CN105426843B - Single-lens lower palm vein and palm print image acquisition device and image enhancement and segmentation method - Google Patents
Single-lens lower palm vein and palm print image acquisition device and image enhancement and segmentation method Download PDFInfo
- Publication number
- CN105426843B CN105426843B CN201510801840.2A CN201510801840A CN105426843B CN 105426843 B CN105426843 B CN 105426843B CN 201510801840 A CN201510801840 A CN 201510801840A CN 105426843 B CN105426843 B CN 105426843B
- Authority
- CN
- China
- Prior art keywords
- palm
- image
- light
- camera lens
- ccd sensor
- 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.)
- Active
Links
- 210000003462 vein Anatomy 0.000 title claims abstract description 62
- 238000000034 method Methods 0.000 title claims abstract description 27
- 230000011218 segmentation Effects 0.000 title claims description 19
- 239000000463 material Substances 0.000 claims abstract description 7
- 238000001914 filtration Methods 0.000 claims description 15
- 238000012545 processing Methods 0.000 claims description 13
- 230000000877 morphologic effect Effects 0.000 claims description 10
- 230000008569 process Effects 0.000 claims description 7
- 238000005286 illumination Methods 0.000 claims description 4
- 238000005192 partition Methods 0.000 claims description 4
- 230000000149 penetrating effect Effects 0.000 claims description 4
- 210000001367 artery Anatomy 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 3
- 230000001186 cumulative effect Effects 0.000 claims description 3
- 230000010339 dilation Effects 0.000 claims description 3
- 230000003628 erosive effect Effects 0.000 claims description 3
- 238000012804 iterative process Methods 0.000 claims description 3
- 230000003595 spectral effect Effects 0.000 claims description 3
- 206010027146 Melanoderma Diseases 0.000 claims description 2
- 125000004122 cyclic group Chemical group 0.000 claims description 2
- 238000012217 deletion Methods 0.000 claims description 2
- 230000037430 deletion Effects 0.000 claims description 2
- 238000002834 transmittance Methods 0.000 claims description 2
- 230000002708 enhancing effect Effects 0.000 abstract description 9
- 230000000694 effects Effects 0.000 description 20
- 238000005516 engineering process Methods 0.000 description 8
- 230000006870 function Effects 0.000 description 6
- 210000003491 skin Anatomy 0.000 description 4
- 230000008859 change Effects 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 239000011800 void material Substances 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 210000004204 blood vessel Anatomy 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 210000002615 epidermis Anatomy 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 238000002310 reflectometry Methods 0.000 description 1
- 230000003014 reinforcing effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000009738 saturating Methods 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 238000007514 turning Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/12—Fingerprints or palmprints
- G06V40/13—Sensors therefor
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/90—Dynamic range modification of images or parts thereof
- G06T5/94—Dynamic range modification of images or parts thereof based on local image properties, e.g. for local contrast enhancement
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/12—Fingerprints or palmprints
- G06V40/13—Sensors therefor
- G06V40/1324—Sensors therefor by using geometrical optics, e.g. using prisms
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Human Computer Interaction (AREA)
- Multimedia (AREA)
- Optics & Photonics (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The invention discloses a device capable of simultaneously acquiring images of palm veins and palm prints, and a method for enhancing and segmenting the images of the palm veins and the palm prints by adopting the device. The device comprises: the LED light source module comprises a shell, a lens, a near-infrared LED light source group, a visible white LED light source group, an annular light-equalizing material, a visible CCD sensor, a near-infrared CCD sensor, a first semi-transparent semi-reflective lens and a second semi-transparent semi-reflective lens. The invention uses one lens to simultaneously acquire the palm vein picture and the palm print picture, thereby ensuring the region consistency of the palm vein picture and the palm print picture; the semi-transparent semi-reflective lens is used for separating light rays with different wavelengths under the same lens, and then the near-infrared CCD sensor and the visible light CCD sensor are used for respectively acquiring pictures of the palm veins and the palm prints, so that the definition of the pictures of the palm veins and the palm prints is guaranteed; and the acquired palm vein and palm print images are enhanced and segmented, so that the phenomenon of uneven gray values of palm areas can be overcome, and the palm vein, the palm print and the skin area can be accurately distinguished.
Description
Technical field
The present invention relates to living things feature recognition fields, specially a kind of to acquire palm vein image simultaneously using single camera lens
With the device of palmprint image, the invention further relates to enhancing and the dividing methods of a kind of low contrast vena metacarpea and palmprint image.
Background
Currently, comparative maturity and most promising several biometrics identification technologies include fingerprint recognition, iris
Identification, face recognition, speech recognition, palm shape identification, signature recognition etc..But above-mentioned biometrics identification technology have it is some total
Same defect: first, it is affected by environment larger；Second, theoretically it can be replicated and usurp.
Hand vein identification technology is the new biometrics identification technology of one kind proposed recent years, have uniqueness,
Stability, unforgeable, it is contactless the advantages that.Hand vein identification technology includes palm vein, finger vena and the back of the hand
Three kinds of forms of hand vein recognition.Wherein the advantages of palm vein identification technology includes two o'clock: first, vena metacarpea relative to refer to vein and
Speech, blood vessel is relatively thick and is located under epidermis, therefore is easy to capture vena metacarpea image；Second, vena metacarpea is relative to hand back vein
For structure, geometry is more complex, can be improved the accuracy of identification.But vena metacarpea structure be only by several compared with
Thick vein blood vessel is constituted, and is unable to satisfy the application requirement in high safety field.It is special that the palmmprint structure of people can be used as a kind of biology
It takes over for use in identification, if the feature of fusion vena metacarpea and palmmprint, undoubtedly can be improved the accuracy of identification.
Disclosure of the invention number is the patent of CN101833647B: " the acquisition equipment of palmprint image and palmprint image processing side
Method " discloses a kind of device for acquiring palmprint image and recognition methods, but palmmprint is relatively simple for structure, relies solely on palmmprint
Structure carries out identification, and there are some potential safety problemss.
Disclosure of the invention number be CN101196987B patent: " online palmprint, palm vein image personal identification method and its
Special-purpose collection instrument ".A kind of device that can acquire palmmprint and vena metacarpea is disclosed, but its collection process is visible alternately to open
Light LED and near-infrared LED, to obtain palmmprint and vena metacarpea image respectively.Alternately due to shooting, this method can not be simultaneously
Capture the completely the same palmmprint in region and vena metacarpea image；The process of video camera imaging needs centainly when due to light source switching
Time results in and increases identification total time；In addition, near-infrared palm vein image contrast is lower, this method, which is not studied, to be had
The image enchancing method of effect.
Summary of the invention
In order to overcome problems of the prior art, the object of the present invention is to provide one kind, can to acquire simultaneously palm quiet
The device of arteries and veins and palmprint image, and enhancing and dividing method are carried out using above-mentioned apparatus acquisition palm vein and palmprint image.
To achieve the goals above, the invention adopts the following technical scheme: under one kind is single-lens while acquiring palm vein
With the device of palmprint image, described device includes:
Shell, four face closures, front aperture is for installing camera lens；
Camera lens passes through shell, luminous ray and near infrared light can be penetrated simultaneously, for adopting to palm image
Collection；
Light source group, surrounding lens, light directive camera lens front, for carrying out illumination and light filling to palm；
Cyclic annular equal luminescent material, is arranged immediately ahead of light source group, and size can shelter from light source group completely, is used to form uniformly
Light；
Ccd sensor group is arranged in shell, for acquiring the image for penetrating camera lens；
Lens set is arranged in shell, and the image for will transmit through camera lens sends ccd sensor group to；
Control storage PC, connects ccd sensor group, for enhancing the palm vein received and palmprint image
And dividing processing, and stored.
The light source group includes near-infrared LED light source group and visible white light LED light source group, the near-infrared LED light source group
For 8 850nm near-infrared LED lamps, it is seen that white LED light source group is 8 visible white light LED light.
The ccd sensor group includes Visible-light CCD sensor and Near Infrared CCD sensor；The Visible-light CCD sensing
The dead astern of camera lens is arranged in device, and the central axis of the two is point-blank, for acquiring the transmitting of visible white light LED light source group
Light after palmar, through the visible light palmprint image of camera lens；The Near Infrared CCD sensor is arranged in shell
Top, just facing towards with horizontal plane angle be 225 degree, for acquire near-infrared LED light source group emit light through palm
After reflection, through the near-infrared palm vein image of camera lens.
The lens set includes two semi-transparent semi-reflecting eyeglasses, and the first semi-transparent semi-reflecting eyeglass is arranged in camera lens and Visible-light CCD
Among sensor, the angle of the first semi-transparent semi-reflecting eyeglass and horizontal plane is 67.5 degree, for transmitting the luminous ray for penetrating camera lens,
The near infrared light of reflectance-transmittance camera lens to the second semi-transparent semi-reflecting eyeglass, the second semi-transparent semi-reflecting eyeglass reflects near infrared light simultaneously
To Near Infrared CCD sensor.
Described device further includes that palm places bracket and middle finger fixed card slot, and it is to be arranged in shell that the palm, which places bracket,
Outside front, camera lens, for limiting the distance between palm and camera lens, the setting of middle finger fixed card slot is placed on bracket in palm,
For keeping the consistency of acquisition palm vein and palmprint image.
Method that is a kind of while acquiring palm vein and palmprint image, comprising the following steps:
(1) palm is placed on 15cm before camera lens, passes through Visible-light CCD sensor and Near Infrared CCD sensor obtains hand
Vena metacarpea and palmprint image；
(2) collected palm vein and palmprint image are split；
(3) palm vein and palmprint image are enhanced based on Retinex iterative filtering；
(4) binarization segmentation is carried out to enhanced palm vein and palmprint image；
(5) to the palm vein and the progress authenticity judgement of palmmprint structure, removal noise and void in the image after binaryzation
False knot structure.
The step 2 is using acquired each 50 pixels of image surrounding of deletion.
The step 3 the following steps are included:
(1) with the mean filter of 3*3 to after segmentation palm vein and palmprint image carry out smothing filtering denoising,
Using following formula:
Wherein I0(x, y) is the palm vein and palmprint image after region of interest regional partition, and I (x, y) is that smothing filtering is gone
Image after making an uproar；
(2) palm vein and palmprint image are enhanced using improved Retinex algorithm:
A, the variation degree d (x, y) of each neighborhood of pixel points pixel of image is calculated using following formula:
D (x, y)=| I (x, y+1)-I (x, y-1) |+| I (x+1, y)-I (x-1, y) |
Wherein I (x, y) be video camera shooting image, d (x, y) be each pixel left and right pixel value difference and up and down
The sum of the absolute value of picture element interpolation；
B, dynamic filter window function is calculated using following formula:
W (x, y)=(1+0.5d2(x,y))-1
C, it is iterated filtering 20 times using each element of the dynamic filter window function w (x, y) to image I (x, y), calculated
Ambient light component L (x, y):
Lt+1(x, y)=max (L 't+1(x,y),Lt(x,y))
Wherein: L0(x, y)=I (x, y)
Wherein I (x, y) is original input picture, and L (x, y) is the ambient light component acquired.The above process is iterative process,
Initial L0(x, y)=I (x, y), N (x, y) be the cumulative of the 3*3 neighborhood of spectral window w (x, y) and；
D, using Retinex algorithm, enhanced image R (x, y) is calculated, and normalizes to [0,1]；
R (x, y)=logI (x, y)-logL (x, y)
R0(x, y)=(R (x, y)-min (R))/(max (R)-min (R))
R0(x, y) is balanced and enhanced palm vein and palmprint image；
(3) to palm vein and palmprint image degree of comparing stretch processing:
A、R0(x, y) is if=0 R0(x,y)<0.6
R0(x, y)=2*R0(x, y) is if -1 R0(x,y)≥0.6
B, gray scale stretching processing is carried out to image using gray scale cosine transform, obtains R1(x, y), calculation formula are as follows:
R1(x, y)=1-cos (0.5* π * R0(x,y))
C, using the Gaussian filter of 3*3 to R1(x, y) is filtered denoising.
It is split in the step 4 using the method opponent palmmprint and palm vein image of global binaryzation, palm line
Optimum segmentation threshold value be 0.45, the optimum segmentation threshold value of palm vein picture is 0.55.
The step 5 the following steps are included:
(1) picture after segmentation being handled using morphological operation, progress morphological dilations first eliminate small holes,
Small gap is connected simultaneously, and morphological erosion is recycled to restore the width of original vena metacarpea and palmmprint；
(2) blackspot spot noise small in binary image is removed；
(3) false vena metacarpea and palmmprint structure are removed.
The beneficial effects of the present invention are: (1) present invention acquires vena metacarpea and palmmprint picture simultaneously using a camera lens, ensure
The region consistency of vena metacarpea picture and palmmprint picture；(2) semi-transparent semi-reflecting eyeglass is used, to different wave length under the same camera lens
Light separated, recycle Near Infrared CCD sensor and Visible-light CCD sensor to collect vena metacarpea and palmmprint respectively
Picture has ensured the clarity of vena metacarpea and palmmprint picture；(3) vena metacarpea of the invention and palmprint image enhancing and partitioning algorithm
The gray value uneven phenomenon that palm area can be overcome accurately distinguishes vena metacarpea, palmmprint and skin area, and this method
Computation complexity it is smaller, in core i7-3770,3.4GHz, on the computer of 4G memory, to the vena metacarpea of 160*120 size
With the average value about 0.1s of the processing time of palmmprint picture, meet the requirement calculated in real time.
Detailed description of the invention
Fig. 1 is the sectional view of embodiment vena metacarpea and palmprint image collecting device.
Fig. 2 is the external structure of embodiment vena metacarpea and palmprint image collecting device.
Fig. 3 is embodiment vena metacarpea and palm-print image capture, enhances and divide flow chart.
Fig. 4 is embodiment vena metacarpea (a) and palmmprint (b) picture area-of-interest segmentation result figure.
Fig. 5 is embodiment vena metacarpea (a) and palmmprint (b) picture reinforcing effect figure.
Fig. 6 is that embodiment vena metacarpea (a) and palmmprint (b) picture compare drawing effect figure.
Fig. 7 is embodiment vena metacarpea (a) and palmmprint (b) picture segmentation effect picture.
Fig. 8 is embodiment vena metacarpea (a) and palmmprint (b) picture Morphological scale-space effect picture.
Fig. 9 is that embodiment vena metacarpea (a) and palmmprint (b) picture denoise effect picture.
Figure 10 is that embodiment vena metacarpea (a) and palmmprint (b) picture remove false structure effect picture.
Specific embodiment
The present invention is further illustrated with attached drawing combined with specific embodiments below.
Vena metacarpea and palmprint image collecting device under a kind of single camera lens, sectional view as shown in Figure 1, its external structure such as
Shown in Fig. 2.
As shown in Figure 1, device includes: shell 1, camera lens 2, near-infrared LED light source group 3, visible white light LED light source group 4, ring
The equal luminescent material 5 of shape, Visible-light CCD sensor 6, Near Infrared CCD sensor 7, the first semi-transparent semi-reflecting eyeglass 8 and second are semi-transparent semi-reflecting
Eyeglass 9.
As shown in Fig. 2, device further includes that palm places bracket 10 and middle finger fixed card slot 11.
Shell effect is immobilising device, while the light in environment being avoided to influence, it is desirable that it is opaque, do not have to material
Particular/special requirement, four face closures, front aperture is for installing camera lens.Camera lens is the common general camera lens of visible light and near infrared light,
Luminous ray and near infrared light can be penetrated simultaneously.850nm near-infrared LED group shares LED light 8, is looped around camera lens week
It encloses, the light source immediately ahead of light directive camera lens, as shooting vena metacarpea image.Visible white light LED group shares LED light 8, surround
Light source around near-infrared LED group, immediately ahead of light directive camera lens, as shooting palmprint image.Annular luminescent material is placed
In the front of two groups of light sources, and size can shelter from two groups of light sources completely, effect be to be formed uniform near-infrared and
Luminous ray, to collect balanced vena metacarpea and palmprint image.
Visible-light CCD sensor is placed on the dead astern of camera lens, and the central axis of the two is point-blank, effect
Be: the light of acquisition visible white light LED group transmitting is after palmar, through the visible light palmprint image of camera lens.Near Infrared CCD
Sensor is located at the top of enclosure interior, and just facing towards being 225 degree with horizontal plane angle, effect is: acquisition second half
The near infrared light of saturating semi-reflective mirror piece reflection, to shoot near-infrared palm vein image.
Semi-transparent semi-reflecting eyeglass has 2, can penetrate luminous ray, while reflecting near infrared light.First semi-transparent semi-reflecting lens
Piece is placed among camera lens and Visible-light CCD sensor, and the angle of eyeglass and horizontal plane is 67.5 degree, and effect is: transmitting lens
The luminous ray of head transmitting, such visible LED can capture palmprint image, while the near infrared light of reflection lens is to the
Two semi-transparent semi-reflecting eyeglasses.Second semi-transparent semi-reflecting eyeglass is placed on the top of left shell, and eyeglass is perpendicular to horizontal plane, effect
It is: the near infrared light of secondary reflection first semi-transparent semi-reflecting eyeglass reflection to Near Infrared CCD sensor again.
The palm vein and palmprint image that Visible-light CCD sensor and Near Infrared CCD sensor receive send control to
Storage PC carries out enhancing and dividing processing to image, and is stored.
It is to be arranged outside shell front, camera lens that palm, which places bracket, high for limiting the distance between palm and camera lens
15 centimetres of degree.The setting of middle finger fixed card slot is placed on bracket in palm, and effect is the position of fixed middle finger and palm intersection, is used
In the consistency for keeping acquisition palm vein and palmprint image.
Vena metacarpea and palm-print image capture, enhancing and segmentation flow chart are shown in Fig. 3, using palm vein and palm-print image capture
The palmmprint and vena metacarpea picture contrast that device takes are lower, if being used directly to be identified, it is difficult to obtain higher knowledge
Not rate, it is therefore desirable to carry out image enhancement and dividing processing, steps are as follows:
Step 1: palm being placed on 15cm before camera lens, passes through Visible-light CCD sensor and Near Infrared CCD sensor obtains
Palm vein and palmprint image.
Step 2: " region of interest regional partition " being carried out to the original image taken, obtains " standard palmmprint and vena metacarpea figure
Picture ".
Specific embodiment are as follows: delete original vena metacarpea collected and each 50 pixels of palmprint image surrounding, effect
It is to remove hand edge vena metacarpea and the less region of palmmprint structure, to obtain structural information key area abundant, obtains
Region of interest area image such as Fig. 4.
Step 3: vena metacarpea and palmprint image enhancing based on Retinex iterative filtering.
Specific embodiment are as follows:
1, smothing filtering denoises.
The disposal of gentle filter is carried out to collected vena metacarpea and palmprint image with the mean filter of 3*3.Due to shooting
To original vena metacarpea and palmprint image contain certain noise, therefore first using the smoothing filter of a 3*3 to image
Smothing filtering is carried out, such as following formula:
Wherein I0(x, y) is the vena metacarpea and palmprint image after region of interest regional partition, and I (x, y) is smothing filtering denoising
Image.
2, improved Retinex algorithm enhances vena metacarpea and palmprint image.
Although acquisition device has used annular equal luminescent material disperse light, relatively uniform light is formd, is adopted
The gray value of the image collected is still less uniform；In addition, since the collected picture contrast of institute is lower, it is therefore desirable to study
Illumination equilibrium and structure enhance algorithm.
Retinex theory is a set of color theory that Edwin Land is proposed, can be applied by research in recent years
Field of image processing.The core concept of Retinex theory is: perception of the human eye to an object color, changes in ambient light
Or in the case where body surface uneven illumination still can accurate judgement, be because the vision system of the mankind is able to carry out at certain
Reason, eliminates the disturbing factors such as the intensity of light source.Irradiation light can be removed from the image taken by Retinex theory,
To obtain reflectivity properties possessed by object.Although near-infrared image does not have color, it is equally applicable Retinex theory
Eliminate the gray scale uneven phenomenon of vena metacarpea and palmprint image.
Retinex theory is defined as follows:
R (x, y)=logI (x, y)-logL (x, y)
Wherein I (x, y) is the image of video camera shooting, and L (x, y) is ambient light component, and R (x, y) is gray scale balance and increasing
Image after strong；By carrying out gaussian filtering to original image come simulated environment light component L (x, y), this hair in standard Retinex theory
Bright ambient light component L (x, y) calculating process is as follows:
(1) the variation degree d (x, y) for calculating each neighborhood of pixel points pixel of image, such as following formula:
D (x, y)=| I (x, y+1)-I (x, y-1) |+| I (x+1, y)-I (x-1, y) |
Discovery is easy from formula, d (x, y) represents the left and right pixel value difference and upper and lower picture element interpolation of each pixel
The sum of absolute value can represent the severe degree of local neighborhood variation.
(2) calculating dynamic filter window function w (x, y) is the function for increasing with certain pixel vegetarian refreshments gradient and successively decreasing, such as
Following formula:
W (x, y)=(1+0.5d2(x,y))-1
It is easy discovery from formula, neighborhood of pixel points variation degree is higher, and the value of the filtering window function is smaller.That is: when d (x,
Y) smooth to weaken when larger (grey scale change is larger, it may be possible to edge), on the contrary enhancing smooth effect.In this way, flat in smoothed image
While smooth region, marginal information is remained.
(3) filtering 20 times, meter are iterated to each element of image I (x, y) using dynamic filter window function w (x, y)
It calculates ambient light component L (x, y):
Lt+1(x, y)=max (L 't+1(x,y),Lt(x,y))
Wherein:
L0(x, y)=I (x, y)
Wherein I (x, y) is original input picture, and L (x, y) is the ambient light component acquired.The above process is iterative process,
Initial L0(x, y)=I (x, y), N (x, y) be the cumulative of the 3*3 neighborhood of spectral window w (x, y) and, effect is before ensureing iteration
The consistency of the interval value of pixel afterwards.The actual effect for solving the larger value of the pixel before and after iteration is to increase skin area
Overall brightness, increase the contrast of picture.
(4) enhanced image is solved, and normalizes to [0,1].
After solving ambient light component, the formula of Retinex algorithm can use, directly calculate balanced and enhanced
Vena metacarpea and palmmprint picture；
R (x, y)=logI (x, y)-logL (x, y)
Pixel value is not or not [0,1] section at this time, it is therefore desirable to pixel value be normalized to [0,1] range using following formula.
R0(x, y)=(R (x, y)-min (R))/(max (R)-min (R))
So far, available enhanced vena metacarpea and palmmprint picture R0(x, y), as shown in Figure 5.
3, contrast stretching is handled.
The gray value of picture is generally higher after above-mentioned calculating, and contrast is lower, it is therefore desirable to further do contrast stretching
Processing, detailed process are as follows:
(1)R0(x, y) is if=0 R0(x,y)<0.6
R0(x, y)=2*R0(x, y) is if -1 R0(x,y)≥0.6
Enhanced vena metacarpea and palmmprint picture R0The gray value of (x, y) is concentrated mainly in the range of [0.6,1], therefore
Using the transformation, its gray scale interval can be stretched, increases contrast.
(2) it reuses gray scale cosine transform and gray scale stretching processing is carried out to image, obtain R1(x, y) further increases the palm
The grey value difference of vein, palmmprint and skin, calculation formula are as follows:
R1(x, y)=1-cos (0.5* π * R0(x,y))
(3) finally using the Gaussian filter of 3*3 to R1(x, y) is filtered denoising, and the enhanced palm is quiet
Arteries and veins and palmprint image are as shown in Figure 6.
Step 4: binarization segmentation is carried out to enhanced vena metacarpea and palmprint image.
Gray value by step 2 treated vena metacarpea and palmmprint picture is more balanced, and vena metacarpea, palmmprint and skin it
Between contrast it is larger, therefore the method that global binaryzation can be used is split palmmprint and vena metacarpea image, palmmprint
Optimum segmentation threshold value is 0.45, and the optimum segmentation threshold value of vena metacarpea picture is 0.55, treatment effect such as Fig. 7.
Step 5: to the vena metacarpea and the progress authenticity judgement of palmmprint structure, removal noise and void in the image after binaryzation
False knot structure.
Specific embodiment is as follows:
(1) picture after segmentation being handled using morphological operation, progress morphological dilations first eliminate small holes,
Small gap is connected simultaneously, and morphological erosion is recycled to restore the width of original vena metacarpea and palmmprint；Treatment effect such as Fig. 8.
(2) remove small black patch noise in binary image, i.e., if in 5 × 5 neighborhoods of a pixel there are five or
More grey scale pixel values are 0, then are set as 0, are otherwise provided as 1；Treatment effect such as Fig. 9.
(3) false vena metacarpea and palmmprint structure are removed, the method is as follows: be marked, count to black block in binary image
Calculate the length and width (image size 160*120) of every piece of area (pixel number), the boundary rectangle for determining every piece, then according to
Lower situation is handled:
● if black patch area deletes this black patch less than 150.Palmmprint and vena metacarpea structure have continuity, have
Number of pixel is more, and the black patch of zonule is usually noise, stain or small opacities.
● if black patch area judges the fold differences D of its length and width between 150 to 600, if D is deleted less than 5
This black patch.It is analyzed as follows: for area in the black patch of 150-600 range, usually single vena metacarpea or palmmprint, having single
Direction, the length of black patch and wide difference are generally large at this time, by a large amount of statistical experiments, the length of single vena metacarpea and palmmprint structure
Wide difference value is greater than 5.This biggish false black patch is typically created in the shaded side at four turnings of image.
● if black patch area is greater than 600, retains this black patch, is true vena metacarpea and palmmprint structure.
, can be accurately not high from contrast by above step, the less uniform vena metacarpea of gray value and palmmprint picture
In extract true structure, such as Figure 10.It can be seen that: distant vena metacarpea and palmmprint structure are gem-pure at this time aobvious originally
It shows and.
The above described is only a preferred embodiment of the present invention, not doing any type of limitation to the present invention.It is all
Any simple modification, equivalent change and modification substantially to the above embodiments of technology and methods according to the present invention, still
Belong in the range of technology and methods scheme of the invention.
Claims (3)
1. a kind of device that is single-lens lower while acquiring palm vein and palmprint image, it is characterised in that；Described device includes:
Shell, four face closures, front aperture is for installing camera lens；
Camera lens passes through shell, luminous ray and near infrared light can be penetrated simultaneously, for being acquired to palm image；
Light source group, surrounding lens, light directive camera lens front, for carrying out illumination and light filling to palm；
Cyclic annular equal luminescent material, is arranged immediately ahead of light source group, size can shelter from light source group completely, be used to form uniform light
Line；
Ccd sensor group is arranged in shell, for acquiring the image for penetrating camera lens；
Lens set is arranged in shell, and the image for will transmit through camera lens sends ccd sensor group to；
Control storage PC, connects ccd sensor group, for being enhanced and being divided the palm vein received and palmprint image
Processing is cut, and is stored；
The light source group includes near-infrared LED light source group and visible white light LED light source group, and the near-infrared LED light source group is 8
850nm near-infrared LED lamp, is looped around around camera lens, immediately ahead of light directive camera lens, it is seen that white LED light source group is 8 visible
White LED lamp is looped around around near-infrared LED group, immediately ahead of light directive camera lens；
The ccd sensor group includes Visible-light CCD sensor and Near Infrared CCD sensor；The Visible-light CCD sensor is set
It sets in the dead astern of camera lens, and the central axis of the two is point-blank, for acquiring the light of visible white light LED light source group transmitting
Line is after palmar, through the visible light palmprint image of camera lens；The Near Infrared CCD sensor is arranged on shell is intracorporal
Portion, just facing towards being 225 degree with horizontal plane angle, for acquiring the light of near-infrared LED light source group transmitting through palmar
Afterwards, through the near-infrared palm vein image of camera lens；
The lens set includes two semi-transparent semi-reflecting eyeglasses, and the first semi-transparent semi-reflecting eyeglass setting is sensed in camera lens and Visible-light CCD
Among device, the angle of the first semi-transparent semi-reflecting eyeglass and horizontal plane is 67.5 degree, for transmitting the luminous ray for penetrating camera lens, simultaneously
For the near infrared light of reflectance-transmittance camera lens to the second semi-transparent semi-reflecting eyeglass, the second semi-transparent semi-reflecting eyeglass reflects near infrared light to close
Infrared CCD sensor.
2. the apparatus according to claim 1, it is characterised in that: described device further includes that palm placement bracket and middle finger are fixed
Card slot, it is to be arranged outside shell front, camera lens that the palm, which places bracket, for limiting the distance between palm and camera lens,
The setting of middle finger fixed card slot is placed on bracket in palm, for keeping the consistency of acquisition palm vein and palmprint image.
3. a kind of according to claim 1 to the method that 2 any described devices acquire palm vein and palmprint image simultaneously, feature
Be the following steps are included:
(1) palm is placed on 15cm before camera lens, it is quiet to obtain palm by Visible-light CCD sensor and Near Infrared CCD sensor
Arteries and veins and palmprint image；
(2) collected palm vein and palmprint image are split；
(3) palm vein and palmprint image are enhanced based on Retinex iterative filtering；
(4) binarization segmentation is carried out to enhanced palm vein and palmprint image；
(5) to the palm vein and the progress authenticity judgement of palmmprint structure, removal noise and false knot in the image after binaryzation
Structure；The step 2 is using acquired each 50 pixels of image surrounding of deletion；
The step 3 the following steps are included:
(1) with the mean filter of 3*3 to after segmentation palm vein and palmprint image carry out smothing filtering denoising, use
Following formula:
Wherein I0(x, y) is the palm vein and palmprint image after region of interest regional partition, and I (x, y) is after smothing filtering denoises
Image；
(2) palm vein and palmprint image are enhanced using improved Retinex algorithm:
A, the variation degree d (x, y) of each neighborhood of pixel points pixel of image is calculated using following formula:
D (x, y)=| I (x, y+1)-I (x, y-1) |+| I (x+1, y)-I (x-1, y) |
Wherein I (x, y) be video camera shooting image, d (x, y) be each pixel left and right pixel value difference and upper and lower pixel
The sum of the absolute value of interpolation；
B, dynamic filter window function is calculated using following formula:
W (x, y)=(1+0.5d2(x,y))-1
C, it is iterated filtering 20 times using each element of the dynamic filter window function w (x, y) to image I (x, y), calculates environment
Light component L (x, y):
Lt+1(x, y)=max (L 't+1(x,y),Lt(x,y))
Wherein: L0(x, y)=I (x, y)
Wherein I (x, y) is original input picture, and L (x, y) is the ambient light component acquired；The above process is iterative process, just
Beginning L0(x, y)=I (x, y), N (x, y) be the cumulative of the 3*3 neighborhood of spectral window w (x, y) and；
D, using Retinex algorithm, enhanced image R (x, y) is calculated, and normalizes to [0,1]；
R (x, y)=logI (x, y)-logL (x, y)
R0(x, y)=(R (x, y)-min (R))/(max (R)-min (R))
R0(x, y) is balanced and enhanced palm vein and palmprint image；
(3) to palm vein and palmprint image degree of comparing stretch processing:
A、R0(x, y) is if=0 R0(x, y) < 0.6
R0(x, y)=2*R0(x, y) is if -1 R0(x,y)≥0.6
B, gray scale stretching processing is carried out to image using gray scale cosine transform, obtains R1(x, y), calculation formula are as follows:
R1(x, y)=1-cos (0.5* π * R0(x,y))
C, using the Gaussian filter of 3*3 to R1(x, y) is filtered denoising；
It is split in the step 4 using the method opponent palmmprint and palm vein image of global binaryzation, palm line is most
Excellent segmentation threshold is 0.45, and the optimum segmentation threshold value of palm vein picture is 0.55；
The step 5 the following steps are included:
(1) picture after segmentation is handled using morphological operation, progress morphological dilations first eliminate small holes, simultaneously
Small gap is connected, morphological erosion is recycled to restore the width of original vena metacarpea and palmmprint；
(2) blackspot spot noise small in binary image is removed；
(3) false vena metacarpea and palmmprint structure are removed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510801840.2A CN105426843B (en) | 2015-11-19 | 2015-11-19 | Single-lens lower palm vein and palm print image acquisition device and image enhancement and segmentation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510801840.2A CN105426843B (en) | 2015-11-19 | 2015-11-19 | Single-lens lower palm vein and palm print image acquisition device and image enhancement and segmentation method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105426843A CN105426843A (en) | 2016-03-23 |
| CN105426843B true CN105426843B (en) | 2019-02-22 |
Family
ID=55505043
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510801840.2A Active CN105426843B (en) | 2015-11-19 | 2015-11-19 | Single-lens lower palm vein and palm print image acquisition device and image enhancement and segmentation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105426843B (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106236016B (en) * | 2016-08-18 | 2019-02-26 | 皖江新兴产业技术发展中心 | A kind of in-situ projection method for vein imaging |
| CN107038445B (en) * | 2017-02-13 | 2021-01-12 | 上海大学 | Binarization and segmentation method for Chinese character verification code |
| CN107977935A (en) * | 2017-11-15 | 2018-05-01 | 合肥中科奔巴科技有限公司 | A kind of vena metacarpea image texture filtering reinforcement method |
| CN109447052A (en) * | 2019-01-09 | 2019-03-08 | 东浓智能科技(上海)有限公司 | A kind of vena metacarpea identification device being accurately positioned palm position and its implementation |
| CN111738076A (en) * | 2020-05-18 | 2020-10-02 | 山东大学 | Non-contact palm print and palm vein identification method and device |
| TWI781459B (en) * | 2020-10-08 | 2022-10-21 | 國立中興大學 | Palm vein feature identification system and method |
| CN112241720A (en) * | 2020-10-27 | 2021-01-19 | 上海掌腾信息科技有限公司 | Method for realizing palm vein image self-adaptive extraction processing of palm vein recognition device aiming at palm inclined condition |
| CN112241721B (en) * | 2020-10-27 | 2022-07-05 | 上海掌腾信息科技有限公司 | Method for realizing complete palm vein image acquisition based on light irradiation of different wavelengths |
| CN114155296A (en) * | 2021-11-19 | 2022-03-08 | 宁波芯然科技有限公司 | Method for determining central area of palm image |
| CN117593341B (en) * | 2024-01-19 | 2024-05-07 | 深圳市超诺科技有限公司 | System and method for processing target object monitoring data based on hunting camera |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103116741A (en) * | 2013-01-28 | 2013-05-22 | 天津理工大学 | Capture and identification system for blending images of palm veins and palm prints |
| US20140019489A1 (en) * | 2012-07-10 | 2014-01-16 | Jinjun Wang | Constructing Incremental Tree Model for Vein Image Recognition and Authentication |
| CN104217195A (en) * | 2014-08-07 | 2014-12-17 | 中国矿业大学 | A stereo imaging and recognition system for hand veins |
-
2015
- 2015-11-19 CN CN201510801840.2A patent/CN105426843B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140019489A1 (en) * | 2012-07-10 | 2014-01-16 | Jinjun Wang | Constructing Incremental Tree Model for Vein Image Recognition and Authentication |
| CN103116741A (en) * | 2013-01-28 | 2013-05-22 | 天津理工大学 | Capture and identification system for blending images of palm veins and palm prints |
| CN104217195A (en) * | 2014-08-07 | 2014-12-17 | 中国矿业大学 | A stereo imaging and recognition system for hand veins |
Non-Patent Citations (3)
| Title |
|---|
| 基于Retinex和小波分解的手背静脉识别方法;王华彬等;《系统工程理论与实践》;20140225;第34卷(第2期);第428-431页 * |
| 基于近红外成像的手背静脉识别系统研究与实现;陈芳;《中国优秀硕士学位论文全文数据库》;20131115(第11期);正文第5-26页 * |
| 手部三模态图像采集系统的设计与实现;苑玮琦等;《传感器与微系统》;20130528;第32卷(第5期);第78-81, 85页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105426843A (en) | 2016-03-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105426843B (en) | Single-lens lower palm vein and palm print image acquisition device and image enhancement and segmentation method | |
| KR101356358B1 (en) | Computer-implemented method and apparatus for biometric authentication based on images of an eye | |
| US9317761B2 (en) | Method and an apparatus for determining vein patterns from a colour image | |
| CN101999900B (en) | Living body detecting method and system applied to human face recognition | |
| WO2010129074A1 (en) | System and method for identifying a person with reference to a sclera image | |
| CN108416291B (en) | Face detection and recognition method, device and system | |
| JP2018106720A (en) | Apparatus and method for image processing | |
| Das et al. | Fuzzy logic based selera recognition | |
| Das et al. | Multi-angle based lively sclera biometrics at a distance | |
| Das et al. | A new method for sclera vessel recognition using OLBP | |
| CN114565531A (en) | Image restoration method, device, equipment and medium | |
| KR101174103B1 (en) | A face recognition method of Mathematics pattern analysis for muscloskeletal in basics | |
| Yang et al. | Finger-vein recognition based on Gabor features | |
| Szczepański et al. | Pupil and iris detection algorithm for near-infrared capture devices | |
| Du et al. | A one-dimensional approach for iris identification | |
| CN113901423B (en) | Intelligent security equipment control method and system based on face recognition | |
| Li et al. | Iris recognition on mobile devices using near-infrared images | |
| Sharma et al. | Improved human identification using finger vein images | |
| Soltany et al. | Daugman’s Algorithm Enhancement for Iris Localization | |
| CN116982093A (en) | Presence attack detection | |
| Crihalmeanu et al. | Multispectral ocular biometrics | |
| Memar Zadeh et al. | Iris localization by means of adaptive thresholding and Circular Hough Transform | |
| Al-Hameed et al. | Face detection based on skin color segmentation and eyes detection in the human face | |
| Marcon et al. | Sclera segmentation for gaze estimation and iris localization in unconstrained images. | |
| AlRifaee et al. | Adaptive Segmentation for Unconstrained Iris Recognition |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |