claims abstract description 9
- 210000000056 organ Anatomy 0.000 claims abstract description 21
- 238000001914 filtration Methods 0.000 claims abstract description 11
- 238000000034 method Methods 0.000 claims description 14
- 238000004458 analytical method Methods 0.000 abstract description 3
- 239000003086 colorant Substances 0.000 abstract 1
- 230000001186 cumulative effect Effects 0.000 description 11
- 230000007812 deficiency Effects 0.000 description 3
- 239000000284 extract Substances 0.000 description 3
- 230000008921 facial expression Effects 0.000 description 3
- 238000005286 illumination Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 210000004709 eyebrow Anatomy 0.000 description 1
- 210000000887 face Anatomy 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000013011 mating Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Abstract
The invention discloses a facial feature extraction method, which comprises the following steps of: (1) dividing H, S, V (Hue, Saturation, Value) components of a facial image in an HSV color space and taking out the V component; (2) filtering the V component image; (3) corroding and expanding the obtained image; (4) carrying out binaryzation on the obtained image, and obtaining an image with black and white colors; (5) respectively carrying out integral projection on the obtained black and white image in a horizontal direction and a vertical direction to obtain a horizontal and vertical integral projection image; (6) determining feature values of primary organs of a face according to the integral projection image obtained through analysis; and (7) locating correct locations of the primary organs on the face according to the obtained feature values. Due to the adoption of the extraction method, accuracy rate of facial feature recognition can be improved.
Description
Technical field
The invention belongs to image processing field, particularly a kind of face feature extraction method.
Background technology
Face feature extraction method commonly used mainly contains following 3 classes now:
1. Extraction of Geometrical Features: architectural feature and the priori of utilizing people's face, by the notable feature to human face expression, shape and change in location such as eyes, nose, eyebrow, mouth etc. position and measure, determine its size, distance, shape and mutually ratio characteristic relation, identify.
2. statistical nature extracts: compare with geometric properties, this statistical nature is based on the overall intensity feature of image, it emphasizes to keep as much as possible the information of original facial expression image, by the training to great amount of samples, obtain its statistical nature, its basic thought is that facial expression image is mapped to feature space, and with the laggard row mode classification of great amount of images Data Dimensionality Reduction, in fact the method for therefore extracting statistical nature is exactly " subspace analysis method ".
3. frequency field feature extraction: because the Gabor Wavelet Kernel Function has the characteristic identical with the two-dimentional echo area of human brain cortex simple cell, namely can catch the partial structurtes information corresponding to spatial frequency, locus and directional selectivity, therefore mating as feature with the coefficient amplitude of Two-Dimensional Gabor Wavelets conversion has good visual characteristic and biology background, is widely used in recognition of face and image processing.
With regard to the Extraction of Geometrical Features method, very easily be subject to the impact that the illumination deficiency is brought, and the facial interfere informations such as noise too much also are one of impact very large factors of extracting accuracy rate, have much room for improvement.
In view of above analysis, the inventor furthers investigate for the lower problem of existing face feature extraction method accuracy rate, and this case produces thus.
Summary of the invention
Technical matters to be solved by this invention is for the defective in the aforementioned background art and deficiency, and a kind of face feature extraction method is provided, and it can improve the accuracy rate of face characteristic identification.
The present invention is for solving above technical matters, and the technical scheme that adopts is:
A kind of face feature extraction method comprises the steps:
(1) facial image is divided H, S, V component at the HSV color space, and take out the V component;
(2) aforementioned V component image being carried out filtering processes;
(3) image that abovementioned steps is obtained corrodes successively and expands;
(4) image that obtains is carried out binaryzation, obtain only having black and white image;
(5) black white image that obtains is carried out respectively integral projection in the horizontal direction with on the vertical direction, to obtain level and vertical integral projection figure;
(6) analyze the integral projection figure that obtains, determine thus the eigenwert of people's face major organs for foundation;
(7) orient major organs in people tram on the face according to the eigenwert of obtaining.
In the above-mentioned steps (2), the V component image is carried out Gauss filtering or medium filtering.
In the above-mentioned steps (6), determine that the detailed process of eigenwert is:
Find out 3 peak values according to the aforementioned horizontal integral projection figure that obtains, wherein, the peak that is positioned at the top is the horizontal central point of eyes, finds out respectively peak point according to vertical integral projection figure in front 1/2 and rear 1/2 scope of horizontal ordinate again, is the vertical centre point of left eye and right eye;
In horizontal integral projection figure, the peak in the middle of being positioned at is the horizontal central point of nose, again according to vertical integral projection figure, finds out its valley in the environs of the center of horizontal ordinate, is the vertical centre point of nose;
In horizontal integral projection figure, be positioned at the horizontal central point that the peak of below is face, again according to vertical integral projection figure, in the environs of the center of horizontal ordinate, find out its peak value, be the vertical centre point of face.
In the above-mentioned steps (6), in horizontal integral projection figure, determine the central point of major organs after, set again a critical value, cumulative all surpass numbers of the ordinate of this critical value, be the thickness of this major organs.
In the above-mentioned steps (6), in vertical integral projection figure, determine the central point of major organs after, set again a critical value, cumulative all surpass numbers of the horizontal ordinate of this critical value, be the width of this major organs.
After adopting such scheme, the present invention improves existing integral projection method, in conjunction with the skin color model to people's face, improvement is integrated the image of projection, thereby overcome the impact that illumination is not enough and interfere information brings, improve the accuracy rate of people's face major organs location, according to test relatively, the just inspection rate of major organs eyes, nose, face can both improve 20%-30%.
Description of drawings
Fig. 1 is workflow diagram of the present invention;
Fig. 2 is the horizontal integral projection figure that adopts method of the present invention to obtain;
Fig. 3 is the horizontal integral projection figure that adopts existing method to obtain.
Embodiment
Below with reference to accompanying drawing, implementation process of the present invention is elaborated.
As shown in Figure 1, the invention provides a kind of face feature extraction method, comprise the steps:
(1) reads in a facial image: from people's face picture library, read in a standard faces;
(2) extract the V component in the HSV space: the aforementioned facial image that reads in is marked off H component, S component and V component at the HSV color space, then take out the V component, the V component that extracts by this kind step can overcome the impact (can cooperate simultaneously Fig. 2 and Fig. 3 display comparison effect) that the illumination deficiency is brought well;
(3) carry out neighborhood processing: the aforementioned V component image that obtains is carried out the filtering such as Gauss filtering or medium filtering and process, can remove well like this impact that the interfere informations such as non-feature organ bring, certainly also can adopt other filtering mode, be not limited to cited herein;
(4) carrying out morphology processes: the image that abovementioned steps obtains corrodes successively and expands, and can effectively strengthen like this display effect of the major organs such as eyes nose;
(5) image binaryzation: the image that obtains is carried out binaryzation according to the adaptive threshold method, obtain only having black and white image;
(6) integral projection: the black white image that obtains is carried out respectively integral projection in the horizontal direction with on the vertical direction, to obtain level and vertical integral projection figure;
(7) take out eigenwert: analyze the integral projection figure that obtains, determine thus the eigenwert of the major organs such as eyes, nose, face for foundation; Detailed process is:
Find out peak value according to the aforementioned horizontal integral projection figure that obtains, co-existing in the drawings 3 peak values, representing respectively eyes, nose and face, wherein, be positioned at the peak of the top, be the horizontal central point of eyes, can also set a critical value herein, cumulative all numbers above the ordinate of this critical value, thickness cumulative and that be eyes; In front 1/2 and rear 1/2 scope of horizontal ordinate, find out respectively peak point according to vertical integral projection figure again, be the vertical centre point of left eye and right eye, can determine like this two oculocentric positions, also set simultaneously a critical value, both sides are cumulative all numbers above the horizontal ordinate of this critical value respectively, are two width.
In horizontal integral projection figure, be positioned at middle peak, be the horizontal central point of nose, set again a critical value, cumulative all numbers above the ordinate of this critical value, thickness cumulative and that be nose; According to vertical integral projection figure, in the environs of the center of horizontal ordinate, find out its valley again, be the vertical centre point of nose, set simultaneously a critical value, cumulative all surpass numbers of the horizontal ordinate of this critical value, be the width of nose.
Among the horizontal integral projection figure, be positioned at the horizontal central point that the peak of below is face, set simultaneously a critical value, cumulative all surpass numbers of the ordinate of this critical value, thickness cumulative and that be face; According to vertical integral projection figure, in the environs of the center of horizontal ordinate, find out its peak value again, be the vertical centre point of face, set simultaneously a critical value, cumulative all surpass numbers of the horizontal ordinate of this critical value, be the width of face.
(8) major organs position, location: orient the major organs such as eyes, nose, face according to the eigenwert of obtaining in people tram on the face.
Above embodiment only for explanation technological thought of the present invention, can not limit protection scope of the present invention with this, every technological thought that proposes according to the present invention, and any change of doing on the technical scheme basis all falls in the protection domain of the present invention.
Claims (4)
1. a face feature extraction method is characterized in that comprising the steps:
(1) facial image is divided H, S, V component at the HSV color space, and take out the V component;
(2) aforementioned V component image being carried out filtering processes;
(3) image that abovementioned steps is obtained corrodes successively and expands;
(4) image that obtains is carried out binaryzation, obtain only having black and white image;
(5) black white image that obtains is carried out respectively integral projection in the horizontal direction with on the vertical direction, to obtain level and vertical integral projection figure;
(6) analyze the integral projection figure that obtains, determine thus the eigenwert of people's face major organs for foundation;
The detailed process of the eigenwert of described definite people's face major organs is:
Find out 3 peak values according to the aforementioned horizontal integral projection figure that obtains, wherein, the peak that is positioned at the top is the horizontal central point of eyes, finds out respectively peak point according to vertical integral projection figure in front 1/2 and rear 1/2 scope of horizontal ordinate again, is the vertical centre point of left eye and right eye;
In horizontal integral projection figure, the peak in the middle of being positioned at is the horizontal central point of nose, again according to vertical integral projection figure, finds out its valley in the environs of the center of horizontal ordinate, is the vertical centre point of nose;
In horizontal integral projection figure, be positioned at the horizontal central point that the peak of below is face, again according to vertical integral projection figure, in the environs of the center of horizontal ordinate, find out its peak value, be the vertical centre point of face;
(7) orient major organs in people tram on the face according to the eigenwert of obtaining.
2. a kind of face feature extraction method as claimed in claim 1 is characterized in that: in the described step (2), the V component image is carried out Gauss filtering or medium filtering.
3. a kind of face feature extraction method as claimed in claim 1, it is characterized in that: in the described step (6), after in horizontal integral projection figure, determining the central point of major organs, set again a critical value, adding up, all surpass numbers of the ordinate of this critical value, are the thickness of this major organs.
4. a kind of face feature extraction method as claimed in claim 1, it is characterized in that: in the described step (6), after in vertical integral projection figure, determining the central point of major organs, set again a critical value, adding up, all surpass numbers of the horizontal ordinate of this critical value, are the width of this major organs.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201110110982.6A CN102184401B (en) | 2011-04-29 | 2011-04-29 | Facial feature extraction method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201110110982.6A CN102184401B (en) | 2011-04-29 | 2011-04-29 | Facial feature extraction method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN102184401A CN102184401A (en) | 2011-09-14 |
CN102184401B true CN102184401B (en) | 2013-01-16 |
Family
ID=44570575
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201110110982.6A Expired - Fee Related CN102184401B (en) | 2011-04-29 | 2011-04-29 | Facial feature extraction method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102184401B (en) |
Families Citing this family (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102842033A (en) * | 2012-08-17 | 2012-12-26 | 苏州两江科技有限公司 | Human expression emotion semantic recognizing method based on face recognition |
CN103112015B (en) * | 2013-01-29 | 2015-03-04 | 山东电力集团公司电力科学研究院 | Operating object position and posture recognition method applicable to industrial robot |
CN106778676B (en) * | 2016-12-31 | 2020-02-18 | 中南大学 | Attention assessment method based on face recognition and image processing |
CN108399598B (en) * | 2018-01-24 | 2021-11-23 | 武汉智博创享科技股份有限公司 | Panoramic image-based face blurring method and system |
CN108596064A (en) * | 2018-04-13 | 2018-09-28 | 长安大学 | Driver based on Multi-information acquisition bows operating handset behavioral value method |
CN109002801B (en) * | 2018-07-20 | 2021-01-15 | 燕山大学 | Face shielding detection method and system based on video monitoring |
CN109543518A (en) * | 2018-10-16 | 2019-03-29 | 天津大学 | A kind of human face precise recognition method based on integral projection |
CN110706415A (en) * | 2019-10-14 | 2020-01-17 | 重庆国翰能源发展有限公司 | Charging pile control system based on biological recognition and face recognition method thereof |
CN112381065B (en) * | 2020-12-07 | 2024-04-05 | 福建天创信息科技有限公司 | Face positioning method and terminal |
CN114529729B (en) * | 2022-04-22 | 2022-08-23 | 珠海视熙科技有限公司 | Strobe detection and elimination method, device, camera and storage medium |
CN114549843B (en) * | 2022-04-22 | 2022-08-23 | 珠海视熙科技有限公司 | Stroboscopic stripe detection and elimination method and device, image pickup device and storage medium |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20050041867A1 (en) * | 2002-03-27 | 2005-02-24 | Gareth Loy | Method and apparatus for the automatic detection of facial features |
US20050063582A1 (en) * | 2003-08-29 | 2005-03-24 | Samsung Electronics Co., Ltd. | Method and apparatus for image-based photorealistic 3D face modeling |
CN101383001A (en) * | 2008-10-17 | 2009-03-11 | 中山大学 | Quick and precise front human face discriminating method |
CN101539992A (en) * | 2008-03-20 | 2009-09-23 | 中国科学院自动化研究所 | Multi-illumination face recognition method based on morphologic quotient images |
-
2011
- 2011-04-29 CN CN201110110982.6A patent/CN102184401B/en not_active Expired - Fee Related
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20050041867A1 (en) * | 2002-03-27 | 2005-02-24 | Gareth Loy | Method and apparatus for the automatic detection of facial features |
US20050063582A1 (en) * | 2003-08-29 | 2005-03-24 | Samsung Electronics Co., Ltd. | Method and apparatus for image-based photorealistic 3D face modeling |
CN101539992A (en) * | 2008-03-20 | 2009-09-23 | 中国科学院自动化研究所 | Multi-illumination face recognition method based on morphologic quotient images |
CN101383001A (en) * | 2008-10-17 | 2009-03-11 | 中山大学 | Quick and precise front human face discriminating method |
Also Published As
Publication number | Publication date |
---|---|
CN102184401A (en) | 2011-09-14 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN102184401B (en) | Facial feature extraction method | |
US20210019601A1 (en) | System and Method for Face Detection and Landmark Localization | |
Galoogahi et al. | Inter-modality face sketch recognition | |
CN101615292B (en) | Accurate positioning method for human eye on the basis of gray gradation information | |
Skodras et al. | An unconstrained method for lip detection in color images | |
Corkery et al. | A preliminary investigation on face recognition as a biometric identifier of sheep | |
US11250249B2 (en) | Human body gender automatic recognition method and apparatus | |
WO2019014813A1 (en) | Method and apparatus for quantitatively detecting skin type parameter of human face, and intelligent terminal | |
CN112232332B (en) | Non-contact palm detection method based on video sequence | |
WenJuan et al. | A real-time lip localization and tacking for lip reading | |
CN110728185B (en) | Detection method for judging existence of handheld mobile phone conversation behavior of driver | |
CN101996317B (en) | Method and device for identifying markers in human body | |
Yang et al. | An efficient method for vehicle model identification via logo recognition | |
Yap et al. | A survey on facial wrinkles detection and inpainting: Datasets, methods, and challenges | |
TWI427545B (en) | Face recognition method based on sift features and head pose estimation | |
CN106650616A (en) | Iris location method and visible light iris identification system | |
CN104063682A (en) | Pedestrian detection method based on edge grading and CENTRIST characteristic | |
CN103207989B (en) | A kind of vein recognition methods of the slight insect pest blade based on principal component analysis | |
Bhaskar et al. | Advanced algorithm for gender prediction with image quality assessment | |
CN110222660B (en) | Signature authentication method and system based on dynamic and static feature fusion | |
CN105138984B (en) | Sharpening image recognition methods based on multiresolution overshoot effect measuring | |
Saha et al. | An approach to detect the region of interest of expressive face images | |
CN107424141B (en) | Face image quality evaluation method based on probability block | |
Fan et al. | Lane detection based on machine learning algorithm | |
CN109145875B (en) | Method and device for removing black frame glasses in face image |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20130116 Termination date: 20150429 |
|
EXPY | Termination of patent right or utility model |