CN104680137B - Embedded monocular iris identification device, system and method - Google Patents
Embedded monocular iris identification device, system and method Download PDFInfo
- Publication number
- CN104680137B CN104680137B CN201510065270.5A CN201510065270A CN104680137B CN 104680137 B CN104680137 B CN 104680137B CN 201510065270 A CN201510065270 A CN 201510065270A CN 104680137 B CN104680137 B CN 104680137B
- Authority
- CN
- China
- Prior art keywords
- image
- cpu
- iris
- built
- storage
- 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
Abstract
The invention discloses a kind of embedded monocular iris identification devices, system and method, belong to field of biological recognition, described device includes CPU and camera lens, the CPU is connected with image capture module, image processing module, built-in storage and external memory, and the CPU is also connected with serial ports and/or USB interface, wherein the data input pin of described image acquisition module connects the camera lens, and data output end connects described image processing module；Described image processing module also connects the built-in storage.The present invention has many advantages, such as that speed is fast, stability is high, flexibility is good, transplantability is strong.
Description
Technical field
The present invention relates to field of biological recognition, a kind of embedded monocular iris identification device, system and its side are particularly related to
Method.
Background technique
It is generally used non-embedded mode for current monocular iris identification device to design, its composition includes single
(so-called host computer just refers to PC machine or other embedded systems, SCM system for mesh iris image acquiring equipment and host computer
Deng the system with certain " control, operation " ability).It acquires monocular at work, by monocular iris image acquiring equipment
Acquired image information, is then sent to host computer by iris image, and host computer will carry out phase to corresponding image information
The operation answered, including extracting iris feature, saving iris feature, compare (by the iris feature extracted from iris image and upper
Stored iris feature compares one by one in the machine database of position), identification (by the comparison to iris feature, confirm identification person's
Identity) etc. operation.
For non-embedded monocular iris identification device, its monocular iris image acquiring equipment only realizes image
Acquisition function, and the substantive work of monocular iris recognition, as the extraction of iris feature, preservation, comparison and identification are all by upper
Position machine is completed.This non-embedded monocular iris identification device is there are many limitations, and first, function is relatively simple, greatly
Limit the application environment of equipment；Second, so that upper computer software realizes that complicated, difficulty is big, third is this non-embedded
Monocular iris identification device is difficult to be integrated into other systems.
Summary of the invention
It is embedding that the technical problem to be solved in the present invention is to provide a kind of speed is fast, stability is high, flexibility is good, transplantability is strong
Enter formula monocular iris identification device, system and method.
In order to solve the above technical problems, present invention offer technical solution is as follows:
A kind of embedded monocular iris identification device, including CPU and camera lens, the CPU are connected with Image Acquisition mould
Block, image processing module, built-in storage and external memory, and the CPU is also connected with serial ports and/or USB interface, in which:
The data input pin of described image acquisition module connects the camera lens, and data output end connects described image and handles mould
Block；
Described image processing module also connects the built-in storage.
A kind of embedded monocular iris authentication system, including host computer and above-mentioned embedded monocular iris identification device.
The method for the iris feature registration that above-mentioned embedded monocular iris identification device carries out, comprising:
Step 1: powering on, CPU reads the user data stored in external memory in built-in storage, wherein the use
User data includes the ID and right and left eyes mark of user；
Step 2:CPU receives register instruction from host computer, includes ID and the left side of user to be registered in the register instruction
Right eye mark；
Step 3:CPU judges whether the eye of the ID number has been registered, if not, executing step 4；
Step 4: image capture module obtains image from camera lens, and is sent to the pretreatment of image processing module row, pretreatment
Image afterwards is stored in built-in storage, and CPU detects whether pretreated image in built-in storage is iris image, if
It is to execute step, 5；
Step 5:CPU judges whether pretreated image meets image quality requirements, if so, executing step 6；
Step 6:CPU is to pretreated image zooming-out iris feature value and saves；
Step 7:CPU feeds back response result to host computer.
The method that above-mentioned embedded monocular iris identification device carries out the certification of 1:1 iris feature, comprising:
Step 1 ': it powers on, CPU reads the user data stored in external memory in built-in storage, wherein the use
User data includes the ID and iris feature value of user；
Step 2 ': CPU receives 1:1 certification instruction from host computer, includes the ID of user to be certified in the certification instruction；
Step 3 ': CPU judges whether the ID number has been registered according to the user data in built-in storage, if so, executing
Step 4 '；
Step 4 ': image capture module obtains image from camera lens, and is sent to image processing module and is pre-processed, pre- to locate
Image after reason is stored in built-in storage, and CPU detects whether pretreated image in built-in storage is iris image, if
It is to execute step 5 '；
Step 5 ': CPU judges whether pretreated image meets image quality requirements, if so, executing step 6 '；
Step 6 ': CPU is to pretreated image zooming-out iris feature value, and the iris feature value and memory that will be extracted
The eye iris characteristic value of the user is compared in reservoir；
Step 7 ': CPU feeds back response result to host computer.
The method that above-mentioned embedded monocular iris identification device carries out the certification of 1:N iris feature characterized by comprising
Step 1 ": powering on, and CPU reads the user data stored in external memory in built-in storage, wherein the use
User data includes the ID and iris feature value of user；
Step 2 ": CPU receives 1:N certification instruction from host computer；
Step 3 ": CPU judges whether that existing subscriber registers, if so, executing step according to the user data in built-in storage
Rapid 4 "；
Step 4 ": image capture module obtains image from camera lens, and is sent to image processing module and is pre-processed, pre- to locate
Image after reason is stored in built-in storage, and CPU detects whether pretreated image in built-in storage is iris image, if
It is to execute step 5 "；
Step 5 ": CPU judges whether pretreated image meets image quality requirements, if so, executing step 6 "；
Step 6 ": CPU is to pretreated image zooming-out iris feature value, and the iris feature value and memory that will be extracted
The iris feature value of user compares one by one in reservoir；
Step 7 ": CPU feeds back response result to host computer.
The invention has the following advantages:
Compared with prior art, embedded monocular iris identification device of the invention is not needing host computer or other external
In the case that equipment is interfered, it will be able to the acquisition and processing for completing image simplify the software realization of host computer in the prior art,
Improve the stability of embedded monocular iris identification device.Setting of the present invention to built-in storage and external memory, realizes
Caching and storage to data, make CPU in data processing, in advance by the data copy in external memory into built-in storage,
CPU can call in time the data in built-in storage to be handled, and effectively increase the data processing speed of CPU.In the present invention
Serial ports and USB interface are additionally provided, can be applied simultaneously when application, can also be used alone, provide selection for client
Leeway, while facilitating the secondary development of user, also widened the application environment of equipment, flexibility and transplantability are reinforced.
Detailed description of the invention
Fig. 1 is the structural schematic diagram of embedded monocular iris identification device of the invention；
Fig. 2 is the flow diagram for the method that iris feature of the invention is registered；
Fig. 3 is the flow diagram for the method that 1:1 iris feature of the invention authenticates；
Fig. 4 is the flow diagram for the method that 1:N iris feature of the invention authenticates；
Fig. 5 is the flow diagram of step 4,4 ' or 4 " in iris feature registration/certification method of the invention；
Fig. 6 is a kind of flow diagram of the embodiment for the method that iris feature of the invention is registered；
Fig. 7 is a kind of flow diagram of the embodiment for the method that 1:1 iris feature of the invention authenticates；
Fig. 8 is a kind of flow diagram of the embodiment for the method that 1:N iris feature of the invention authenticates.
Specific embodiment
To keep the technical problem to be solved in the present invention, technical solution and advantage clearer, below in conjunction with attached drawing and tool
Body embodiment is described in detail.
On the one hand, the present invention provides a kind of embedded monocular iris identification device, as shown in Figure 1, including CPU101 and one
A camera lens 102, CPU101 are connected with image capture module 104, image processing module 103, built-in storage 105 and external memory
106, and CPU101 is also connected with serial ports 107 and/or usb 1 08；The data input pin of image capture module 104 connects
Camera lens 102, data output end connect image processing module 103；Image processing module 103 also connects built-in storage 105.
Compared with prior art, the present invention is not in the case where needing host computer or other external devices are interfered, it will be able to
The acquisition and processing for completing image, effectively simplify the software realization of host computer in the prior art.It is directed to the prior art simultaneously
Middle Image Acquisition uses speed caused by the software collection mode of CPLD slow and the disadvantage of stability difference, the present invention use hardware system
System realizes Image Acquisition, and speed is fast, stability is high.Serial ports 107 and usb 1 08 are additionally provided in the present invention, application when
Time can be applied simultaneously, can also be used alone, provide the leeway of selection for client, facilitate the secondary development of user, also widen
The application environment of equipment, can readily be integrated into other embedded systems, SCM system, dsp system, CPLD or
FPGA system etc. has in the system of certain processing operational capability, keeps embedded monocular iris identification device of the invention flexible
Property is good, transplantability is strong.
In the present invention, CPU101, image processing module 103, image capture module 104 both can be three independent cores
Piece can also not influence the reality of technical solution of the present invention using the chip of three's unification or the chip of two of them unification
It is existing.It when circuit design, can design on a pcb board, be preferably designed to two pcb boards, one of plate is for connecting mirror
Head mould group, another plate is for connecting the chip modules such as CPU101, image capture module 104.The present invention can select integrated level
High chip allows the area of pcb board to be made small.
As an improvement of the present invention, Image Acquisition, while image are realized using cmos image sensor in the present invention
Acquisition module 104 passes through CMOS interface connection lens 102.Cmos image sensor is compared to commonly used in the prior art other one
For kind mainstream CCD image sensor, have many advantages, such as that low in energy consumption, speed is fast, integrated level is high.
Image Acquisition condition is carried out in extraneous illumination deficiency in order to meet, and can be equipped with light filling at camera lens 102
Lamp；In order to facilitate prompt/prompting user, CPU101 can also be connected with LED light (or dual-colored LED indicator light) and/or
Buzzer；Built-in storage 105 is preferably DDR2, and external memory 106 can be FLASH, preferably NAND FLASH, serial ports 107
Preferably RS232 interface.
Preferably, CPU101 can select the ARM chip of Cortex-A8 framework；DDR2 memory chip is compatible with 128MB/
256MB/512MB/1GB is even more big, is configured according to customer demand；Flash memory chip is compatible with 128MB/256MB/512MB/
1GB is even more big, is configured according to customer demand；Imaging sensor can use the CMOS of the even more big pixel of 30W/200W；It mends
Light lamp can use 850nm infrared light compensating lamp；Whole system is powered by UART or usb 1 08, is 5V/500mA, figure
As the connection of FPC soft arranging wire can be used between acquisition module 104 and camera lens 102.
On the other hand, the present invention provides a kind of embedded monocular iris authentication system, including host computer and above-mentioned insertion
Formula monocular iris identification device.Structure with it is upper identical, details are not described herein again.
Host computer in the present invention can be sent to embedded monocular iris identification device when system works register instruction,
The necessary instruction of certification instruction and other devices in the application, starts the working procedure of embedded monocular iris identification device, embedding
After the completion of entering the work of formula monocular iris identification device, result can be uploaded to host computer and be shown, intuitively show use
Family facilitates human-computer interaction.
In another aspect, the present invention provides a kind of side for carrying out iris feature registration using embedded monocular iris identification device
Method, as shown in Figure 2, comprising:
Step 1: it powers on, CPU101 reads the user data stored in external memory 106 in built-in storage 105,
In, the user data includes the ID and right and left eyes mark of user；
In this step, the ID number of user data is to show the number of user identity, has uniqueness, it can be individually
Number, individually letter, and the forms such as number and alphabetical combination.Right and left eyes are indicated for distinguishing right and left eyes, can be used
Various expression-forms.
Step 2:CPU101 receives register instruction from host computer, include in the register instruction user to be registered ID and
Right and left eyes mark；
In this step, embedded monocular iris identification device receives the register instruction of host computer, and then starts accreditation process.
Step 3:CPU101 judges whether the eye of the ID number has been registered, if not, executing step 4；
It is whether registered to user to judge in this step, avoid repeated registration.It, can be with if having registered
Refusal is registered and jumps directly to step 7 feedback host computer.
Step 4: image capture module 104 obtains image from camera lens 102, and is sent to 103 row of image processing module and locates in advance
Reason, pretreated image are stored in built-in storage 105, and pretreated image is in CPU101 detection built-in storage 105
No is iris image, if so, executing step 105；
In this step, image processing module 103 pre-processes image, so that the image be made to meet subsequent processing
Requirement；Meanwhile pretreated image being stored in built-in storage 105, it reads and uses for CPU101；Finally,
CPU101 also detects whether to be iris image in advance, to avoid the processing capacity of the non-iris image waste CPU101 of acquisition.Such as
If fruit is not iris image, can go to step 4 resurveys image again.
Step 5:CPU101 judges whether pretreated image meets image quality requirements, if so, executing step 6；
In this step, the reliable extraction for subsequent iris feature value provides guarantee.If being unsatisfactory for quality requirement,
It can go to step and 4 resurvey image again.
Step 6:CPU101 is to pretreated image zooming-out iris feature value and saves；
In this step, ID and the right and left eyes mark of iris feature value and user, which are collectively stored in external memory, (to be realized
Iris registration), foundation is provided for later iris authentication process.
Step 7:CPU101 feeds back response result to host computer.
The present invention realizes the extraction and preservation and user's registration identification process of iris feature, eliminates the prior art
The process of middle extraction and the preservation and user's registration identification that iris feature is carried out by host computer, this invention simplifies upper
The development difficulty of machine software, improves the stability of device operation, while also strengthening flexibility and the transplantability of device application.
Pretreated image is stored in built-in storage 105 in the present invention, it, can be timely when CPU101 being facilitated to carry out data processing
It calls, improves the data processing speed of CPU101.
Another aspect carries out 1:1 iris feature the present invention also provides a kind of embedded monocular iris identification device of application and recognizes
The method of card, as shown in Figure 3, comprising:
Step 1 ': it powers on, CPU101 reads the user data stored in external memory 106 in built-in storage 105,
In, the user data includes the ID of user, right and left eyes mark and iris feature value；
In this step, the ID number of user data is to show the number of user identity, has uniqueness, it can be individually
Number, individually letter, and the forms such as number and alphabetical combination.Right and left eyes are indicated for distinguishing right and left eyes, can be used
Various expression-forms.Iris feature value provides alignment parameters for subsequent certification.
Step 2 ': CPU101 receives 1:1 certification instruction from host computer, includes user to be certified in the certification instruction
ID and right and left eyes mark；
In this step, embedded monocular iris identification device receives the certification instruction of host computer, and then starts authentication procedure.
Step 3 ': CPU1 judges whether the ID number has been registered according to the user data in built-in storage 105, if so,
Execute step 4 '；
In this step, the phenomenon that user not registered is directly entered certification waste of resource is avoided.If do not registered
If, step 7 can be jumped directly to ' feedback host computer.
Step 4 ': image capture module 104 from camera lens 102 obtain image, and be sent to image processing module 103 carry out it is pre-
Processing, pretreated image is stored in built-in storage, in CPU101 detection built-in storage pretreated image whether be
Iris image, if so, executing step 5 '；
In this step, image processing module 103 pre-processes image, so that the image be made to meet subsequent processing
Requirement；Meanwhile pretreated image being stored in built-in storage 105, it reads and uses for CPU101；Finally,
CPU101 also detects whether to be iris image in advance, to avoid the processing capacity of the non-iris image waste CPU101 of acquisition.Such as
If fruit is not iris image, can go to step 4 ' resurveys image again.
Step 5 ': CPU101 judges whether pretreated image meets image quality requirements, if so, executing step
6'；
In this step, the reliable extraction for subsequent iris feature value provides guarantee.If being unsatisfactory for quality requirement,
It can go to step and 4 ' resurvey image again.
Step 6 ': CPU101 to pretreated image zooming-out iris feature value, and by the iris feature value extracted with
The eye iris characteristic value of the user is compared in built-in storage 105；
Step 7 ': CPU101 feeds back response result to host computer.
It is index with ID number in the method for above-mentioned 1:1 iris feature certification, ID use is found in built-in storage 105
Family also just has found the iris feature value of the user, is then compared with the iris feature value of extraction, returns to host computer and answer
Answer result.Pretreated image is stored in built-in storage 105 in the present invention, when CPU101 being facilitated to carry out data processing,
It can call in time, improve the data processing speed of CPU101.The method achieve inside embedded monocular iris device
Iris feature extraction, the preservation of iris feature, comparison and identification etc. a series of non-embedded monocular iris identification devices are upper
The function that position machine is realized, has widened the application environment of device, has simplified the development difficulty of upper computer software, improves device operation
Stability, while also strengthen device application flexibility and transplantability.
Another aspect carries out 1:N iris feature the present invention also provides a kind of embedded monocular iris identification device of application and recognizes
The method of card, as shown in Figure 4, comprising:
Step 1 ": powering on, and CPU101 reads the user data stored in external memory 106 in built-in storage 105,
In, the user data includes the ID and iris feature value of user；
In the present invention, in this step, the ID number of user data is to show the number of user identity, has uniqueness, it can
To be individual number, individually letter, and the forms such as number and alphabetical combination.Right and left eyes mark is for distinguishing left and right
Eye, can use various expression-forms.Iris feature value provides alignment parameters for subsequent certification.
Step 2 ": CPU101 receives 1:N certification instruction from host computer；
In the present invention, embedded monocular iris identification device receives the certification instruction of host computer, and then starts authentication procedure.
Step 3 ": CPU101 judges whether that existing subscriber registers according to the user data in built-in storage 105, if so,
Execute step 4 "；
In the present invention, the phenomenon that user not registered is directly entered certification waste of resource is avoided.If do not registered
If, step 7 can be jumped directly to " feedback host computer.
Step 4 ": image capture module 104 from camera lens 102 obtain image, and be sent to image processing module 103 carry out it is pre-
Processing, pretreated image are stored in built-in storage 105, and CPU101 detects pretreated image in built-in storage 105
It whether is iris image, if so, executing step 5 "；
In the present invention, image processing module 103 pre-processes image, so that the image be made to meet subsequent processing
Requirement；Meanwhile pretreated image being stored in built-in storage 105, it reads and uses for CPU101；Finally,
CPU101 also detects whether to be iris image in advance, to avoid the processing capacity of the non-iris image waste CPU101 of acquisition.Such as
If fruit is not iris image, can go to step 4 " resurveys image again.
Step 5 ": CPU101 judges whether pretreated image meets image quality requirements, if so, executing step
6"；
In the present invention, the reliable extraction for subsequent iris feature value provides guarantee.If being unsatisfactory for quality requirement,
It can go to step and 4 " resurvey image again.
Step 6 ": CPU101 to pretreated image zooming-out iris feature value, and by the iris feature value extracted with
The iris feature value of user compares one by one in built-in storage 105；
Step 7 ": CPU feeds back response result to host computer.
In the method for above-mentioned 1:N iris feature certification, it will extract and deposited in client iris characteristic value and built-in storage 105
The client iris characteristic value of storage compares one by one, returns to host computer response result, includes to compare through user's in response result
ID number.Pretreated image is stored in built-in storage 105 in the present invention, when CPU101 being facilitated to carry out data processing, energy
It is enough to call in time, improve the data processing speed of CPU101.The method achieve the rainbows inside embedded monocular iris device
A series of non-embedded monocular iris identification devices of film feature extraction, the preservation of iris feature, comparison and identification etc. are upper
The function that machine is realized, has widened the application environment of device, has simplified the development difficulty of upper computer software, improves device operation
Stability, while also strengthening flexibility and the transplantability of device application.
In the present invention, by the iris feature value on external memory and in user data all copies to when system electrification
In memory, the certification of all iris features is all completed in memory.It is every in compared with the existing technology to compare once just from external memory
The method that a characteristic value is taken in reservoir, the present invention realize that registration, the process speed of identification are many fastly, and main cause is exactly pair
On the visit speed of built-in storage is tens times faster than the access speed to external memory.
In the present invention, embedded monocular iris identification device in addition to from host computer receive register instruction, 1:1 certification instruction,
Other than 1:N certification instruction, other instructions can also be received from host computer, such as: user deletes, user edits and user searches
Deng.Device can also work independently in the case where being detached from host computer, further enhance flexibility, the transplantability of device simultaneously,
Human-computer interaction is enhanced simultaneously.
In the present invention, image capture module 104 can obtain image from camera lens 102 in a manner of conventional continuous scanning, so
And in order to improve speed, it can preferably use mode shown in fig. 5.For the method for iris feature registration/certification step 4,
4 ' or 4 " can further include:
Step 41: image capture module 104 obtains image from camera lens 102 with interlace mode, and is sent at image
Reason module 103 is pre-processed, and pretreated image is stored in built-in storage 105；
In this step, interlacing scan refers to that each frame image is divided into two parts, and each section contains a frame image
In all odd-numbered scan lines or odd scan line.
Step 42:CPU101 detects whether pretreated image in built-in storage 105 is iris image, if it is,
Instruction image capture module 104 is switched to continuous scanning mode and reacquires image, the figure that image capture module 104 reacquires
It is pre-processed as being still sent to image processing module 103, pretreated image is stored in built-in storage 105；
In this step, image information can be acquired by interleaved mode, and then whether detection image is iris figure
Picture during this, only obtains the half of complete image, compared to continuous scanning mode is used, the speed of iris detection is mentioned
It is one times high, the experience effect of user is also enhanced while improving detection speed.
Step 43:CPU101 detects whether newest pretreated image in built-in storage 105 is iris image, if
It is to execute step 5,5 ' or 5 ".
Due to needing whether be that iris image determines, and then reaches the acquisition to iris image to image, step 41-
In 43, the judgement of image is carried out using interlace mode, the acquisition of image is carried out using continuous scanning mode, and this method guarantees
Image is distortionless while again improving work efficiency.
For above-mentioned steps 42, when work, can be set to timing acquiring, make 104 separated in time of image capture module
Continuous collecting image, correspondingly, image processing module 103 also persistently pre-processes image；
Also, step 5,5 ' or 5 " can further include:
CPU continuously judges whether a certain number of pretreated images meet image quality requirements, and only retains image
The best piece image of effect, then goes to step 6,6 ' or 6 ".
In this step, embedded monocular iris identification device can be with continuous acquisition certain amount (such as 2,3,4,5,6)
Iris image is opened, the selection of specific number can take into account the processing speed of image effect and whole device to determine, then by it
In a top-quality extraction characteristic value and save.The quality of the client iris image of preservation can be greatly improved in this way, protected
Demonstrate,prove the accuracy of identification.
In order to guarantee the time of Image Acquisition, CPU101 can be with 4,4 ' or 4 " holding to step 6,6 ' or 6 " of monitoring step
The row time, if it times out, going to step 7,7 ' or 7 ".
In this step, judge whether that time-out is to realize by software instruction to the timing of timer, timing can be with
It is flexibly chosen according to the needs of actual use scene, such as 30s, 1min.
In order to cope with the strong and weak variation of extraneous light, guarantee the acquisition quality of image, in step 4,4 ' or 4 ", works as image
When acquisition module 104 obtains image from camera lens 102, light compensating lamp is opened simultaneously.
When it is iris image that CPU101, which is detected, it can control LED light and/or buzzer provide prompt, realize
Human-computer interaction, enhances the experience of the process of user.Wherein, LED light can using red and green color LED light or other
Double-colored or monochromatic LED indicator light, by taking red and green color LED light as an example, whens specific works, is referred to following method:
When device carries out Image Acquisition, red LED indicator light is lighted；
When device detects iris image, red LED indicator light is closed, green LED indicator lamp is lighted；
When device needs to resurvey image, green LED indicator lamp is closed, red LED indicator light is lighted.
By the prompt of above-mentioned LED light, user enhances user's it is known that whether equipment detects iris
Experience effect embodies a kind of humanized design.
To reduce power consumption, when embedded monocular iris identification device " free time ", light compensating lamp and LED light can be with
It is all closed, only when receiving instruction, just opens, after having executed, be closed again.While in order to save
The use of CPU101, when CPU101 is in " free time ", the present invention can lower the dominant frequency of system, when receiving work order,
Dominant frequency is risen to normal working frequency again by CPU101, after having executed, frequency reducing again.
The monocular iris identification equipment of the prior art is a kind of non-embedded system using the method for CPLD acquisition image,
Essence is a kind of acquisition mode of software, its speed is slow, and stability is poor.Compared with prior art, embedded single of the invention
Mesh iris identification device is built using hardware module, wherein including CPU, image processing module and image capture module etc.
Electronic component, the acquisition of image are faster, more stable.
The embedded monocular iris identification device that the present invention realizes, can (we be logical in DDR2 when acquiring image
Often say " memory ") in cache multiple images, CPU only needs to take image from caching when needing to handle image,
Without waiting for a current frame image；Rather than embedded monocular iris identification equipment is using DSP+CPLD structure or similar hard
Part cannot achieve the caching of image, can only adopt a frame image, handle a frame image due to being limited by memory space.
The embedded monocular iris identification equipment that the present invention realizes is acquiring image, and image is then transmitted to image procossing
Resume module, the image for handling completion are placed to the process stored in DDR2, and centre does not need the participation of CPU, all by hard
Part is automatically performed, and is thus made CPU in image acquisition and processing while be can handle other things, and the use effect of CPU is improved
Rate；Rather than embedded monocular iris identification equipment using DSP+CPLD structure or similar hardware acquisition process image when
It waits, DSP cannot handle other things simultaneously.
Below with reference to a specific embodiment, the method flow that the present invention carries out iris feature registration is discussed in detail, such as
Shown in Fig. 6, comprising:
Step 11: host computer sends register instruction to embedded monocular iris identification device, wherein contains in register instruction
Register the ID number and right and left eyes mark of user；
It after step 12:CPU receives register instruction, is parsed, while obtaining the ID number and right and left eyes mark of user；
Step 13:CPU judges whether the eye of the ID number is registered by above-mentioned ID number and right and left eyes mark, if
It does not register, thens follow the steps 14, otherwise, set " repeated registration " for response value, be uploaded to host computer；
Step 14: opening light compensating lamp, light red LED indicator light, starting timer (can timing 30s), then carry out
Image Acquisition thens follow the steps 15 if image successful acquisition before the deadline, otherwise sets " time-out " for response value,
It is uploaded to host computer, simultaneously closes off light compensating lamp, LED light and timer；
Step 15:CPU takes out image data from image buffer storage, and detects whether the image is iris image, if so,
Step 16 is executed, otherwise, return step 14 resurveys image information；
Step 16: the quality score of image is obtained by CPU, while point gives a green light, and closes red LED indicator light, prompts
It detects iris image, and judges whether the image is corresponding the first width of eye iris image, if it is not, 17 are thened follow the steps, it is no
Then, then the image score value and number are recorded, and the image is copied in interim buffering and is saved, while covering upper piece image
Data；
Step 17:CPU judges whether the quality of the width image is good with upper piece image by quality score, if it is not, then
Step 18 is executed, otherwise, then records the image score value and number, and the image is copied in interim buffering and is saved, simultaneously
Piece image data in covering；
Step 18: judging whether to be extracted n (n can be 5) iris image by number, if so, executing step
Rapid 19, otherwise, red LED indicator light is lighted, closes green LED indicator lamp, and return step 14, resurvey image information；
Step 19: extracting iris feature value and save, set success for response value, be uploaded to host computer, simultaneously close off
Light compensating lamp, LED light and timer；
Step 20: host computer parses response instruction, obtains processing result.
Below with reference to a specific embodiment, the method flow that the present invention carries out the certification of 1:1 iris feature is discussed in detail,
As shown in fig. 7, comprises:
Step 21: host computer authenticates to contain in instruction and recognize to the transmission 1:1 certification instruction of embedded monocular iris identification device
Demonstrate,prove the ID number and right and left eyes mark of user；
Step 22:CPU parses instruction, obtains 1:1 certification instruction, while obtaining ID number and the left and right of certification user
Eye mark；
Step 23:CPU judges whether the eye of the ID has been registered, if having registered, thens follow the steps 24, otherwise, will
Response value is set as " empty user ", is uploaded to host computer；
Step 24: opening light compensating lamp, light red LED indicator light, starting timer (can timing 30s), then carry out
Image Acquisition thens follow the steps 25 if image successful acquisition before the deadline, otherwise sets " time-out " for response value,
It is uploaded to host computer, simultaneously closes off light compensating lamp, LED light and timer；
Step 25:CPU takes out image data from image buffer storage, and detects whether the image is iris image, if so,
Step 26 is executed, otherwise, return step 24 resurveys image information；
Step 26: obtaining the quality score of image by CPU, while lighting green LED indicator lamp, close red LED and refer to
Show lamp, prompts to detect iris image, and judge whether the image is corresponding the first width of eye iris image, if it is not, then executing
Step 27, otherwise, then the image score value and number are recorded, and the image is copied in interim buffering and is saved, is covered simultaneously
Upper piece image data；
Step 27:CPU judges whether the quality of the width image is good with upper piece image by quality score, if it is not, then
Step 28 is executed, otherwise, then records the image score value and number, and the image is copied in interim buffering and is saved, simultaneously
Piece image data in covering；
Step 28: judging whether to be extracted n (n can be 5) iris image by number, if so, executing step
Rapid 29, otherwise, red LED indicator light is lighted, closes green LED indicator lamp, and return step 24, resurvey image information；
Step 29: it extracts iris feature and saves, and be compared with the feature of this of the user stored in device,
If comparison passes through, response value is set as " succeeding ", is uploaded to host computer, simultaneously closes off light compensating lamp, LED light and timing
Device, if not over the picture number recorded is recorded and reset by CPU, lights red LED indicator light, closes green LED instruction
Lamp, and return step 24, resurvey image information；
Step 30: host computer parses response instruction, obtains processing result.
Below with reference to a specific embodiment, the method flow that the present invention carries out the certification of 1:N iris feature is discussed in detail,
As shown in Figure 8, comprising:
Step 31: host computer sends 1:N certification instruction to embedded monocular iris identification device；
Step 32:CPU parses instruction, obtains 1:N certification instruction；
Step 33:CPU judges whether that existing subscriber registers, if having registered, 34 is thened follow the steps, otherwise, by response value
It is set as " empty user ", is uploaded to host computer；
Step 34: opening light compensating lamp, light red LED indicator light, starting timer (can timing 30s), then carry out
Image Acquisition thens follow the steps 35 if image successful acquisition before the deadline, otherwise sets " time-out " for response value,
It is uploaded to host computer, simultaneously closes off light compensating lamp, LED light and timer；
Step 35:CPU takes out image data from image buffer storage, and detects whether the image is iris image, if so,
Step 36 is executed, otherwise, return step 34 resurveys image information；
Step 36: obtaining the quality score of image by CPU, while lighting green LED indicator lamp, close red LED and refer to
Show lamp, prompts to detect iris image, and judge whether the image is corresponding the first width of eye iris image, if it is not, then executing
Step 37, otherwise, then the image score value and number are recorded, and the image is copied in interim buffering and is saved, is covered simultaneously
Upper piece image data；
Step 37:CPU judges whether the quality of the width image is good with upper piece image by quality score, if it is not, then
Step 38 is executed, otherwise, then records the image score value and number, and the image is copied in interim buffering and is saved, simultaneously
Piece image data in covering；
Step 38: judging whether to be extracted n (n can be 5) iris image by number, if so, executing step
Rapid 39, otherwise, red LED indicator light is lighted, closes green LED indicator lamp, and return step 34, resurvey image information；
Step 39: extracting iris feature value and save, and compared one by one with the iris feature stored in device, if comparing logical
It crosses, is then set as " succeeding " by response value, be uploaded to host computer, simultaneously close off light compensating lamp, LED light and timer, if not having
Have and pass through, the picture number recorded is recorded and reset by CPU, lights red LED indicator light, closes green LED indicator lamp, and return
Step 34 is returned, image information is resurveyed；
Step 40: host computer parses response instruction, obtains processing result.
In conclusion benefit of the invention is:
(1) in the case where not needing host computer or other external devices are interfered, it will be able to complete acquisition and the place of image
Reason, departing from the dependence to host computer, is effectively simplified the software realization of host computer in the prior art, makes embedded monocular iris
The stability of identification device improves.
(2) setting of the present invention to built-in storage and external memory, realizes the caching to data and storage, CPU is made to exist
When data processing, in advance by the data copy in external memory into built-in storage, CPU can be called in built-in storage in time
Data are handled, and the data processing speed of CPU is effectively increased.
(3) in the present invention, image information can be acquired by interleaved mode, and then whether detection image is iris
Image during this, only obtains the half of complete image, compared to using continuous scanning mode, the speed of iris detection
It is doubled, further increases the working efficiency of CPU, also enhance the experience effect of user.
(4) serial ports and USB interface are additionally provided in the present invention, can applies, can also be individually answered simultaneously when application
With, provide the leeway of selection for client, facilitate the secondary development of user, also widened the application environment of equipment, flexibility and
Transplantability is reinforced.
The above is a preferred embodiment of the present invention, it is noted that for those skilled in the art
For, without departing from the principles of the present invention, it can also make several improvements and retouch, these improvements and modifications
It should be regarded as protection scope of the present invention.
Claims (6)
1. a kind of embedded monocular iris authentication system, which is characterized in that filled including host computer and embedded monocular iris recognition
It sets, the embedded monocular iris identification device includes CPU and camera lens, and the CPU is connected with image capture module, image
Processing module, built-in storage and external memory, and the CPU is also connected with serial ports and/or USB interface, in which:
The data input pin of described image acquisition module connects the camera lens, and data output end connects described image processing module；
Described image processing module also connects the built-in storage；
The method that the embedded monocular iris authentication system carries out the certification of 1:1 iris feature includes:
Step 1 ': it powers on, CPU reads the user data stored in external memory in built-in storage, wherein the number of users
According to ID, right and left eyes mark and the iris feature value for including user；
Step 2 ': CPU receives 1:1 certification instruction from host computer, includes ID and the left side of user to be certified in the certification instruction
Right eye mark；
Step 3 ': CPU judges whether the eye of the ID number has been registered according to the user data in built-in storage, if so, holding
Row step 4 '；
Step 4 ': image capture module obtains image from camera lens, and is sent to image processing module and is pre-processed, after pretreatment
Image be stored in built-in storage, CPU detect built-in storage in pretreated image whether be iris image, if so,
Execute step 5 '；
Step 5 ': CPU judges whether pretreated image meets image quality requirements, if so, executing step 6 '；
Step 6 ': CPU is to pretreated image zooming-out iris feature value, and the iris feature value and built-in storage that will be extracted
In the eye iris characteristic value of the user be compared；
Step 7 ': CPU feeds back response result to host computer；
Alternatively,
The method that the embedded monocular iris authentication system carries out the certification of 1:N iris feature includes:
Step 1 ": powering on, and CPU reads the user data stored in external memory in built-in storage, wherein the number of users
According to ID and iris feature value including user；
Step 2 ": CPU receives 1:N certification instruction from host computer；
Step 3 ": CPU judges whether that existing subscriber registers according to the user data in built-in storage, if so, executing step
4"；
Step 4 ": image capture module obtains image from camera lens, and is sent to image processing module and is pre-processed, after pretreatment
Image be stored in built-in storage, CPU detect built-in storage in pretreated image whether be iris image, if so,
Execute step 5 "；
Step 5 ": CPU judges whether pretreated image meets image quality requirements, if so, executing step 6 "；
Step 6 ": CPU is to pretreated image zooming-out iris feature value, and the iris feature value and built-in storage that will be extracted
The iris feature value of middle user compares one by one；
Step 7 ": CPU feeds back response result to host computer；
The step 4 ' further comprise:
Step 41: image capture module obtains image from camera lens with interlace mode, and is sent to image processing module progress
Pretreatment, pretreated image are stored in built-in storage；
Step 42:CPU detects whether pretreated image in built-in storage is iris image, if it is, instruction image is adopted
Collection module is switched to continuous scanning mode and reacquires image, and the image that image capture module reacquires still is sent at image
Reason module is pre-processed, and pretreated image is stored in built-in storage；
Step 43:CPU detects whether newest pretreated image in built-in storage is iris image, if so, executing step
5'；
In the step 42, image capture module separated in time continuous collecting image, correspondingly, image processing module is also held
It is continuous that image is pre-processed；
The step 5 ' further comprise:
CPU continuously judges whether a certain number of pretreated images meet image quality requirements, and only retains image effect
Best piece image, then goes to step 6 '；
The step 4 " further comprises:
Step 41: image capture module obtains image from camera lens with interlace mode, and is sent to image processing module progress
Pretreatment, pretreated image are stored in built-in storage；
Step 42:CPU detects whether pretreated image in built-in storage is iris image, if it is, instruction image is adopted
Collection module is switched to continuous scanning mode and reacquires image, and the image that image capture module reacquires still is sent at image
Reason module is pre-processed, and pretreated image is stored in built-in storage；
Step 43:CPU detects whether newest pretreated image in built-in storage is iris image, if so, executing step
5"；
In the step 42, image capture module separated in time continuous collecting image, correspondingly, image processing module is also held
It is continuous that image is pre-processed；
The step 5 " further comprises:
CPU continuously judges whether a certain number of pretreated images meet image quality requirements, and only retains image effect
Best piece image, then goes to step 6 ".
2. embedded monocular iris authentication system according to claim 1, which is characterized in that described image acquisition module is
Cmos image acquisition module, described image acquisition module connect the camera lens by CMOS interface；
Light compensating lamp and LED light are equipped at the camera lens；
The CPU is also connected with buzzer；
The built-in storage is DDR2, and the external memory is FLASH, and the serial ports is RS232 interface.
3. embedded monocular iris authentication system according to claim 1 or 2, which is characterized in that the system carries out iris
Feature registration method include:
Step 1: powering on, CPU reads the user data stored in external memory in built-in storage, wherein the number of users
According to ID and the right and left eyes mark for including user；
Step 2:CPU receives register instruction from host computer, includes the ID and right and left eyes of user to be registered in the register instruction
Mark；
Step 3:CPU judges whether the eye of the ID number has been registered, if not, executing step 4；
Step 4: image capture module obtains image from camera lens, and is sent to the pretreatment of image processing module row, pretreated
Image is stored in built-in storage, and CPU detects whether pretreated image in built-in storage is iris image, if so, holding
Row step 5；
Step 5:CPU judges whether pretreated image meets image quality requirements, if so, executing step 6；
Step 6:CPU is to pretreated image zooming-out iris feature value and saves；
Step 7:CPU feeds back response result to host computer.
4. embedded monocular iris authentication system according to claim 3, which is characterized in that CPU also monitoring step 4 to step
The rapid 6 execution time, if it times out, going to step 7；CPU also monitoring step 4 ' is to step 6 ' the execution time, if super
When, then go to step 7 '；CPU also monitoring step 4 " is to step 6 " the execution time, if it times out, going to step 7 ".
5. embedded monocular iris authentication system according to claim 3, which is characterized in that the step 4,4 ' or 4 "
In, when image capture module obtains image from camera lens, open simultaneously light compensating lamp.
6. embedded monocular iris authentication system according to claim 3, which is characterized in that the step 4,4 ' or 4 "
In, when it is iris image that CPU, which is detected, controls the LED light and/or buzzer provides prompt.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510065270.5A CN104680137B (en) | 2015-02-06 | 2015-02-06 | Embedded monocular iris identification device, system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510065270.5A CN104680137B (en) | 2015-02-06 | 2015-02-06 | Embedded monocular iris identification device, system and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104680137A CN104680137A (en) | 2015-06-03 |
| CN104680137B true CN104680137B (en) | 2019-01-25 |
Family
ID=53315155
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510065270.5A Active CN104680137B (en) | 2015-02-06 | 2015-02-06 | Embedded monocular iris identification device, system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104680137B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105447443B (en) * | 2015-06-16 | 2019-03-19 | 北京眼神智能科技有限公司 | Improve the method and device of iris identification device identification accuracy |
| CN105426838A (en) * | 2015-11-18 | 2016-03-23 | 北京天诚盛业科技有限公司 | Method for improving iris image detection speed in iris identification device |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2001040982A2 (en) * | 1999-12-01 | 2001-06-07 | Iridian Technologies | System and method of fast biometric database searching using digital certificates |
| CN2768661Y (en) * | 2004-09-21 | 2006-04-05 | 中国科学院自动化研究所 | Portable irides image collecting device |
| CN101369311A (en) * | 2008-09-26 | 2009-02-18 | 北京中科虹霸科技有限公司 | Miniature iris recognition module adopting active visual feedback |
| CN103150565A (en) * | 2013-02-06 | 2013-06-12 | 北京中科虹霸科技有限公司 | Portable two-eye iris image acquisition and processing equipment |
-
2015
- 2015-02-06 CN CN201510065270.5A patent/CN104680137B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2001040982A2 (en) * | 1999-12-01 | 2001-06-07 | Iridian Technologies | System and method of fast biometric database searching using digital certificates |
| CN2768661Y (en) * | 2004-09-21 | 2006-04-05 | 中国科学院自动化研究所 | Portable irides image collecting device |
| CN101369311A (en) * | 2008-09-26 | 2009-02-18 | 北京中科虹霸科技有限公司 | Miniature iris recognition module adopting active visual feedback |
| CN103150565A (en) * | 2013-02-06 | 2013-06-12 | 北京中科虹霸科技有限公司 | Portable two-eye iris image acquisition and processing equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104680137A (en) | 2015-06-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104700077B (en) | Embedded binocular iris identification device, system and method | |
| CN105868613A (en) | Biometric feature recognition method, biometric feature recognition device and mobile terminal | |
| CN107483811A (en) | Imaging method and electronic installation | |
| US20100080422A1 (en) | Finger Vein Authentication Apparatus and Finger Vein Authentication Method | |
| CN109727344A (en) | 3D Intelligent door lock capable of recognizing face and 3D face unlocking method | |
| CN102542263A (en) | Multi-mode identity authentication method and device based on biological characteristics of fingers | |
| CN111368780B (en) | Combined recognition processing method and system based on fingerprint and finger vein | |
| CN106056064A (en) | Face recognition method and face recognition device | |
| CN104680137B (en) | Embedded monocular iris identification device, system and method | |
| CN204791017U (en) | Mobile terminal users authentication device based on many biological characteristics mode | |
| CN107239772A (en) | Palm print and palm vein image collecting device and clearance gate | |
| CN101282265B (en) | Method for driving vision sensor based on wireless sensor network node | |
| CN202771437U (en) | Multi-modal identity authentication device based on biological characteristics of finger part | |
| CN106161916A (en) | A kind of image-pickup method and electronic equipment | |
| CN107527025A (en) | A kind of recognition of face intelligent monitor system | |
| CN110013217A (en) | Detection consumes illegal drugs method, apparatus, terminal device and the medium of personnel | |
| CN109727194A (en) | A kind of method, electronic equipment and storage medium obtaining pet noseprint | |
| CN107800962A (en) | Image processing method, device and electronic installation | |
| CN104933427A (en) | Embedded finger three-modal characteristic acquisition equipment | |
| CN203552331U (en) | Intelligent identification door control system | |
| CN106548530A (en) | Wireless mobile attendance recorder | |
| CN205845057U (en) | A kind of identity card reader carrying testimony of a witness comparison function | |
| CN107979725A (en) | Reduce the method and terminal of intelligent terminal power consumption when taking pictures | |
| CN209122167U (en) | A kind of low-power consumption capsule endoscope Image Acquisition and three-dimensional reconstruction system | |
| CN205959217U (en) | Authentication device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: 100085, 1 floor 8, 1 Street, ten Street, Haidian District, Beijing. Applicant after: Beijing Eyes Intelligent Technology Co.,Ltd. Address before: 100085, 1 floor 8, 1 Street, ten Street, Haidian District, Beijing. Applicant before: BEIJING TECHSHINO TECHNOLOGY Co.,Ltd. |
|
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20181228 Address after: 100085, 1 floor 8, 1 Street, ten Street, Haidian District, Beijing. Applicant after: Beijing Eyes Intelligent Technology Co.,Ltd. Applicant after: BEIJING EYECOOL TECHNOLOGY Co.,Ltd. Address before: 100085, 1 floor 8, 1 Street, ten Street, Haidian District, Beijing. Applicant before: Beijing Eyes Intelligent Technology Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |