CN1520161A - Face recognition method, face extraction method and pick-up device - Google Patents
Face recognition method, face extraction method and pick-up device Download PDFInfo
- Publication number
- CN1520161A CN1520161A CNA2003101204192A CN200310120419A CN1520161A CN 1520161 A CN1520161 A CN 1520161A CN A2003101204192 A CNA2003101204192 A CN A2003101204192A CN 200310120419 A CN200310120419 A CN 200310120419A CN 1520161 A CN1520161 A CN 1520161A
- Authority
- CN
- China
- Prior art keywords
- face
- image
- photographs
- blood
- eye illness
- 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.)
- Granted
Links
Images
Abstract
Provided is a face recognition method for more accurately recognizing face portions contained in an image, face extraction method for extracting the recognized faces, and image pickup apparatus. A subject is photographed with a camera whose incidence of defects such as red-eye or other eye color change has been increased intentionally by designing the flash unit and lens to be extremely close to each other, and so on. Discolored eye portions are detected in a photographic image and the faces of people contained in the photographic image are recognized based on the discolored eye portions. The discolored eye portions in the photographic image are corrected and the recognized face portions are extracted from the corrected photographic image. This makes it possible to generate corrected facial images more accurately.
Description
Technical field
Facial extracting method that the face recognition method that the present invention relates to the personage's face that is included in the image is discerned, the face that will identify extract and the photography of carrying out subject, obtain photographs, and discern the camera head of the personage's face that is comprised in this photographs.
Background technology
In the past, followed popularizing of digital camera, carried out widely the operation of photographs numberization.For example, the photographs that is stored in the film increases in the keeping upper volume, when the developing and printing photographs, must rely on photofinishing shop usually, makes troubles thus.Relative therewith, by the photographs of numberization, its advantage be can a plurality of photographs centralized stores in FD etc., so volume does not increase, no matter when and use PC and printer can print photographs.Have again,, for example, can use PC that photographs is implemented desirable image processing, revise the unfavorable condition of photographs etc. as the advantage of number photographs.As above-mentioned image processing, flash of light when pair strobo photography is for example arranged takes eyes the processing revised for red or golden blood-shot eye illness or gold eye after the reflection of amphiblestroid depths, to the processing revised because of closing one's eyes of the closed eyes of flash of light, the color correct of the color of sky or skin is the processing of good-looking color in appearance and gray scale correcting process etc., by the quilt photographs of numberization is implemented these image processing, can obtain better image (for example, with reference to patent documentation 1 and patent documentation 2).
In addition, in recent years, also begin the photographic image data of collective's photograph usefulness is implemented the image processing of regulation, from photographs, extract everyone face, generate the view data of expression face-image,, make individual photograph according to this view data.
[patent documentation 1] spy opens flat 10-233929 communique
[patent documentation 2] spy opens flat 11-127371 communique
For from by the view data that generates individual photograph the photographic image data of collective's photograph, at first, the face organ who becomes mark who detects in the photographs to be comprised (below be referred to as the mark organ), according to the face in this mark organ identification photographs, generate facial face-image is extracted in representative from photographs view data.
Here, a plurality of when facial when extracting from photographs, for example, if the mark organ is eyes, if produced blood-shot eye illness in the photographs or close one's eyes etc., the deviation of mark organ is big, and it is difficult then detecting the mark organ.Consequently, existence can not correctly detect the mark organ, obscures the possibility of the face that can discern and the face that can not discern.Have, because the flash light emission device and the distance between the camera lens of camera can not be fully long, so increased the generation of blood-shot eye illness or gold eye, there is the difficult more problem of eyes that correctly detects in the photographs in nearest camera to miniaturization development again.
Summary of the invention
The present invention in view of the above problems, its purpose is to provide a kind of facial face recognition method that comprises in can the high accuracy recognition image, extraction identifies facial facial extracting method, and the shooting of carrying out subject, obtain photographs, can high accuracy discern the camera head of the personage's face that comprises in this photographs.
The face recognition method of the present invention that achieves the above object, facial in the recognition image according to the view data of presentation video, have:, detect the testing process of the eyes part that in image, generates the regulation variable color according to view data; With basis detected eyes part in testing process, the identifying of the face in the recognition image.
In the past, the known eyes that detect in the photographs were discerned the face recognition method that comprises face in the photographs according to detected eyes.During eyes in detecting photographs, though be general eye color (black or indigo plant etc.) in the detection photographs and the image section of eye shape (circle), but for example produce blood-shot eye illness or gold at the moment in photographs, the whole eyes that correctly detect in the photographs to be comprised are difficult.Consequently, discernible face and the face that can not discern might appear in photographs.
Yet in recent years, the miniaturization of camera is in development, because the flash light emission device of camera and the distance between the camera lens can not grow, so the generation rate rising of blood-shot eye illness or gold eye.In addition, follow the rising of the generation rate of eyes variable color, descend gradually based on the face recognition rate of eyes in the above-mentioned photographs.Relative therewith, recognition methods of the present invention is conceived to be created in the eyes part of detected variable color in the testing process, according to this variable color the eyes part, identification is facial in identifying.Though it is difficult avoiding the generation of eyes variable color, it is easy deliberately making the eyes variable color, thus according to face recognition method of the present invention, use the photographs that the eyes variable color is taken place, can high accuracy and identify face easily.
In addition, in face recognition method of the present invention, preferred above-mentioned testing process is the process of the blood-shot eye illness part in the detected image.
If the flash light emission device of the camera of shooting subject and the position of camera lens are extremely near, then can improve the generation rate of blood-shot eye illness especially.In addition, detect the processing of the blood-shot eye illness in the photographs, because widely-used in the past, the savings of technology is abundant, uses these skill keys high Precision Detection to go out blood-shot eye illness.Therefore, by using the camera subject that causes this blood-shot eye illness design easily, detect the blood-shot eye illness in the photographs, identification is facial, can discern facial more accurate and effectively.
In addition, reach the of the present invention facial extracting method of above-mentioned purpose, according to the view data of presentation video, the face in the extraction image has to generate face-image: according to the testing process of the part of the blood-shot eye illness in the view data detected image; According to detected blood-shot eye illness part in testing process, the identifying of the face in the recognition image; The makeover process of correction detected blood-shot eye illness part in testing process; In identifying, identified and in makeover process, seen red the face that part has been corrected with from image, extracting, to generate the face-image generative process of face-image.
In the past, carry out everyone face of identification from the photographs that group picture is used, extracted face, generated individual photograph.At this moment, by being suitable for as above-mentioned face recognition method, partly discern facial method according to the blood-shot eye illness in the photographs, identification is facial, extracts the face that identifies, and generates face-image, can be accurately and generate face-image effectively.
In addition, facial extracting method of the present invention, also can be a kind of view data, extract face in the image, possess: the testing process of the generation regulation variable color that in image, detects according to view data to generate the facial extracting method of face-image according to representative image; According to detected eyes part in testing process, the identifying of the face in the recognition image; The makeover process of correction variable color of detected eyes part in testing process; With the face that extraction has identified in identifying and variable color has been corrected in makeover process from image, to generate the face-image generative process of face-image.
Extracting method of the present invention is not defined in blood-shot eye illness, for example can be to discern face according to the eyes variable color of gold eye etc. yet, extracts the face that identifies to generate the method for face-image.Owing to deliberately make the blood-shot eye illness generation also easily,, can precision more discern face in the highland, thereby can correctly generate face-image by discerning face according to blood-shot eye illness than the eyes variable color generation that makes other.
Have again, in order to reach above-mentioned purpose, face recognition device of the present invention, it is a kind of view data according to presentation video, the face recognition device of face in the recognition image has: according to view data, detect the test section of the eyes part of the variable color that generates regulation in image; With basis detected eyes part in test section, the identification part of the face in the recognition image.
Use face recognition device of the present invention, for example extremely near in the position of flash light emission device and camera lens, be easy to generate in the camera of blood-shot eye illness or gold eye etc., can identify the face that is comprised in the image of shooting more accurately.
Also have, in order to reach above-mentioned purpose, camera head of the present invention, be a kind of subject to be photographed, generate the camera head of the photographic image data of expression photographs, possess:, in photographs, detect the test section of the eyes part of the variable color that generates regulation according to photographic image data; According to by the detected eyes part of test section, discern the identification part of the face in the photographs; Correction is in the correction portion of the variable color of the detected eyes part of test section; With the face that extraction has identified in the identification part and variable color has been corrected in correction portion from photographs, to generate the face-image generating unit of face-image.
Camera head of the present invention, it is extremely near etc. to be designed to the position of flash light emission device and camera lens, by improving the incidence of eyes variable color, detects the variable color part of the eyes in the photographs, extraction can generate face-image more accurately according to the face that the variable color of these eyes part identifies.
Have again, in camera head of the present invention, the blood-shot eye illness part in the preferred above-mentioned test section detected image, the correction portion correction is by the detected blood-shot eye illness part of test section.
According to the present invention, can provide a kind of and can high accuracy be identified in the face recognition method that comprises face in the image, extraction identifies facial facial extracting method, and the shooting of carrying out subject, obtain photographs, can high accuracy be identified in the camera head of the personage's face that comprises in this photographs.
Description of drawings
Fig. 1 is the outside drawing of the digital camera of expression one embodiment of the present invention.
Fig. 2 is the block diagram of digital camera.
Fig. 3 is face recognition processing and the facial functional block diagram of handling relevant portion that extracts in the image processor.
Fig. 4 is that expression obtains photographs, when this photographs is stored in video memory shown in Figure 2, and the flow chart of a series of processing of in image processor, carrying out.
Fig. 5 is the figure that the photographs of the photographic image data that is input to measuring ability is represented in expression.
Fig. 6 is the figure of the testing result of the expression blood-shot eye illness that detects in the photographs to be comprised.
Fig. 7 is the testing result of expression according to blood-shot eye illness, the result's of the personage's face that identifies in the photographs to be comprised figure.
Fig. 8 is the testing result of expression according to blood-shot eye illness, has revised the figure of photographs after the correction of the blood-shot eye illness that is comprised in the photographs.
Fig. 9 is that the figure based on face recognition result's face-image face-image has partly been extracted in expression from revise the back photographs.
Figure 10 is the flow chart of a series of photograph processing of carrying out in the digital camera of expression the 2nd execution mode.
Figure 11 is the functional block diagram that is suitable for the personal identification system of the 3rd execution mode of the present invention.
Figure 12 is the photographs of representing according to the camera head shooting, the flow chart of a series of processing of the personage in the identification photographs.
Among the figure: 100-digital camera, 110-view finder, 120-shutter release button, the 130-phtographic lens, 140-flash light emission device, 150-dimmer sensor, the 200-image processor, 201-measuring ability, 202-recognition function, the 203-debugging functions, 204-face-image systematic function, 210-timing sequencer, 211-CCD, the 220-AD transducer, the 230-image shows LCD, 240-high-speed computation memory, 250-video memory, 300-control microcomputer, the 310-control part that exposes, the 311-shutter, 320-focus control portion, 330-zoom control part, 340-flash light emission portion, the 350-power control part, the 360-switches set, 370-state LCD, 400-power supply, the 500-external device (ED), the 600-photographs, 601-personage, the testing result of 610-blood-shot eye illness, the 611-eyes, the recognition result of 620-face, 621-face, 630-are revised the back photographs, 631-personage, the 640-face-image, 700-personal identification system, 710-camera head, the 720-PC, 721-sees red test section, 722-face recognition portion, 723-individual identification portion, the 724-storage part, the 725-image displaying part.
Embodiment
Below, embodiments of the present invention are described.
Fig. 1 is the outside drawing of the digital camera of expression one embodiment of the present invention.
On the front of digital camera 100, possess in appearance: the shutter release button 120 that is pressed during photography, with the flash light emission device 140 of pressing synchronous light-emitting of shutter release button 120, the dimmer sensor 150 of measuring the light quantity of sending from flash light emission device 140, view finder 110 that the cameraman observes for the position of determining subject and by focus being accumulated in the condenser lens on the subject or switching the phtographic lens 130 that the zoom lens etc. at photography visual angle constitute.Shutter release button 120 can be divided into 2 stages of partly pressing and pressing entirely, if partly press shutter release button 120, then along the motor on the condenser lens of direction drive installation in phtographic lens 130 of optical axis extending, focus is focused on the middle section at photography visual angle, setting is maintained to the position of condenser lens the focus-lock of reading (exposure) of shot object image, if press shutter release button 120 entirely, then shutter action is carried out actual photography.In addition, flash light emission device 140 is arranged on the extremely near position of phtographic lens 130, becomes the design that blood-shot eye illness is taken place.
Here, the internal structure of logarithmic code camera 100 describes.
Fig. 2 is the block diagram of digital camera.
This digital camera 100, possess: image processor 200, timing sequencer, CCD (ChargeCouple Device: charge coupled device) 211, AD (Analog-Degital) transducer 220, image shows LCD (Liquid Crystal Display) 230, high-speed computation memory 240, video memory 250, control microcomputer 300, exposure control part 310, shutter 311, focus control portion 320, zoom control part 330, flash light emission portion 340, power control part 350, switches set 360, the phtographic lens 130 that also illustrates among state LCD370 and Fig. 1 can be connected with external device (ED)s 500 such as PC.
At first, switches set 360 is described.
In switches set 360, though not shown, possess: follow the pressing of shutter release button 120 of Fig. 1 and the shutter release going into to cut; Start the zoom lens in the phtographic lens 130, being visible dolly-out,s dolly-back switches the ZSW at photography visual angle in the shadow wide-angle photography; Any one the mode selector switch of face-image memory module of the face-image of the personage's face that is comprised in the photographs is extracted in the common memory module that memory module is switched to the storage photographs and storage; Photographs is presented at image shows image display switch on the LCD230; And the state of digital camera 100 is presented at status switch on the state LCD370.Shutter release is the switch that can be set at 2 stages, if shutter release button 120 is partly pressed, then connects the switch in the 1st stage, sets focus-lock, if shutter release button 120 is pressed entirely, then connects the switch in the 2nd stage, shutter 311 actions.
Then, the each several part beyond the switches set 360 is described.
CCD211 receives object light, this object light is transformed to the subject signal that belongs to analog signal.The subject signal that has been transformed according to the sequential that generates, is sent to AD converter 220 from CCD211 in timing sequencer.The subject signal transformation that AD converter 220 will obtain from CCD211 is the photographic image data that belongs to numerical data.
Image shows that LCD230 is the LCD monitor that shows based on the image of the view data that sends from image processor 200.In addition, high-speed computation is the scratchpad memories that use in image processor 200 with memory 240, the memory of video memory 250 view data that to be storages send here from image processor 200.
After if exposure control part 310 is obtained the information of aperture amount and shutter speed from control with microcomputer 300, the motor of drive installation on shutter 311 then, regulate aperture amount, allow shutter action according to indicated shutter speed simultaneously, the exposure when control receives object light by CCD211.
After focus control portion 320 obtained the information of condenser lens position from control with microcomputer 300, then the motor in the condenser lens that comprises of drive installation phtographic lens 130 made condenser lens move to the condenser lens position that is instructed to.
If zoom control part 330 is from controlling the information that obtains the zoom lens position with microcomputer 300, then the motor of drive installation in the zoom lens that phtographic lens 130 comprises makes zoom lens move to the zoom lens position that is instructed to.
If flash light emission portion 340 then flashes from flash light emission device 140 shown in Figure 1 from controlling the information that obtains the suitable luminous quantity of flash of light with microcomputer 300.The flash light emission amount of sending this moment is measured by dimmer sensor 150, if luminous quantity reaches suitable luminous quantity, then stops luminous.
The digital camera 100 of present embodiment is basically as above-mentioned formation.Here, in digital camera 100, be to handle by red-eye correction processing, face recognition processing and facial extraction that image processor 200 carries out as the feature of one embodiment of the present invention, below these processing are described in detail.
Fig. 3 is red-eye correction processing, face recognition processing and the facial functional block diagram of handling correlation function that extracts of image processor.
Measuring ability 201 is obtained the photographs that has been digitized from the AD converter 220 of Fig. 2.Have again, measuring ability 201, color is the image section of redness and toroidal in the detection photographs, detects the blood-shot eye illness that photographs comprises, the redeye location information of the redeye location that the generation expression has been detected.Blood-shot eye illness is equivalent to an example of the said blood-shot eye illness of the present invention, is equivalent to produce an eyes example partly of regulation variable color simultaneously.This measuring ability 201 is equivalent to an example of test section function in camera head of the present invention and the face recognition device.
Recognition function 202, according to the redeye location information that generates in the measuring ability 201, identification personage's face that photographs comprised, the facial positions information of the facial positions that the generation expression identifies.This recognition function 202 is equivalent to an example of the function of the identification part in camera head of the present invention and the face recognition device.
Debugging functions 203, according to the redeye location information that generates in photographs and the measuring ability 201, the colourity of the image section of seeing red in the photographs is adjusted into the value of regulation, the image section of blood-shot eye illness is modified to color and the brightness that preferably equates with general personage's eye color and brightness as photograph, generates and revise photographs afterwards.Debugging functions 203 are equivalent to an example of the function of the correction portion in the camera head of the present invention.
Face-image systematic function 204 is obtained in the debugging functions 203 the facial positions information that generates in the photographs and recognition function 202 after the correction that generates, and generates from revising the face-image that the back photographs extracts personage's face-image.This face-image systematic function 204 is equivalent to an example of the function of the face-image generating unit in the camera head of the present invention.
Here, below, the cameraman is taken subject, a series of flow processs of storage photographs describe.
At first, the cameraman selects the face-image memory module as memory module, and the example of storage photographs describes.
The cameraman adopts not shown memory module switching push button, selects the face-image memory module as memory module.
If select the face-image memory module by the cameraman, then in the inside of digital camera 100, set the face-image memory module with the mode selector switch that the switches set 360 of Fig. 2 is comprised, the memory module after being set is sent to control usefulness microcomputer 300.Control after receiving memory module, is then delivered to image processor 200 with this memory module with microcomputer 300.
Have, the cameraman observes the view finder 110 of Fig. 1 again, so that desirable subject is disposed at the mode mobile digital camera 100 of the middle section in the photography visual angle, partly presses shutter release button 120.
If the shutter release button 120 by the cameraman partly presses Fig. 1 then in digital camera 100 inside, in the 1st stage of the shutter release that the switches set 360 of connection Fig. 2 is comprised, transmits the information in the 1st stage of on shutter releases to control with microcomputer 300.
At this moment, in image processor 200, obtain the low photographic image data of resolution of processing such as being applied to find range.That is, read object light roughly, generate the subject signal of low resolution, transmit the subject signal of low resolution to AD converter 220 with CCD211.The subject signal of low resolution is transformed to the photographic image data that belongs to digital signal in AD converter 220, and the photographic image data of this low resolution is sent to image processor 200.
Control is determined shutter speed and aperture amount according to the brightness information that obtains from image processor 200 with microcomputer 300, is identified for the condenser lens position of subject in focus by ranging information.In addition, control transmits the information of condenser lens position with microcomputer 300 to focus control portion 320, keeps shutter speed and aperture amount, up to by the cameraman shutter release button 120 of Fig. 1 being pressed entirely.
If a series of preparations of above-mentioned formal photography usefulness are handled (following should a series of preparation processing be called pre-photograph processing) and finished, then begin following illustrated formal photograph processing.
The cameraman presses the shutter release button 120 of Fig. 1 entirely.
If shutter release button 120 pressed entirely, then connect the 2nd stage of the shutter release that switches set shown in Figure 2 360 comprised, and the information in the 2nd stage of on shutter release is delivered to control with microcomputer 300.
Control if be transmitted the information in the 2nd stage of on shutter release, then transmits the information of shutter speed and aperture amount with microcomputer 300 to exposure control part 310.Expose control part 310 if from controlling the information that obtains shutter speed and aperture amount with microcomputer 300, then according to the shutter speed and the aperture amount of indication, regulate aperture, allow shutter 311 move.
If shutter 311 actions, then the object light that is received by CCD211 is transformed to the high subject signal of resolution, delivers to AD converter 220, and this subject signal is transformed to the high photographs signal of resolution in AD converter 220.Then, the photographs signal is sent to image processor 200.
Fig. 4 is that the expression photographs is sent to image processor, the flow chart of a series of processing of carrying out when this photographs is stored in video memory shown in Figure 2 in image processor.Below, use the flow chart of this Fig. 4, describe to the flow process of the stored a series of processing of this photographs be sent to image processor 200 from photographs.Have again, in the time of with reference to Fig. 4, also describe to each figure of Fig. 9 with reference to Fig. 5.
In the measuring ability 201 of image processor shown in Figure 3 200, transmit the situation that is set at the face-image memory module as memory module from the control of Fig. 2 with microcomputer 300, transmit photographic image datas (the step S1 of Fig. 4) from AD converter 220.
As mentioned above, carry out formal photograph processing, obtain photographs.
Fig. 5 is the figure that expression inputs to the represented photographs of the photographic image data of measuring ability.Photographs 600 is to comprise the photographs that a plurality of personages' 601 group picture is used.Digital camera 100 is the cameras that deliberately produce blood-shot eye illness, makes whole personages' 601 that photographs 600 comprised eyes produce blood-shot eye illness.Blood-shot eye illness is to flash when taking the personage, and the high light of the state glitter down of opening at the pupil of eyes is by the capillary reflection on eyeground, and the captured personage's of photographs eye color becomes red phenomenon as a result.This blood-shot eye illness is owing to be to take place easily under the near situation of the camera lens of camera and flash light emission device, the digital camera 100 of present embodiment, and as shown in Figure 1, phtographic lens 130 and flash light emission device 140 are configured on the extremely near position.
The measuring ability 201 of Fig. 3, color is red and is shaped as circular image section in the photographs 600 of detection Fig. 5, detects the blood-shot eye illness (the step S2 of Fig. 4) that photographs 600 is comprised.
Fig. 6 is the figure that expression detects the testing result of the blood-shot eye illness that photographs comprises.In the personage's 601 who in the photographs 600 of Fig. 5, is comprised the eyes,, in the testing result 610 of blood-shot eye illness, comprise the eyes 611 of the whole personages in the photographs 600 owing to all see red.
The measuring ability 201 of Fig. 3 generates the redeye location information that expression detects the position of blood-shot eye illness, transmits the positional information of photographic image datas, blood-shot eye illness and as the face-image memory module of memory module to debugging functions 203.Have, measuring ability 201 also transmits the positional information of photographic image data and blood-shot eye illness to recognition function 202 again.The processing of the step S2 that carries out in measuring ability 201 is equivalent to an example of the testing process in face recognition method of the present invention and the facial extracting method.
In the flow chart of Fig. 4, because what here select as memory module is the face-image memory module, so enter step S5 from step S3.
The recognition function 202 of Fig. 3 after the positional information and photographic image data that transmit blood-shot eye illness from measuring ability, then according to the positional information of this blood-shot eye illness, identifies the personage (the step S5 of Fig. 4) in the photographs of photographic image data representative.
Fig. 7 be expression according to the blood-shot eye illness testing result of Fig. 6, identify result's the figure of the personage's who is comprised in the photographs of Fig. 5 face.In the recognition result 620 of face, the periphery of the eyes of being represented by the blood-shot eye illness testing result 610 of Fig. 6 in the photographs 600 of Fig. 5 611 is identified as facial 621.
The recognition function 202 of Fig. 3 generates the facial positions information of the facial positions that expression identifies, and positional information that should face transmits to debugging functions 203.The processing of the step S5 that carries out in recognition function 202 is equivalent to an example of the identifying in face recognition method of the present invention and the facial extracting method.
Debugging functions 203, according to the photographic image data that transmits from measuring ability 201 and the positional information of blood-shot eye illness, the image section colourity of the position of the expression redeye location information in the photographs of reduction photographic image data representative is revised the blood-shot eye illness (the step S6 of Fig. 4) that is comprised in the photographs.
Fig. 8 be expression according to the blood-shot eye illness testing result correction of Fig. 6 the figure of photographs after the correction of the blood-shot eye illness that comprised in the photographs of Fig. 5.The blood-shot eye illness that the personage's 601 that the photographs 600 of Fig. 5 is comprised eyes produce is corrected for, the personage's 631 that photographs 630 is comprised after revising eyes.
Photographic image data reached the facial positions information that is transmitted from recognition function 202 after the debugging functions 203 of Fig. 3 transmitted the correction of representing correction back photographs 630 to face-image systematic function 204.The processing of the step S5 that carries out in debugging functions 203 is equivalent to an example of the makeover process in the facial extracting method of the present invention.
Face-image systematic function 204, if transmit the positional information of revising back photographic image data and face from debugging functions 203, then photographs extracts the image section of expression facial positions information after the correction of revising back photographic image data representative, generates face-image.(the step S6 of Fig. 4).
Fig. 9 is that expression photographs after the correction of Fig. 8 has extracted the figure based on the face-image of the image section of the face recognition result's of Fig. 7 face.As shown in Figure 9, the personage's who is comprised in the face-image 640 eyes, blood-shot eye illness is corrected, and generates personage 601 crews' that photographs shown in Figure 5 600 comprised face-image 640.
Face-image systematic function 204 will be delivered to video memory shown in Figure 2 250 based on the face image data of face-image.The processing of the step S6 that carries out in this face-image systematic function 204 is equivalent to an example of the face-image generative process in the facial extracting method of the present invention.
Face image data is sent to video memory 250, is stored in the video memory 250.(the step S8 of Fig. 4).
Above-mentioned each processing is carried out finishing (the step S9 of Fig. 4) up to photography repeatedly.
Like this,, blood-shot eye illness is taken place, discern face goodly, extract facial exactly according to this blood-shot eye illness precision if utilize the digital camera 100 of present embodiment.
More than, to select the face-image memory module as memory module, store the explanation that the example of photographs carries out and finish, next illustrate as memory module and select common memory module to store the example of photographs.
The cameraman selects common memory module with not shown memory module switching push button as memory module.
If select common memory module by the cameraman, then in the inside of digital camera 100, the mode selector switch that is comprised by the switches set 360 of Fig. 2 is set at common memory module, the memory module that has been set with microcomputer 300 transmission to control.Same during with above-mentioned face-image memory module, control is delivered to image processor 200 with microcomputer 300 with memory module.
Have, same when cameraman and face-image memory module are selected, the view finder 110 of observation Fig. 1 is partly pressed shutter release button 120, sets focus-lock again.
If the shutter release button 120 half of Fig. 1 is pressed by the cameraman, same during with above-mentioned face-image memory module then in digital camera 100 inside, carry out a series of pre-photograph processing.
Here, the cameraman presses the shutter release button 120 of Fig. 1 entirely.
If shutter release button 120 is pressed entirely, then carry out the formal photograph processing same with above-mentioned face-image memory module, will represent the photographic image data input image processor 200 shown in Figure 2 of photographs.
Below, with select as memory module be the face-image memory module time above-mentioned explanation same, use the flow chart of Fig. 4 to describe.
The measuring ability 201 of image processor 200 shown in Figure 3, when transmitting photographic image data, transmitting with microcomputer 300 from the control of Fig. 2 is the situation (the step S1 of Fig. 4) of common memory module as the memory module setting.Measuring ability 201, though it is same during with above-mentioned face-image memory module, detect blood-shot eye illness and generate the positional information (the step S2 of Fig. 4) of seeing red, but do not carry out facial extraction (the step S3 of Fig. 4) in this example, do not transmit the positional information of photographss and blood-shot eye illness, only transmit the positional information of photographss, blood-shot eye illness and as the common memory module of memory module to debugging functions 203 to recognition function 202.In the flow chart of Fig. 4, enter step S4 from step S3.
In debugging functions 203, according to the positional information of photographic image data of sending here from measuring ability 201 and blood-shot eye illness, carry out with the same photographs of above-mentioned face-image memory module in the correcting process (the step S4 of Fig. 4) of blood-shot eye illness.In addition, in this example, debugging functions 203, not to face-image systematic function 204 transmit representative revised the correction of photographs after the correction of the blood-shot eye illness in the photographs after photographic image data, but directly deliver in the video memory shown in Figure 2 250.
Revise the back photographic image data, be transferred into the storage (the step S8 of Fig. 4) in video memory 250 equally of video memory 250 and face image data, (the step S9 of Fig. 4) carried out in processing as described above repeatedly before photography finishes.
Like this, if utilize the digital camera of present embodiment,, also can store and revise the good-looking image of looking of blood-shot eye illness even under the situation of having selected common memory module.
More than, the explanation end to the 1st execution mode of the present invention then illustrates the 2nd execution mode of the present invention.Though the digital camera of present embodiment has the inscape same with the digital camera 100 of the 1st execution mode, different with the 1st execution mode is the sequential of seeing red detection.Below, the Fig. 1 that uses in the explanation of the 1st execution mode and Fig. 2 here also use, the different aspect of main explanation and the 1st execution mode.
Figure 10 is the flow chart that usefulness is represented a series of photograph processing that the digital camera of present embodiment carries out.
At first, the cameraman moves the digital camera 100 of Fig. 1 towards desirable subject, partly presses shutter release button 120.
If the shutter release button 120 of Fig. 1 is partly pressed, then in the inside of digital camera 100, the 1st stage of the shutter release that the switches set 360 of connection Fig. 2 is comprised.At this moment, provide the indication of luminous quantity, flash light emission portion 340, the flash of light (the step S21 of Figure 10) of from shown in Figure 1 flash light emission device 140 sending indicated luminous quantity with microcomputer 300 to flash light emission portion 340 from control.
If flash, then identical with the 1st execution mode, read object light roughly by CCD211, generate the low photographic image data (the step S22 of Figure 10) of resolution.The image of the photographic image data representative of this low resolution, become with photographs shown in Figure 5 600 same, but the coarse image of image quality.
Detecting under the situation of blood-shot eye illness, entering step S25 from the step S24 of Figure 10.Image processor 200 carries out the processing same with the step S5 of Fig. 4, detects the face (the step S25 of Figure 10) of personage in the image of photographic image data representative of low resolution, generates the positional information that expression detects facial positions.
Control is determined shutter speed and aperture amount by the brightness information that obtains from image processor 200 with microcomputer 300, is identified for allowing the condenser lens position of subject in focus by ranging information.According to the shutter speed of determining like this or aperture amount, condenser lens position, adjust each inscape shown in Figure 2 (the step S26 of Figure 10).
If finish to a series of pre-photograph processing the step S26 from the step S21 of Figure 10, then the shutter release button 120 of Fig. 1 is pressed entirely by the cameraman, carry out with the 1st execution mode that in the step S1 of Fig. 4, illustrates in the same formal photograph processing (the step S27 of Figure 10) of formal photograph processing.
Here, by ranging information or the brightness information that image processor 200 calculates, be the blood-shot eye illness that photographs comprised according to photographic image data representative, detect personage's face, cooperate with this personage's position and the information that calculates.Therefore, by according to this ranging information or brightness information, carry out AF (focus on and adjust), AE (exposure is adjusted), AWB (white-balance correction) etc., high accuracy allows focus aim at the personage who is detected, thereby can photograph with the exposure of the best.In addition, digital camera with can shooting motion detects the personage, in advance for example by following the trail of this personage's who is detected color sequence photography in above-mentioned pre-photograph processing, focus is aimed at this detected personage all the time, can obtain the good animation of image quality.
In addition, in the step S24 of Figure 10, in the photographs that the photographic image data of low resolution is represented, do not detect under the situation of blood-shot eye illness, represent not comprise in this photographs the personage.In this case, entering step S27 from step S24, formally photographs in shutter speed according to the rules, aperture amount or condenser lens position.
More than, next the explanation end to the 2nd execution mode of the present invention illustrates the 3rd execution mode of the present invention.Present embodiment possesses the digital camera 100 same camera heads that constitute that have with the 1st execution mode, is the system that the personage in the photographs of taking with this camera head is carried out individual identification.
Figure 11 is the functional block diagram that is applicable to the personal identification system of the 3rd execution mode of the present invention.
Personal identification system 700 constitutes with the PC 720 that is connected with camera head 710 by having the camera head 710 that constitutes equally with the digital camera 100 of the 1st execution mode shown in Figure 2.In PC 720, possess respectively as the function of seeing red test section 721, face recognition portion 722, individual identification portion 723, storage part 724 and image displaying part 725.This sees red test section 721, is equivalent to an example of test section of the present invention, and facial test section 722 is equivalent to an example of face recognition of the present invention portion.In the storage part 724 in advance with personage's face-image with preserve after relevant this personage's information is associated.
Figure 12 is a series of processing of the personage in the photographs are discerned in expression according to the photographs of taking with camera head a flow chart.
In camera head 710, carry out the same a series of photograph processing of a series of photograph processing (the step S1 of Fig. 4) in the digital camera 100 with the 1st execution mode, generate the photographic image data (the step S31 of Figure 12) of representing photographs.In this camera head 710, same with digital camera 100 shown in Figure 1, have flasher and camera lens and be configured on the very approaching position, make photographs be easy to generate the feature of blood-shot eye illness.The photographic image data that generates in this camera head 710 is sent to blood-shot eye illness test section 721.
In blood-shot eye illness test section 721, according to the same processing of step S23 of step S2 or Figure 10 of Fig. 4, detect the blood-shot eye illness (the step S32 of Figure 12) in the photographs of the photographic image data representative of being sent here.
In photographs, detect under the situation of blood-shot eye illness by blood-shot eye illness test section 721, enter step S34, photographic image data and redeye location information are delivered to face recognition portion 722, the beginning face recognition processing from step S33.In face recognition portion 722, carry out the same processing of step S25 with step S4 or Figure 10 of Fig. 4, detect the face (the step S34 of Figure 12) of the personage in the photographs.Representative detects facial face-image and is sent to individual identification portion 723 in face recognition portion 722.
In the face-image of individual identification portion 723 in being stored in storage part 724, retrieve the identical personage's face-image of personage with appearance with face-image of sending here from face recognition portion 722.Discerning this individual's processing, is the processing of in the past extensively carrying out, and omits its detailed description in this manual.Have, individual identification portion 723 obtains the personage's corresponding with the face-image that retrieves relevant information (the step S35 of Figure 12) again.The personage's relevant information that obtains and personage's face-image are delivered to image displaying part 725, and information that these are relevant with the personage and personage's face-image are displayed on the display frame (not shown).
In addition, in the step S33 of Figure 12, in photographs, do not detect under the situation of blood-shot eye illness, omit step S34 and step S35, do not carry out individual identification and handle.In this case, photographs is delivered to image displaying part 725 from the blood-shot eye illness test section 721 of Figure 11, on the display frame (not shown), show the information of photographs and " can not discern the individual ".
The photographs that obtains according to the photography personage in extensive use in the past carries out in the personal identification system of individual identification, for example, if before personage's photograph or mask be disposed at camera, then just can carry out personage's identification, the personal identification system of might out-tricking according to the photographs of taking these photographs or mask.If personal identification system according to present embodiment, then only the personage in the photographs that produces eyes variable colors such as so-called blood-shot eye illness is carried out individual identification, the personage who takes photograph or mask but do not produce in the photographs of seeing red is not carried out individual identification, output error message.Therefore, for example,, then can constitute firm safety system if the personal identification system of present embodiment is applicable in the safety system.
Here, though above-mentionedly discern facial digital camera, face recognition method and facial extracting method and be illustrated to detecting blood-shot eye illness in the photographs, but camera head of the present invention, face recognition method and facial extracting method also can be to detect the eyes that produce the regulation variable color partly to discern facial device and method in image, for example also can be that the gold eye that detects in the photographs is discerned facial device and method.
In addition, in above-mentioned, though as the example that has been to use digital camera of camera head explanation, camera head of the present invention for example can be to be loaded into minicam on the mobile phone etc.
Have again, in above-mentioned, though to sending visible light by the flash light emission device, detect the example of the eyes in the photographs of photographing with this visible light, but camera head of the present invention, face recognition method and facial extracting method also can be the light that sends from the flash light emission device beyond the visible light such as infrared ray, detect the device and method of the eyes in the photographs of photographing with this light.By taking the photograph subject with the photo-beat beyond the visible light, the eyes in can easier detection photographs.
Claims (5)
1. face recognition method, according to the view data of presentation video, the face in the recognition image is characterized in that, has:
According to described view data, detect the testing process of the eyes part that in image, generates the regulation variable color; With
According to detected eyes part in described testing process, discern the identifying of the face in the described image.
2. face recognition method according to claim 1 is characterized in that, described testing process is the process of the blood-shot eye illness part in the detected image.
3. facial extracting method according to the view data of presentation video, is extracted face in the image to generate face-image, it is characterized in that having:
Testing process according to the part of the blood-shot eye illness in the view data detected image;
According to detected blood-shot eye illness part in described testing process, discern the identifying of the face in the described image;
The makeover process of correction detected blood-shot eye illness part in described testing process; With
From image, extract and in described identifying, identified and in described makeover process, seen red the face that part has been corrected, to generate the face-image generative process of face-image.
4. a camera head is that subject is photographed, and generates the camera head of the photographic image data of expression photographs, it is characterized in that having:
According to described photographic image data, in photographs, detect the test section of the eyes part that generates the regulation variable color;
According to by the detected eyes part of described test section, discern the identification part of the face in the described photographs;
Correction is in the correction portion of the variable color of the detected eyes part of described test section; With
The face that extraction has identified in described identification part and variable color has been corrected in described correction portion from photographs is to generate the face-image generating unit of face-image.
5. camera head according to claim 4 is characterized in that, the blood-shot eye illness part in the described test section detected image, and described correction portion correction is by the detected blood-shot eye illness part of described test section.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2002360611 | 2002-12-12 | ||
| JP2002360611 | 2002-12-12 |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1520161A true CN1520161A (en) | 2004-08-11 |
| CN1270516C CN1270516C (en) | 2006-08-16 |
Family
ID=34308263
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 200310120419 Expired - Fee Related CN1270516C (en) | 2002-12-12 | 2003-12-11 | Face recognition method, face extraction method and pick-up device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1270516C (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100440247C (en) * | 2005-08-04 | 2008-12-03 | 奥林巴斯映像株式会社 | Device and method for image searching |
| CN100448267C (en) * | 2004-02-06 | 2008-12-31 | 株式会社尼康 | Digital camera |
| CN1841187B (en) * | 2005-04-01 | 2010-05-12 | 佳能株式会社 | Image sensing device and control method thereof |
| CN101710383A (en) * | 2009-10-26 | 2010-05-19 | 北京中星微电子有限公司 | Method and device for identity authentication |
| CN101335839B (en) * | 2004-02-06 | 2010-10-13 | 株式会社尼康 | Digital camera |
| CN101526719B (en) * | 2008-03-07 | 2010-12-01 | 欧姆龙株式会社 | Measurement device and method, imaging device |
| CN102063753A (en) * | 2011-01-18 | 2011-05-18 | 王芳 | Method for identification and statistic of visitors based on Internet of things |
| CN103826031A (en) * | 2005-04-29 | 2014-05-28 | 惠普开发有限公司 | Method and apparatus for digital-image red-eye correction that facilitates undo operation |
| CN104798103A (en) * | 2012-11-28 | 2015-07-22 | Nec卡西欧移动通信株式会社 | Facial recognition device, recognition method and program therefor, and information device |
-
2003
- 2003-12-11 CN CN 200310120419 patent/CN1270516C/en not_active Expired - Fee Related
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100448267C (en) * | 2004-02-06 | 2008-12-31 | 株式会社尼康 | Digital camera |
| CN101335839B (en) * | 2004-02-06 | 2010-10-13 | 株式会社尼康 | Digital camera |
| CN1841187B (en) * | 2005-04-01 | 2010-05-12 | 佳能株式会社 | Image sensing device and control method thereof |
| CN103826031A (en) * | 2005-04-29 | 2014-05-28 | 惠普开发有限公司 | Method and apparatus for digital-image red-eye correction that facilitates undo operation |
| CN100440247C (en) * | 2005-08-04 | 2008-12-03 | 奥林巴斯映像株式会社 | Device and method for image searching |
| CN101526719B (en) * | 2008-03-07 | 2010-12-01 | 欧姆龙株式会社 | Measurement device and method, imaging device |
| CN101710383B (en) * | 2009-10-26 | 2015-06-10 | 北京中星微电子有限公司 | Method and device for identity authentication |
| CN101710383A (en) * | 2009-10-26 | 2010-05-19 | 北京中星微电子有限公司 | Method and device for identity authentication |
| CN102063753A (en) * | 2011-01-18 | 2011-05-18 | 王芳 | Method for identification and statistic of visitors based on Internet of things |
| CN104798103A (en) * | 2012-11-28 | 2015-07-22 | Nec卡西欧移动通信株式会社 | Facial recognition device, recognition method and program therefor, and information device |
| US10083344B2 (en) | 2012-11-28 | 2018-09-25 | Nec Corporation | Facial recognition apparatus, recognition method and program therefor, and information device |
| US10303926B2 (en) | 2012-11-28 | 2019-05-28 | Nec Corporation | Decreasing lighting-induced false facial recognition |
| US10614293B2 (en) | 2012-11-28 | 2020-04-07 | Nec Corporation | Facial recognition apparatus, recognition method and program therefor, and information device |
| US10956715B2 (en) | 2012-11-28 | 2021-03-23 | Nec Corporation | Decreasing lighting-induced false facial recognition |
| CN104798103B (en) * | 2012-11-28 | 2021-08-03 | 日本电气株式会社 | Face recognition device, face recognition method, program for the same, and information apparatus |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1270516C (en) | 2006-08-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1220111C (en) | Camera | |
| CN1263286C (en) | Image pickup device | |
| CN1744672A (en) | Image capture apparatus and control method therefor | |
| US20040119851A1 (en) | Face recognition method, face recognition apparatus, face extraction method, and image pickup apparatus | |
| US7894715B2 (en) | Image pickup apparatus, camera system, and control method for image pickup apparatus | |
| CN101399914B (en) | Image capture device and image capture method | |
| CN1648934A (en) | Face detecting apparatus and method | |
| CN1928692A (en) | Camera apparatus having a plurality of image pickup elements | |
| KR20080101780A (en) | Image device, focus control method and storage medium recording a focus control program | |
| CN1917611A (en) | Imaging device and imaging method | |
| JP2007150601A (en) | Electronic camera | |
| JP2007067870A (en) | Digital camera system and calibration method of photographing condition | |
| CN1756309A (en) | Image-capturing device, image-processing device, method for controlling image-capturing device, and program for causing computer to execute the method | |
| CN106210524B (en) | A kind of image pickup method and camera module of camera module | |
| CN1270516C (en) | Face recognition method, face extraction method and pick-up device | |
| CN1311288C (en) | Focusing device, camera and focusing position checking method | |
| CN1514297A (en) | Camera and its range finding method | |
| JP2007067934A (en) | Imaging apparatus and its control method | |
| CN108289170B (en) | Photographing apparatus, method and computer readable medium capable of detecting measurement area | |
| CN1260951C (en) | Camera | |
| JP2007067874A (en) | Digital camera system and calibration method of photographing condition | |
| CN1265620C (en) | Digital camera | |
| CN100592760C (en) | Automatic red eye detection and correction method when photo taking | |
| JP2007005966A (en) | System for calculation amount of exposure, and control method and control program thereof | |
| CN103379273A (en) | Imaging 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 | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| ASS | Succession or assignment of patent right |
Owner name: FUJI PHOTO FILM CO., LTD. Free format text: FORMER OWNER: FUJIFILM HOLDINGS CORP. Effective date: 20070803 |
|
| C41 | Transfer of patent application or patent right or utility model | ||
| C56 | Change in the name or address of the patentee |
Owner name: FUJIFILM HOLDINGS CORP. Free format text: FORMER NAME OR ADDRESS: FUJI PHOTO FILM CO., LTD. |
|
| CP01 | Change in the name or title of a patent holder |
Address after: Tokyo, Japan Patentee after: Fujifilm Corp. Address before: Kanagawa Patentee before: FUJIFILM Corp. |
|
| TR01 | Transfer of patent right |
Effective date of registration: 20070803 Address after: Tokyo, Japan Patentee after: FUJIFILM Corp. Address before: Tokyo, Japan Patentee before: Fujifilm Corp. |
|
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20060816 Termination date: 20181211 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |