CN107724900B - A kind of family security door based on personal recognition - Google Patents
A kind of family security door based on personal recognition Download PDFInfo
- Publication number
- CN107724900B CN107724900B CN201710900903.9A CN201710900903A CN107724900B CN 107724900 B CN107724900 B CN 107724900B CN 201710900903 A CN201710900903 A CN 201710900903A CN 107724900 B CN107724900 B CN 107724900B
- Authority
- CN
- China
- Prior art keywords
- roi region
- denoising
- palmmprint
- transformation
- image
- 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
Classifications
-
- E—FIXED CONSTRUCTIONS
- E06—DOORS, WINDOWS, SHUTTERS, OR ROLLER BLINDS IN GENERAL; LADDERS
- E06B—FIXED OR MOVABLE CLOSURES FOR OPENINGS IN BUILDINGS, VEHICLES, FENCES OR LIKE ENCLOSURES IN GENERAL, e.g. DOORS, WINDOWS, BLINDS, GATES
- E06B5/00—Doors, windows, or like closures for special purposes; Border constructions therefor
- E06B5/10—Doors, windows, or like closures for special purposes; Border constructions therefor for protection against air-raid or other war-like action; for other protective purposes
- E06B5/11—Doors, windows, or like closures for special purposes; Border constructions therefor for protection against air-raid or other war-like action; for other protective purposes against burglary
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
- G07C9/00—Individual registration on entry or exit
- G07C9/00174—Electronically operated locks; Circuits therefor; Nonmechanical keys therefor, e.g. passive or active electrical keys or other data carriers without mechanical keys
- G07C9/00563—Electronically operated locks; Circuits therefor; Nonmechanical keys therefor, e.g. passive or active electrical keys or other data carriers without mechanical keys using personal physical data of the operator, e.g. finger prints, retinal images, voicepatterns
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B19/00—Alarms responsive to two or more different undesired or abnormal conditions, e.g. burglary and fire, abnormal temperature and abnormal rate of flow
Abstract
The present invention provides a kind of family security door based on personal recognition, and the family security door is equipped with palmmprint acquisition module, information processing centre and monitoring center；The monitoring center connects electronic lock；The palm print collecting device is used to acquire the palmprint image of enabling people；Palmprint image of the described information processing center for the people that opens the door is identified, and identification result is transferred to monitoring center；The monitoring center is used to control the opening and closing of electronic lock according to recognition result.The present invention is identified identity using palm print and palm vein cognitron, improves identification accuracy, using Touchless manipulation, use more freely, safety, health, efficiently；The present invention also acquires the image information and environmental information on doorway, exception and alarm occurs, reduces resident family's loss, improves safety；Framework of the present invention is simple, at low cost, and intelligence degree is high, and safety is good, is suitble to promote the use of.
Description
Technical field
The present invention relates to antitheft door fields, more particularly, to a kind of family security door based on personal recognition.
Background technology
Now due to urbanization is universal, people live a cell together but from different places, interpersonal letter
Sense is appointed to decline, more and more families have used burglary-resisting installation, especially antitheft door, and it is big that existing antitheft door carries out identification
IC card, image recognition, fingerprint recognition etc. are mostly used, but its safety is relatively low and is easy to be stolen by offender, to people's
Life, property safety bring hidden danger.Compared with the biological characteristics such as face, the condition that palmmprint vena metacarpea obtains is relatively easy to control, less
Ground is by interference such as light, expression, makeups, and situations such as twins can be effectively treated, therefore can preferably accuracy of identification, acquisition side
Formula is more convenient.
Invention content
In view of the above-mentioned problems, the present invention is intended to provide a kind of family security door based on personal recognition, is pacified with improving family
The safety and reliability of anti-management.
The purpose of the present invention is realized using following technical scheme：
A kind of family security door based on personal recognition, it is characterised in that：The family security door is acquired equipped with palmmprint
Module, information processing centre and monitoring center；
The palm print collecting device is used to acquire the palmprint image of enabling people；
Described information processing center is used to be identified according to the identity of the palmprint image clamshell doors people of enabling people, and will know
Other result is transferred to monitoring center；
The monitoring center is used to control the opening and closing of electronic lock according to handling result.
The monitoring center include controller and the power supply being connect with the controller, warning device, lighting device and
Memory；
The controller is microcontroller, the unlatching for controlling electronic lock and closure, when identification result is correct, control
Device controls electronic lock and opens；When identification result is incorrect, electronic lock is not turned on, while the warning device carries out alarm and carries
It wakes up；
The power supply is used to power to controller, ensures that controller is in normal operating conditions；The lighting device is used for
Automatic illuminating is carried out at night, when there are abnormal conditions or power-off, automatic illuminating is carried out by built-in accumulator, is prevented
It is flurried to occur；The memory is used to store the identification result of information processing centre.
The family security door further includes information collecting center；Described information acquire center include image acquisition device, it is infrared
Human inductor, 3D trail detector, temperature-detecting device, mist detecting device；Described image collector is for acquiring doorway
The image stream information at place, collected result is sent to by wireless network in the memory of monitoring center and is stored；Institute
Infrared human body inductor is stated for incuding human body information, when detecting human body information, image acquisition device is opened by controller
Detector is trailed with 3D, human body information is can't detect then in a dormant state, reduces the waste of energy；The 3D trails detector
Whether it is more than regulation number for detecting the number in monitoring area, more than then sending out alarm；The temperature-detecting device and
Mist detecting device is used to detect the temperature and smokescope of environment, is more than preset range then by warning device warning reminding,
Prevent accident.
Described information processing center includes pretreatment unit, palmmprint enhancement unit, palmmprint extraction unit, palm print and palm vein identification
Machine；The pretreatment unit is used to carry out denoising, the palmmprint enhancing to the collected palmprint image of palmmprint acquisition module
Unit is used to carry out image enhancement effects processing to the palmprint image that denoising obtains；The palmmprint extraction unit enhances for extracting
Treated palmmprint textural characteristics；The palm print and palm vein cognitron is used for according to enhancing treated palmmprint textural characteristics clamshell doors
The identity of people is identified, and recognition result is correct, then controlling electronic lock by controller opens, and identification is incorrect, then passes through report
Alarm device warning reminding.
Beneficial effects of the present invention：Identification is carried out using personal recognition machine, identification accuracy is improved, is connect using non-
Touch operation, use more freely, safety, health, efficiently；The present invention also acquires the image information and environmental information on doorway, goes out
Now exception and alarm reduce the loss of resident family, improve safety；The configuration of the present invention is simple, at low cost, intelligence degree height,
Safety is good, is suitble to promote the use of.
Description of the drawings
Fig. 1 is the frame construction drawing of the present invention；
Fig. 2 is the principle of the present invention block diagram；
Fig. 3 is the structure chart of the identification device of the present invention.
Reference numeral：
Palmmprint acquisition module 1；Monitoring center 2；Information processing centre 3；Electronic lock 4；Information collecting center 5；Image Acquisition
Device 6；Infrared human body inductor 7；3D trails detector 8；Temperature-detecting device 9；Mist detecting device 10；Controller 11；Power supply
12；Warning device 13；Lighting device 14；Memory 15；Pretreatment unit 31；Palmmprint enhancement unit 32；Palmmprint extraction unit 33；
Palm print and palm vein cognitron 34.
Specific implementation mode
In conjunction with following application scenarios, the invention will be further described.
Referring to Fig. 1, a kind of family security door based on personal recognition, it is characterised in that：The family security door is equipped with
Palmmprint acquisition module 1, monitoring center 2 and information processing centre 3, the palm print collecting device 1 are used to acquire the palmmprint of enabling people
Image；Described information processing center 3 is used to be identified according to the identity of the palmprint image clamshell doors people of enabling people, and will identification
As a result it is transferred to monitoring center 2；The monitoring center 2 is used to control the opening and closing of electronic lock 4 according to recognition result.
Preferably, referring to Fig. 2, the monitoring center 2 includes controller 11 and the power supply being connect with the controller 11
12, warning device 13, lighting device 14 and memory 15；The controller 11 is microcontroller, the unlatching for controlling electronic lock
And closure, when identification result is correct, controller 11 controls electronic lock and opens；When identification result is incorrect, electronic lock
It is not turned on, while the warning device 13 carries out warning reminding；The power supply 12 is used to power to controller 11, ensures controller
11 are in normal operating conditions；The lighting device 14 is used to carry out automatic illuminating at night, when there are abnormal conditions or disconnected
When electric, automatic illuminating is carried out by built-in accumulator, it is flurried to prevent；The memory 15 is for storing in information processing
The recognition result of the heart 3.
Preferably, described information acquisition center 5 includes image acquisition device 6, infrared human body inductor 7,3D trailing detectors
8, temperature-detecting device 9, mist detecting device 10；Described image collector 6 is used to acquire the image stream information at doorway, will adopt
The memory 15 that the result collected is sent to monitoring center 2 by wireless network is stored and is monitored；The infrared human body sense
It answers device 7 for incuding human body information, when detecting human body information, image acquisition device 6 is opened by controller 11 and 3D is trailed
Detector 8 can't detect human body information then in a dormant state, reduce the waste of energy；The 3D trails detector 8 for examining
Survey whether the number in monitoring area is more than regulation number, more than then sending out alarm；The temperature-detecting device 7 and smog inspection
Temperature and smokescope that device 10 is used to detect environment are surveyed, more than preset range then by 13 warning reminding of warning device, is prevented
Only occur unexpected.
Preferably, referring to Fig. 3, described information processing center 3 includes pretreatment unit 31, palmmprint enhancement unit 32, palmmprint
Extraction unit 33, palm print and palm vein cognitron 34；The pretreatment unit 31 is used to carry out at denoising collected palmprint image
Reason, the palmmprint enhancement unit 32 are used to carry out image enhancement effects processing to the palmprint image that denoising obtains；The palmmprint carries
Take unit 33 for extract enhance treated palmmprint textural characteristics；The palm print and palm vein cognitron 34 according to enhancing for handling
The identification of palmmprint textural characteristics clamshell doors people afterwards, recognition result is correct, then controlling electronic lock by controller 10 opens,
It identifies incorrect, then passes through 13 warning reminding of warning device.
The above embodiment of the present invention is carried out identification using personal recognition machine, improves identification accuracy, connect using non-
Touch operation, use more freely, safety, health, efficiently；The present invention also acquires the image information and environmental information on doorway, goes out
Now exception and alarm reduce the loss of resident family, improve safety；The configuration of the present invention is simple, at low cost, intelligence degree height,
Safety is good, is suitble to promote the use of.
Preferably, the pretreatment unit 31 for chosen from palmprint image size be 256 × 256 it is interested
Region (ROI region), and denoising is carried out to ROI region, the ROI region after denoising is obtained, specially：
1 according to 1 collected palmprint image of palmmprint acquisition module, and one is chosen out of palmprint image and is believed containing detailed palmmprint
The ROI region of breath, if the area size is 256 × 256；
2) it uses Fourier transformation by ROI region by space field transformation to frequency domain, and utilizes attenuation function F (i, j)
Denoising is carried out to the ROI region in frequency domain, obtains the ROI region after denoising, attenuation function F (i, j) formula is：
Wherein, if (x, y) is pixel point coordinates in ROI region, (i, j) is pixel (x, y) in ROI region in the areas ROI
Respective coordinates in the frequency domain in domain；F (i, j) is attenuation function value, P0For cutoff frequency, P (i, j) is the frequency of ROI region
For coordinate origin in domain to the Euclidean distance of point (i, j), m is decay factor, is a positive integer more than 0；
Decay to the high fdrequency component in the ROI region Jing Guo Fourier transformation, it can be with by attenuation function F (i, j)
Make to be less than cutoff frequency P0All frequencies all pass through, and be higher than cutoff frequency P0Frequency according to the Euclidean distance of point (i, j)
With cutoff frequency P0Ratio gradually decayed；
3) inverse transformation is carried out to the ROI region after denoising using inverse Fourier transform, by the areas denoising Hou ROI
Domain is by frequency domain transformation to spatial domain, obtaining the ROI region after denoising.
This preferred embodiment, selection carry out denoising, energy using attenuation function F (i, j) in frequency domain to ROI region
The random noise in ROI region is enough effectively removed, retains the palmmprint texture information in palmprint image, while recycling attenuation function
When F (i, j) decays to high fdrequency component, a smooth intermediate zone can be formed in transitional region so that the figure after denoising
As being not in ringing.
Preferably, the palmmprint enhancement unit 32 is used to transform to the ROI region after denoising in fuzzy field, and to fuzzy
ROI region in domain carries out logarithmic transformation and fuzzy inverse transformation, obtains enhanced ROI region, specially：
1) by the ROI region after denoising by space field transformation to fuzzy field, used transforming function transformation function is：
Wherein, Ua,bFor the transforming function transformation function value at the ROI region pixel (a, b) after denoising, Za,bFor the areas denoising Hou ROI
Gray value at domain pixel (a, b), ZminFor the minimum gradation value of the ROI image after denoising, ZmaxFor the ROI image after denoising
Maximum gradation value；
It seeks all pixels point in the ROI region after denoising all over, all pixels point in the ROI region after denoising can be obtained
Transforming function transformation function value；
2) use non-linear transform function formula to Ua,bNonlinear Processing is carried out, it is flat that a new fuzzy characteristics can be obtained
Face U ', the non-linear transform function formula defined are：
Ua,b'=Br(Ua,b)=Br(Br-1(Ua,b)), r=1,2,3 ...
Wherein, Ua,b' is the fuzzy ownership angle value after nonlinear transformation, ZcFor customized fuzzy degree of membership threshold value；
Br(Ua,b) it is a mapping function；
3) to Ua,b' carries out nonlinear inversion transformation, and the ROI image after the denoising in fuzzy field is transformed to sky by fuzzy field
Between domain, inverse transformation formula is：
Wherein, Za,b' is the functional value of the inverse transformation obtained after inverse transformation；
Traverse all pixels point in fuzzy field, all Za,bThe collection of ' compositions is combined into enhancing treated ROI region.
This preferred embodiment carries out conversion process to the ROI image after denoising in fuzzy field using transforming function transformation function, passes through
Increase the angle value that is subordinate to more than fuzzy membership threshold value, while reducing and being subordinate to angle value, Jin Ershi less than fuzzy membership threshold value
Now the enhancing of the ROI image after denoising is handled, image enhancement can not only be achieved the purpose that by this algorithm, and remain
The brightness of image itself also can be good at inhibiting the random noise in image.
Preferably, the palmmprint extraction unit 33 is used to extract the textural characteristics by enhancing treated ROI region, obtains
To enhancing treated palmmprint textural characteristics, wherein the size of enhancing treated ROI region is 256 × 256, specially：
1) convolutional neural networks model is used to carry out texture information extraction to enhancing treated ROI region；Implementation method
It is：Will enhancing treated ROI region as input picture, use size for 11 × 11 convolution filter to input picture into
Row convolution operation, obtains the characteristic pattern that 96 sizes are 55 × 55, and transformation and normalized are carried out to the data that convolution obtains,
It normalizes formula：
Wherein,Indicate that g-th of convolution filter passes through application volume at enhancing treated ROI region pixel (s, t)
The calculated neuronal activation degree of product core, n are in the close convolution nuclear mapping number of same spatial location, and N is that convolution kernel is total
Number, K, α, β are preset value,For the value obtained after normalized；
Down-sampling is carried out to the data after normalized, wherein down-sampling window is 3 × 3, and step-length 2 has obtained 96
The characteristic pattern that a size is 27 × 27；Use the spy that 256 sizes are 27 × 27 for 5 × 5 × 48 96 sizes of filter pair
Sign figure carries out convolution operation, obtains 384 13 × 13 characteristic patterns；Use 256 sizes for 3 × 3 × 192 filter pair
384 13 × 13 characteristic patterns carry out convolution operation, 256 13 × 13 characteristic patterns are obtained, by obtain 256 13 × 13 features
Figure carries out down-sampling operation, and wherein down-sampling window size is 3 × 3, and step-length 2 obtains the feature that 256 sizes are 6 × 6
Figure；The pixel for 256 6 × 6 characteristic patterns that down-sampling obtains is formed a line, dimensionality reduction operation is carried out using neural network, obtains
To the output result of 4096 dimensions；After result after dimension-reduction treatment is input to full Connection Neural Network, output is at enhancing
4096 dimension palmmprint textural characteristics of the ROI region after reason；
2) dimensionality reductions are carried out to 4096 dimension palmmprint textural characteristics of enhancing treated ROI region using Principal Component Analysis Algorithm
Operation obtains enhancing treated palmmprint textural characteristics.
This preferred embodiment, when carrying out convolution operation to enhancing treated ROI region, selection to neuronal activation degree into
Row normalized has and utilizes the extensive operation realized to model, the error rate for the convolutional layer that degraded；Moreover carrying out convolution
When operation, the pond algorithm of selection overlapping prevents over-fitting, uses full Connection Neural Network and handles enhancing
ROI region afterwards extracts palmmprint textural characteristics, and the palmmprint which can be good at showing in palmprint image is thin
Information is saved, is conducive to the identification of follow-up clamshell doors personnel identity, improves the safety of antitheft door.
Preferably, treated slaps for the enhancing that the palm print and palm vein cognitron 34 is used to extract according to palmmprint extraction unit 33
The identity of line textural characteristics clamshell doors people is identified, and identification result is transferred to monitoring center 2.
Finally it should be noted that the above embodiments are merely illustrative of the technical solutions of the present invention, rather than the present invention is protected
The limitation of range is protected, although being explained in detail to the present invention with reference to preferred embodiment, those skilled in the art should
Understand, technical scheme of the present invention can be modified or replaced equivalently, without departing from the essence of technical solution of the present invention
And range.
Claims (3)
1. a kind of family security door based on personal recognition, it is characterised in that：The family security door is equipped with palmmprint and acquires mould
Block, information processing centre and monitoring center；
The palm print collecting device is used to acquire the palmprint image of enabling people；
Described information processing center is used to be identified according to the identity of the palmprint image clamshell doors people of enabling people, and identity is known
Other result is transferred to monitoring center；
The monitoring center is used for the opening and closing according to identification output control electronic lock；
The monitoring center includes controller and the power supply being connect with the controller, warning device, lighting device and storage
Device；
The controller is microcontroller, the unlatching for controlling electronic lock and closure, when identification result is correct, controller control
Electronic lock processed is opened；When identification result is incorrect, electronic lock is not turned on, while the warning device carries out warning reminding；
The power supply is used to power to controller, ensures that controller is in normal operating conditions；
The lighting device is used to carry out automatic illuminating at night, when there are abnormal conditions or power-off, passes through built-in storage
Battery carries out automatic illuminating, and it is flurried to prevent；
The memory is used to store the identification result of information processing centre；
The family security door further includes information collecting center, and described information is acquired centrally through wireless network and monitoring center phase
Even；
Described information acquisition center include image acquisition device, infrared human body inductor, 3D trail detector, temperature-detecting device,
Mist detecting device；Described image collector is used to acquire the image stream information at doorway, collected result is passed through wireless
Network is sent in the memory of monitoring center and is stored；The infrared human body inductor works as inspection for incuding human body information
When measuring human body information, image acquisition device is opened by controller and 3D trails detector, human body information is can't detect and is then in
Dormant state reduces the waste of energy；Whether it is more than rule that the 3D trails detector for detecting the number in monitoring area
Number is determined, more than then sending out alarm；The temperature-detecting device and mist detecting device are used to detect the temperature and smog of environment
Concentration is more than preset range then by warning device warning reminding, prevents accident；
Described information processing center includes pretreatment unit, palmmprint enhancement unit, palmmprint extraction unit, palm print and palm vein cognitron；
The pretreatment unit is used to carry out denoising, the palmmprint enhancement unit to the collected palmprint image of palmmprint acquisition module
Palmprint image for being obtained to denoising carries out image enhancement processing；Treated for extracting enhancing for the palmmprint extraction unit
Palmmprint textural characteristics；The palm print and palm vein cognitron is used for the identity according to enhancing treated palmmprint textural characteristics clamshell doors people
It is identified, and identification result is transferred to monitoring center；
The pretreatment unit from palmprint image for choosing a size for 256 × 256 ROI region, and to ROI region
Denoising is carried out, obtains the ROI region after denoising, specially：
1) according to the collected palmprint image of palmmprint acquisition module, one is chosen out of palmprint image containing detailed palm print information
ROI region, if the area size is 256 × 256；
2) use Fourier transformation by ROI region by space field transformation to frequency domain, and using attenuation function F (i, j) to frequency
ROI region in rate domain carries out denoising, obtains the ROI region after denoising, attenuation function F (i, j) formula is：
Wherein, if (x, y) is pixel point coordinates in ROI region, (i, j) is pixel (x, y) in ROI region in ROI region
Respective coordinates in frequency domain；F (i, j) is attenuation function value, P0For cutoff frequency, P (i, j) is in the frequency domain of ROI region
Coordinate origin to the Euclidean distance of point (i, j), m is decay factor, be one be more than 0 positive integer；
Decay to the high fdrequency component in the ROI region Jing Guo Fourier transformation, can be made by attenuation function F (i, j) low
In cutoff frequency P0All frequencies all pass through, and be higher than cutoff frequency P0Frequency according to the Euclidean distance of point (i, j) and cut
Only frequency P0Ratio gradually decayed；
3) using inverse Fourier transform to after denoising ROI region carry out inverse transformation, by the ROI region after denoising by
In frequency domain transformation to spatial domain, the ROI region after denoising is obtained.
2. family security door according to claim 1, it is characterised in that：The palmmprint enhancement unit is used for will be after denoising
ROI region transforms in fuzzy field, and carries out logarithmic transformation and fuzzy inverse transformation to the ROI region in fuzzy field, is enhanced
ROI region afterwards, specially：
1) by the ROI region after denoising by space field transformation to fuzzy field, used logarithmic transformation function is：
Wherein, UA, bFor the transforming function transformation function value at the ROI region pixel (a, b) after denoising, ZA, bFor the ROI region picture after denoising
Gray value at vegetarian refreshments (a, b), ZminFor the minimum gradation value of the ROI image after denoising, ZmaxMost for the ROI image after denoising
High-gray level value；
It seeks all pixels point in the ROI region after denoising all over, the transformation of all pixels point in the ROI region after denoising can be obtained
Functional value；
2) use non-linear transform function formula to UA, bNonlinear Processing is carried out, a new Fuzzy property domain U ' can be obtained,
Its define non-linear transform function formula be：
UA, b'=Br(UA, b)=Br(Br-1(UA, b)), r=1,2,3...
Wherein, UA, b' for fuzzy ownership angle value after nonlinear transformation, ZcFor customized fuzzy degree of membership threshold value；Br
(UA, b) it is a mapping function；
3) to UA, b' nonlinear inversion transformation is carried out, the ROI image after the denoising in fuzzy field is transformed into spatial domain by fuzzy field,
Its inverse transformation formula is：
Wherein, ZA, b' it is the inverse transform function value obtained after inverse transformation；
Traverse all pixels point in fuzzy field, all ZA, bThe collection of ' composition is combined into enhancing treated ROI region.
3. family security door according to claim 2, it is characterised in that：The palmmprint extraction unit is for extracting by increasing
The textural characteristics of strong treated ROI region obtain enhancing treated palmmprint textural characteristics, wherein enhancing treated ROI
The size in region is 256 × 256, specially：
1) convolutional neural networks model is used to carry out texture information extraction to enhancing treated ROI region；Implementation method is：It will
Enhancing treated ROI region uses size to be rolled up to input picture for 11 × 11 convolution filter as input picture
Product operation, obtains the characteristic pattern that 96 sizes are 55 × 55, carries out transformation and normalized to the data that convolution obtains, returns
One, which changes formula, is：
Wherein,Indicate that g-th of convolution filter is by applying convolution kernel at enhancing treated ROI region pixel (s, t)
Calculated neuronal activation degree, n are in the close convolution nuclear mapping number of same spatial location, and N is convolution kernel total number,
K, α, β are preset value,For the value obtained after normalized；
Down-sampling is carried out to the data after normalized, wherein down-sampling window is 3 × 3, step-length 2, has been obtained 96 big
The small characteristic pattern for being 27 × 27；Use the characteristic pattern that 256 sizes are 27 × 27 for 5 × 5 × 48 96 sizes of filter pair
Convolution operation is carried out, 384 13 × 13 characteristic patterns are obtained；Use 256 sizes for 3 × 3 × 192 384, filter pair
13 × 13 characteristic pattern carries out convolution operation, obtains 256 13 × 13 characteristic patterns, by obtain 256 13 × 13 characteristic patterns into
Row down-sampling operates, and wherein down-sampling window size is 3 × 3, step-length 2, obtains the characteristic pattern that 256 sizes are 6 × 6；It will
The pixel for 256 6 × 6 characteristic patterns that down-sampling obtains forms a line, and carries out dimensionality reduction operation using neural network, obtains 4096
The output result of dimension；After result after dimension-reduction treatment is input to full Connection Neural Network, output is that treated for enhancing
4096 dimension palmmprint textural characteristics of ROI region；
2) dimensionality reduction behaviour are carried out to 4096 dimension palmmprint textural characteristics of enhancing treated ROI region using Principal Component Analysis Algorithm
Make, obtains enhancing treated palmmprint textural characteristics.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710900903.9A CN107724900B (en) | 2017-09-28 | 2017-09-28 | A kind of family security door based on personal recognition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710900903.9A CN107724900B (en) | 2017-09-28 | 2017-09-28 | A kind of family security door based on personal recognition |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107724900A CN107724900A (en) | 2018-02-23 |
| CN107724900B true CN107724900B (en) | 2018-10-23 |
Family
ID=61208764
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710900903.9A Active CN107724900B (en) | 2017-09-28 | 2017-09-28 | A kind of family security door based on personal recognition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107724900B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110321871B (en) * | 2019-07-11 | 2023-04-07 | 电子科技大学成都学院 | Palm vein identification system and method based on LSTM |
| CN112700576B (en) * | 2020-12-29 | 2021-08-03 | 成都启源西普科技有限公司 | Multi-modal recognition algorithm based on images and characters |
| CN114241633B (en) * | 2021-12-02 | 2022-11-08 | 泰仕达电子（深圳）有限公司 | Remote control system of intelligent door lock |
| CN114973467B (en) * | 2022-07-07 | 2022-10-18 | 深圳市经纬纵横科技有限公司 | Household intelligent door lock control system based on communication platform |
Family Cites Families (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN201225071Y (en) * | 2008-07-01 | 2009-04-22 | 黄耀荣 | Anti-theft door |
| US8350680B2 (en) * | 2009-05-27 | 2013-01-08 | Cole Stephen F | System for maintaining electrically operated folding partition systems |
| CN102254188B (en) * | 2011-08-04 | 2013-03-13 | 汉王科技股份有限公司 | Palmprint recognizing method and device |
| CN102797390B (en) * | 2012-08-27 | 2015-04-22 | 方永秋 | Intelligent burglarproof door |
| CN203640408U (en) * | 2013-11-21 | 2014-06-11 | 云南启顺科技有限公司 | Multifunctional safe-guard door lock |
| CN104268963A (en) * | 2014-08-06 | 2015-01-07 | 成都百威讯科技有限责任公司 | Intelligent door lock system, intelligent door lock and intelligent alarm door |
| CN105184216B (en) * | 2015-07-24 | 2019-01-11 | 山东大学 | A kind of digital extraction method of two area's palmmprint of the heart |
| CN105426821B (en) * | 2015-11-04 | 2018-09-07 | 浙江工业大学 | A kind of palm vein feature extraction and matching method based on eight neighborhood and Secondary Match |
| CN205354169U (en) * | 2016-02-25 | 2016-06-29 | 曾丽琴 | Domestic entrance guard's device |
| CN106355825A (en) * | 2016-11-17 | 2017-01-25 | 天津稻恩科技有限公司 | Security and protection system based on palm print and palm pulse recognition |
-
2017
- 2017-09-28 CN CN201710900903.9A patent/CN107724900B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN107724900A (en) | 2018-02-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107724900B (en) | A kind of family security door based on personal recognition | |
| CN102163283B (en) | Method for extracting face characteristic based on local three-value mode | |
| Gong et al. | A real-time fire detection method from video with multifeature fusion | |
| Tzeng et al. | Design of fall detection system with floor pressure and infrared image | |
| CN103093234B (en) | Based on the personal identification method of ground reaction force during walking | |
| CN106803301A (en) | A kind of recognition of face guard method and system based on deep learning | |
| CN203250345U (en) | Access control system based on brain wave identity recognition | |
| CN105139003B (en) | A kind of dynamic human face recognition system and method | |
| CN107392112A (en) | A kind of facial expression recognizing method and its intelligent lock system of application | |
| CN108053615A (en) | Driver tired driving condition detection method based on micro- expression | |
| CN106250821A (en) | The face identification method that a kind of cluster is classified again | |
| CN202196458U (en) | Face recognition system | |
| CN104239851A (en) | Intelligent cell inspection system based on behavior analysis and control method thereof | |
| CN108492404A (en) | A kind of the face lock method for unlocking and device of additional Expression Recognition | |
| CN110149604B (en) | Personnel detection method based on WIFI signal | |
| CN108205869A (en) | Intelligent household security system | |
| CN108009488A (en) | The street security joint defense system of Behavior-based control analysis | |
| CN109493470A (en) | A kind of intelligent access control system based on block chain | |
| CN109389793A (en) | A kind of super anti-theft monitoring system of quotient | |
| CN104809780A (en) | Access control method based on fingerprint recognition and face recognition | |
| CN108305366A (en) | A kind of intelligent access control system with face identification functions | |
| CN108537921A (en) | Visitor's recognition methods based on recognition of face and system | |
| CN209433456U (en) | A kind of face identification system for computer room | |
| CN204331938U (en) | A kind of intelligent safety and defence system for burglary | |
| CN116631068B (en) | Palm vein living body detection method based on deep learning feature fusion |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20180904 Address after: 221300 the east side of Pei Cang Road, Xinhua Village, Guan Hu Town, Pizhou City, Xuzhou, Jiangsu (50 meters south of the town government) Applicant after: Paragon products (Jiangsu) Co., Ltd. Address before: 518000 B356, 2124 Li Wan mansion, Nanhai street, Nanshan District, Shenzhen, Guangdong. Applicant before: Shenzhen Shengda Machine Design Co., Ltd. |
|
| TA01 | Transfer of patent application right | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |