CN101296386B - Picture processing apparatus and method - Google Patents
Picture processing apparatus and method Download PDFInfo
- Publication number
- CN101296386B CN101296386B CN200810095922XA CN200810095922A CN101296386B CN 101296386 B CN101296386 B CN 101296386B CN 200810095922X A CN200810095922X A CN 200810095922XA CN 200810095922 A CN200810095922 A CN 200810095922A CN 101296386 B CN101296386 B CN 101296386B
- Authority
- CN
- China
- Prior art keywords
- image
- mentioned
- downscaled images
- identified
- view image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Images
Landscapes
- Image Processing (AREA)
- Image Generation (AREA)
- Controls And Circuits For Display Device (AREA)
Abstract
The present invention realizes picture recording in simple configuration such that a user can easily and intuitively identify a minified picture of a multi-view picture from among minified pictures of pictures being taken. A 3-dimensional thumbnail picture is created by performing 3-dimensional CG processing on a multi-view picture thumbnail picture, and a thumbnail picture is created from a normal picture and recorded in association with a primary picture. The 3-dimensional CG thumbnail picture X is given a stereoscopic effect as a result of the 3-dimensional CG graphic processing so that the difference from a normal 2-dimensional thumbnail Y is clear at a glance. Consequently, a viewer can readily recognize the picture X as a minified picture of a multi-view picture without using a special apparatus to realize a graphical approach.
Description
Technical field
The guide look that the present invention relates to the downscaled images of multi-view image and single view image shows.
Background technology
General stereo image display system is that right eye is shown with the right eye is viewpoint and image represented, and it is viewpoint and image represented is shown separately that left eye is shown with the left eye.Usually, such be in 2 images (following note is made multi-view image) of viewpoint with right eye and left eye, be in from the object of viewpoint position far away because, promptly watch mode identical, so be depicted in roughly the same position from two sight line almost parallel.On the other hand, be in from the object of the near position of viewpoint, promptly watch the mode difference, so be depicted in the position of staggering mutually because not parallel from two sight line.When watching this image (following note is made anaglyph) that parallax arranged respectively with the right eye left eye, the observer can obtain far and near sense to object.
Aspect the generation of anaglyph, there are about employing 2 camera heads take and are replicated body, utilize the illusion of eye and the realistic gimmick of the anaglyph that is obtained is showed on 3 dimension ground, and adopt computer graphics techniques by with about two 3 dimension computer graphicals (3 dimension CG) that are the visual field conversion of the model data of viewpoint generates anaglyph show.3 dimension Computer Graphic Demonstrations are commonly called as imagination 3 dimension demonstrations, doubtful property 3 dimensions show.
Realistic gimmick adopts the 3D monitor of parallax barrier bar or cylindrical lens formula to carry out the stereo display of anaglyph.The parallax barrier bar is to produce by light transmissive portion and light shielding portion alternately to arrange the parallax barrier that figure was constituted that forms with given pitch on the parallax barrier display layer, and on the picture display face of its lower floor, alternately arrange, show the image segments of the strip of the image about expression, thereby make the stereos copic viewing of anaglyph become possibility.
Perhaps, realistic gimmick can adopt the cylindrical lens mode of utilizing the cylindrical lens sheet material, the integrated graphics mode of utilizing the microlens array sheet material, the holographic mode of utilizing interference etc.
General computer graphical is to show the object that defines in 3 dimension spaces by a plurality of polygonal set.As polygon, the simplest triangle of many usually employings.Each polygon that constitutes object is to adopt mathematics gimmicks such as vector or matrix, is that 3 dimension coordinates of benchmark are that visual field coordinate carries out coordinate transform from the object coordinate that himself has to the position with observer's viewpoint.And, be used for the polygon with the performance of visual field coordinate is carried out to 2 dimensional planes the coordinate transform (perspective transform) of perspective projection in order to be shown on the display unit.Final polygonal position data after this a succession of coordinate transform is carried out just becomes the Z value of expression distance of (Z axle) from viewpoint to the direction vertical with display and the X value and the Y value of the position coordinates in the expression display face.
Aspect the determining method of the colour that constitutes each polygonal pixel, the light source that configuration imagination in the coordinate of the visual field is arranged, physically calculate from the light intensity of this light source to polygonal each summit irradiation, the method for the light intensity on each location of pixels in the rasterizing process in polygon being obtained by linear interpolation according to these values again.
Perhaps, shown in patent documentation 1, the method for the texture of pasting the expression body surface that is called texture, special images such as apperance is arranged on polygon surface.In this occasion, in the stage of the object coordinate before coordinate transform,, except position coordinates, revest texture coordinate to constituting polygonal each summit of object.And same with the occasion of above-mentioned luminous intensity, in polygonal rasterizing process, carry out linear interpolation by each location of pixels one a pair of texture coordinate, decide the colour of pixel according to the texture coordinate of being obtained.
According to just generating the graph image of seeing from specific viewpoint with upper type, but, in order to generate the anaglyph of aforesaid stereos copic viewing, the position of viewpoint is changed by right eye and left eye position separately, same object is carried out 2 times describe, thereby generate multi-view image.In the high price system, possess the display unit that can realize aforesaid realistic gimmick, generate multi-view image.
On the other hand, in the past, left occasion in the storage device in that multi-view image and parallax free common image are mixed, the technology of the downscaled images that shows them was diacritically arranged.The electronic equipment of patent documentation 2 possesses the display unit that can realize realistic gimmick, when the guide look of the flatness of thumbnail image shows, adds for the pairing thumbnail image of 3 d image
The mark of " 3D ", and indicate with circle or ellipse, add the mark of " 2D " for 2 dimension images pairing thumbnail images, indicate with triangle, thereby can distinguish thumbnail that becomes according to 2 dimension image makings and the thumbnail that is made into according to 3 d image.
Patent documentation 1: the spy opens the 2002-24850 communique
Patent documentation 2: the spy opens the 2004-120165 communique
Summary of the invention
The problem that solution is planned in invention
In patent documentation 2, be the thumbnail of indicating 2 dimension images and 3 d image with the such mark of " 3D ", " 2D " and circle, ellipse, the such figure of triangle, but, so then both are difficult to difference on directly perceived.
Also have, in patent documentation 2, can realize that the configuration of display unit of the demonstration of the formed 3 d image of realistic gimmick is necessary, and the demonstrations of 2 dimension images and 3 d image to switch be necessary, so the shortcoming that exists the cost of entire system to uprise.
Therefore, the present invention realizes that with easy formation the user can discern the device of the representative image of multi-view image easily and intuitively from the representative image of photographic images.
Be used to solve the scheme of problem
Image processing apparatus involved in the present invention possesses: image input part, input comprise the multi-view image that photographs from the viewpoint more than 2 and the image of the single view image that photographs from single viewpoint; Storage part, the image of having imported in the memory image input part; Downscaled images preparing department makes the 1st downscaled images of dwindling the image gained of storing in the storage part; Image recognition portion, the image of storing in the identification storage part is multi-view image or single view image; Image processing part is implemented 3 dimension computer graphicals to the 1st downscaled images corresponding with the image that is identified as multi-view image by image recognition portion and is handled, thereby makes the 2nd downscaled images; And display part, according to the indication of the image of storing in the storage part being had a guide look of demonstration, guide look shows the 1st downscaled images that is made into by downscaled images preparing department as the representative image that is identified as the image of single view image by image recognition portion, and guide look shows that the 2nd downscaled images that is made into by image processing part is as the representative image that is identified as the image of multi-view image by image recognition portion.
According to the present invention, guide look shows that having implemented 3 ties up the representative image of the downscaled images of computer graphicals processing as multi-view image, and guide look shows the representative image of simple downscaled images as the single view image.The representative image of multi-view image has been carried out 3 dimension computer graphicals handled, so it is very clear with the difference of simple downscaled images.Also having, do not need to use special realistic gimmick for the representative image that shows multi-view image, is cheap.
Image processing apparatus involved in the present invention possesses: image input part, input comprise the multi-view image that photographs from the viewpoint more than 2 and the image of the single view image that photographs from single viewpoint; Storage part, the image of having imported in the memory image input part; Downscaled images preparing department makes the downscaled images of dwindling the image gained of storing in the storage part; Image recognition portion, the image of storing in the identification storage part is multi-view image or single view image; The face test section for the image that is identified as multi-view image by image recognition portion, detects and is replicated having or not of body face zone; And display part, according to the indication of the image of storing in the storage part being had a guide look of demonstration, guide look shows that downscaled images is identified as multi-view image and is the representative image that is not detected the image that is replicated body face zone by the face test section as the downscaled images of the image that is identified as the single view image by image recognition portion and by image recognition portion, and guide look shows and implemented given sample downscaled images that 3 dimension computer graphicals handle as being identified as multi-view image by image recognition portion and being the representative image that is detected the image that is replicated body face zone by the face test section.
According to the present invention, guide look shows has implemented given sample images that 3 dimension computer graphicals handle as the representative image that detects the multi-view image of face, and guide look shows that simple downscaled images is as the single view image or do not detect the representative image of the multi-view image of face.The given sample of the representative image that becomes multi-view image has been carried out 3 dimension computer graphicals handled, so the difference of the representative image of the multi-view image of having the face and simple downscaled images comes into plain view.Also have,, handle so need not carry out 3 dimension computer graphicals to each multi-view image for the representative image that shows multi-view image because use given sample.
Image processing apparatus involved in the present invention possesses: image input part, input comprise the multi-view image that photographs from the viewpoint more than 2 and the image of the single view image that photographs from single viewpoint; Storage part, the image of having imported in the memory image input part; Downscaled images preparing department makes the downscaled images of dwindling the image gained of storing in the storage part; Image recognition portion, the image of storing in the identification storage part is multi-view image or single view image; The flash light emission identification part, for the image that is identified as multi-view image by image recognition portion, the having or not of the flash light emission when identification is taken; And display part, according to the indication of the image of storing in the storage part being had a guide look of demonstration, guide look shows that downscaled images is as the representative image that is identified as the image of single view image by image recognition portion, the guide look demonstration shows that as being identified as multi-view image by image recognition portion and being the representative image that is identified as the image that does not have flash light emission by the flash light emission identification part, having a guide look of the given sample downscaled images of having implemented 3 dimension computer graphicals processing is as being identified as multi-view image by image recognition portion and being the representative image that is identified as the image of flash light emission by the flash light emission identification part by the downscaled images that downscaled images preparing department is made into.
According to the present invention, guide look shows has implemented the representative image that downscaled images that 3 dimension computer graphicals handle has the multi-view image of flash light emission when taking, and guide look shows that simple downscaled images does not have the representative image of the multi-view image of flash light emission as the single view image or when taking.There is when taking the representative image of the multi-view image of flash light emission to carry out 3 dimension computer graphicals and handles, so it does not have the difference of representative image of multi-view image of flash light emission very clear with simple downscaled images or when taking.Also having, do not need to use special realistic gimmick for the representative image that shows multi-view image, is cheap.
Also possess and from the image that is identified as multi-view image by image recognition portion, detect the face test section that has or not that is replicated body face zone, image processing part has been set the 1st downscaled images and has been replicated after the corresponding zone, body face zone by the face test section is detected, 3 dimension computer graphicals are carried out in the zone of having set to be handled, thereby make the 2nd downscaled images, the display part guide look shows the 1st downscaled images as the representative image that is not detected the image that is replicated body face zone by the face test section in the image that is identified as multi-view image by image recognition portion, and guide look shows that the 2nd downscaled images is as the representative image that is detected the image that is replicated body face zone by the face test section.
Be only limited to the such pith of face is implemented image processing, can substantial doubtful stereo-picture be shown as representative image by processing seldom.
Image processing apparatus involved in the present invention possesses: image input part, input comprise the multi-view image that photographs from the viewpoint more than 2 and the image of the single view image that photographs from single viewpoint; Storage part, the image of having imported in the memory image input part; Downscaled images preparing department makes the 1st downscaled images of dwindling the image gained of storing in the storage part; Image recognition portion, the image of storing in the identification storage part is multi-view image or single view image; Apart from the identification part, identification is at a distance still closely by the body that is replicated that image recognition portion is identified as in the image of multi-view image; Image processing part to according to being identified as multi-view image by image recognition portion and being to be identified as by the distance identification part to be replicated body and to be in the 1st downscaled images that in-plant image is made into and to implement 3 dimension computer graphicals and handle, thereby is made the 2nd downscaled images; And display part, according to the indication of the image of storing in the storage part being had a guide look of demonstration, guide look shows that the 1st downscaled images that is made into by downscaled images preparing department is as the representative image that is identified as the image of single view image by image recognition portion, guide look shows the 1st downscaled images that is made into by downscaled images preparing department as being identified as multi-view image by image recognition portion and being to be identified as by the distance identification part to be replicated the representative image that body is in remote image, and guide look shows that the 2nd downscaled images that is made into by image processing part is as being identified as multi-view image by image recognition portion and being to be identified as by the distance identification part to be replicated the representative image that body is in in-plant image.
According to the present invention, only be in in-plant multi-view image and implement 3 dimension computer graphicals and handle, so need not be in remote and implement the also insignificant image of this processing and carry out this processing to being replicated body to being replicated body.
Image processing method involved in the present invention may further comprise the steps: input comprises the multi-view image that photographs from the viewpoint more than 2 and the image of the single view image that photographs from single viewpoint; The image that storage has been imported; The 1st downscaled images of the image gained of being stored is dwindled in making; The image that identification is stored is multi-view image or single view image; 1st downscaled images corresponding with the image that is identified as multi-view image implemented 3 dimension computer graphicals handle, thereby make the 2nd downscaled images; And according to the indication of the image of being stored being had a guide look of demonstration, guide look shows the 1st downscaled images as the representative image that is identified as the image of single view image, and guide look shows that the 2nd downscaled images is as the representative image that is identified as the image of multi-view image.
Image processing method involved in the present invention may further comprise the steps: input comprises the multi-view image that photographs from the viewpoint more than 2 and the image of the single view image that photographs from single viewpoint; The image that storage has been imported; The downscaled images of the image gained of being stored is dwindled in making; The image that identification is stored is multi-view image or single view image; For the image that is identified as multi-view image, detect and be replicated having or not of body face zone; And according to the indication of the image of being stored being had a guide look of demonstration, guide look shows downscaled images as the downscaled images of the image that is identified as the single view image and be identified as multi-view image and be the representative image that does not detect the image that is replicated body face zone, and guide look shows and implemented given sample downscaled images that 3 dimension computer graphicals handle as being identified as multi-view image and being the representative image that detects the image that is replicated body face zone.
Image processing method involved in the present invention may further comprise the steps: input comprises the multi-view image that photographs from the viewpoint more than 2 and the image of the single view image that photographs from single viewpoint; The image that storage has been imported; The downscaled images of the image gained of being stored is dwindled in making; The image that identification is stored is multi-view image or single view image; For the image that is identified as multi-view image, the having or not of the flash light emission when identification is taken; And according to the indication of the image of being stored being had a guide look of demonstration, guide look shows that downscaled images is as the representative image that is identified as the image of single view image, guide look demonstration downscaled images is as being identified as multi-view image and being the representative image that is identified as the image that does not have flash light emission, and guide look shows that the given sample downscaled images of having implemented 3 dimension computer graphicals processing is as being identified as multi-view image and being the representative image that is identified as the image of flash light emission.
Further comprising the steps of: as from the image that is identified as multi-view image, to detect and be replicated having or not of body face zone; The 1st downscaled images set after the zone corresponding with being replicated body face zone, 3 dimension computer graphicals processing have been carried out in the zone of having set, thereby made the 2nd downscaled images; And have a guide look of demonstration the 1st downscaled images as the representative image that does not detect the image that is replicated body face zone in the image that is identified as multi-view image, and guide look shows that the conduct of the 2nd downscaled images detects the representative image of the image that is replicated body face zone.
Image processing method involved in the present invention may further comprise the steps: input comprises the multi-view image that photographs from the viewpoint more than 2 and the image of the single view image that photographs from single viewpoint; The image that storage has been imported; The 1st downscaled images of the image gained of being stored is dwindled in making; The image that identification is stored is multi-view image or single view image; The body that is replicated that is identified as in the image of multi-view image is at a distance still closely; To according to being identified as multi-view image and being to be identified as to be replicated body and to be in the 1st downscaled images that in-plant image is made into and to implement 3 dimension computer graphicals and handle, thereby make the 2nd downscaled images; And according to the indication of the image of being stored being had a guide look of demonstration, guide look shows that the 1st downscaled images is as the representative image that is identified as the image of single view image, guide look shows the 1st downscaled images as being identified as multi-view image and being to be identified as to be replicated the representative image that body is in remote image, and guide look shows that the 2nd downscaled images is as being identified as multi-view image and being to be identified as to be replicated the representative image that body is in in-plant image.
The program that makes computer carry out this image processing method is also contained among the present invention.
The invention effect
According to the present invention, guide look shows has implemented downscaled images that 3 dimension computer graphicals the handle representative image as multi-view image, guide look show simple downscaled images as the single view image, do not detect the multi-view image of face or be replicated the representative image that body is in remote multi-view image.The given sample of the representative image that becomes multi-view image has been carried out 3 dimension computer graphicals handled, so it is very clear with the difference of simple downscaled images to have the representative image of multi-view image of face.Also having, do not need to use special realistic gimmick for the representative image that shows multi-view image, is cheap.
Description of drawings
Fig. 1 is the block diagram of camera head.
Fig. 2 is that the related general chart picture of the 1st execution mode shows the flow chart of handling.
Fig. 3 is the figure of an example of expression texture mapping.
Fig. 4 is the figure of expression example of thumbnail images shown by the related general chart picture demonstration processing of the 1st execution mode.
Fig. 5 is that the related general chart picture of the 2nd execution mode shows the flow chart of handling.
Fig. 6 is the figure of expression example of thumbnail images shown by the related general chart picture demonstration processing of the 2nd execution mode.
Fig. 7 is that the related general chart picture of the 3rd execution mode shows the flow chart of handling.
Fig. 8 is the figure of expression example of thumbnail images shown by the related general chart picture demonstration processing of the 3rd execution mode.
Fig. 9 is that the related general chart picture of the 4th execution mode shows the flow chart of handling.
Figure 10 is the figure of expression example of thumbnail images shown by the related general chart picture demonstration processing of the 4th execution mode.
Figure 11 is that the related general chart picture of the 5th execution mode shows the flow chart of handling.
Figure 12 is the figure of expression example of thumbnail images shown by the related general chart picture demonstration processing of the 5th execution mode.
Figure 13 is the block diagram of personal computer.
Symbol description
1: camera head, 2a: the 1st pinch of picture portion, 2b: the 2nd pinch of picture portion, 10:LCD, 62: host CPU, 38: storage card, 100: personal computer, 102:CPU, 108: hard disk unit
Embodiment
＜the 1 execution mode 〉
Fig. 1 represents that the electricity of compound eye imaging device 1 constitutes.Arrange setting as lens axis L1, the L2 of the 2a of portion, 2b by parallel mode or mode with given angle for the 1st and the 2nd pinch.
The 1st pinch is made of the 1st zoom lens 11a, the 1st aperture 12a, the 1st condenser lens 13a and the 1st imageing sensor 14a that arrange along lens axis L1 as the 2a of portion.The 1st aperture 12a is connected with aperture control part 16a, also has, and the 1st imageing sensor 14a is connected with timing generator (TG) 18a.The action of the 1st aperture 12a, the 1st condenser lens 13a is by photometry range finding CPU19a control.The action of TG18a is by host CPU 62 controls.
Be provided with the operating portion 70 that is used for when the user uses this camera head 1, carrying out all operations in the camera head 1.The cross key of this operating portion 70 possesses mains switch that the power connection that is used to make camera head 1 action uses, be used to select to automatically snap, manually take etc. pattern dish, the setting that is used to carry out various menus, selection or zoom, flash light emission are with switch and be used to carry out the information position assignment key of the execution, cancellation etc. of the menu selected with cross key.Just can carry out switching, zoom of the connection/cut-out of power supply, various pattern (screening-mode, replay mode etc.) etc. by proper handling to operating portion 70.
Also have, possess the driver 148 of host CPU 62, EEPROM146, YC/RGB transformation component 147 and demonstration usefulness in the camera head 1.Host CPU 62 carries out the control of this camera head 1 integral body.Deposit the intrinsic curing data of this camera head 1, program etc. among the EEPROM146.
YC/RGB transformation component 147 shows LCD10 output to rgb signal that the colour-video signal YC that by YC handling part 35a35b generated be transformed to 3 looks via the driver 148 that shows usefulness to image with it.
According to the zoom operation from input operation part 9, the 1st zoom lens 11a moves to NEAR side (stretching out side) or INF side (retraction side) along lens axis L1, changes zoom ratio.Should move by not shown motor driven.Aperture 12a changes opening value (f-number) and the restriction luminous flux when AE (Auto Exposure) moves, adjustment exposes.Condenser lens 13a when AF (Auto Focus) moves along lens axis L1 to NEAR side or INF side shifting and change focal position, carry out the focus adjustment.Should move by not shown motor driven.
Detecting static picture with half during by state of release-push, host CPU 62 obtains ranging data from distance with imaging apparatus 51a51b respectively.Host CPU 62 carries out the adjustment of focus, aperture etc. based on the ranging data that is obtained.
The 1st imageing sensor 14a accept by the 1st zoom lens 11a, the 1st aperture 12a and the 1st condenser lens 13a imaging gained be replicated body light, on photo detector, put aside optical charge according to light income.The optical charge savings of the 1st imageing sensor 14a passes on action by TG18a control, by timing signal (clock pulse) the decision electronic shutter speed (optical charge memory time) from the TG18a input.The 1st imageing sensor 14a obtains the picture signal of the amount of 1 picture by each period demand when screening-mode.
The 2nd pinch is and the 1st pinch of identical formation of the 2a of picture portion to reach the 2nd imageing sensor 14b that is connected with timing generator (TG) 14b by the 2nd zoom lens 11b, the 2nd aperture 12b, the 2nd condenser lens 13b and constitute as the 2b of portion.Their action is by host CPU 62 controls.The 1st pinch is interlock and moving as the 2a of portion and the 2nd pinch as the 2b of portion basically, but, also can make its action individually separately.In addition, as the 1st and the 2nd imageing sensor 14a14b, can adopt CCD type, cmos type imageing sensor.
Be input to respectively the A/D converter 30a30B from pinch image signal of the 1st and the 2nd imageing sensor 14a14b output.A/D converter 30a30b is numeral to the view data of being imported from analog converting.By A/D converter 30a30b, pinch image signal of the 1st imageing sensor 14a is exported as the 1st view data (right eye view data), pinch image signal of the 2nd imageing sensor 14b is exported as the 2nd view data (left eye view data).
Imaging signal processing circuit 31a31b is respectively to implementing various image processing such as tonal gradation conversion, white balance revisal, γ revisal processing from the 1st and the 2nd view data of A/D converter 30a31b input.Buffer storage 32a32b deposits the 1st and the 2nd view data of having been implemented various image processing by imaging signal processing circuit 31a31b temporarily.
Photometry range finding CPU19a19b calculates AF evaluation of estimate and AE evaluation of estimate one by one according to the 1st and the 2nd view data of depositing among the buffer storage 32a32b respectively.The AF evaluation of estimate is that the region-wide or given area (for example central portion) at each view data is added up the radio-frequency component of brightness value to calculate the definition of presentation video.The radio-frequency component of brightness value be in the given area the thing of the luminance difference between pixel adjacent to each other (contrast) addition gained.Also have, the AE evaluation of estimate is that the region-wide or given area (for example central portion) at each view data is added up brightness value to calculate the brightness of presentation video.Use in AF action that AF evaluation of estimate and AE evaluation of estimate are carried out when shooting described later prepares to handle respectively and the AE action.
The view data of storing among the buffer storage 32a 32b is transformed to luminance signal (Y-signal) and color difference signal (Cr, Cb signal) by YC handling part 35a 35b respectively, and accepts given processing such as gamma correction.YC signal after the processing leaves in respectively among the working storage 128a128b.
The YC signal of the 1st and the 2nd view data of storing among the working storage 128a128b is read in the YC/RGB transformation component 147 by controller 34 respectively.YC/RGB transformation component 147 the YC signal transformation of the 1st and the 2nd view data be given way vision signal (for example composite color video signal of NTSC mode) afterwards, it synthesized be used to carry out the stereoscopic image data that image shows the stereo display on the LCD10.LCD10 is presented on the LCD10 as the live view image by lcd driver 148 by YC/RGB transformation component 147 synthetic stereoscopic image datas as under the situation of electronic viewfinder when screening-mode.
Compressed and decompressed treatment circuit 36a36b is the 1st and the 2nd view data to storing among the working storage 128a128b respectively, if static picture is then according to JPEG, if animation is then according to MPEG2, MPEG4, H.264 given compressed format such as mode is implemented compression and handled.Medium controller 37 is being compressed each Imagery Data Recording after the processing by compressed and decompressed treatment circuit 36a36 in the storage card 38 that connects via I/F39 and other recording mediums.
Resetting and to be presented at occasion on the LCD10 being recorded in the 1st and the 2nd view data in the storage card 38 like this, each view data of storage card 38 is read by medium controller 37, carry out decompression by compressed and decompressed treatment circuit 36a36, be transformed to after the stereoscopic image data by YC/RGB transformation component 147, be presented on the LCD 10 as reproduced picture by lcd driver 148.
LCD10 is the common monitor that does not adopt realistic gimmick, as electronic viewfinder, carries out the demonstration of the view data that obtains by shooting when image replaying when image taking.
The action of host CPU 62 unified control compound eye imaging device 1 integral body.Host CPU 62 is except with release-push 5a5b, operating portion 70 are connected, and also is connected with EEPROM 146 as nonvolatile memory.Depositing program, set information of various control usefulness etc. among the EEPROM 146.Host CPU 62 is based on this program, set information and carry out various processing.
Also have, host CPU 62 is connected with optical system control instruction unit 127, changes the shooting multiplying power of the 1st crowd of 2a of picture portion, the 2nd crowd of 2b of picture portion according to optical system is controlled the zoom operation of instruction unit 127.
Release-push 5a is 2 grades of switch structures of pushing.In screening-mode, if push (half presses) release-push 5a very lightly, then can carry out AF action and AE action, prepare to handle thereby take.Under this state,, then can carry out photographing process, the 1st and the 2nd view data of the amount of 1 picture is passed on, is recorded in the storage card 38 from frame memory 32 if further push (complete pressing) release-push 5a heavily.
The AF action is by host CPU 62 control the 1st and the 2nd condenser lens 13a13b it to be moved respectively at assigned direction on one side, Yi Bian obtain the maximum of the AF evaluation of estimate of calculating respectively according to the 1st and the 2nd view data that obtains successively, thereby carry out.The AE action is after the AF action is finished, and sets the opening value of the 1st and the 2nd aperture 12a12b and the electronic shutter speed of the 1st and the 2nd imageing sensor 14a14b based on the AE evaluation of estimate of being calculated, thereby carry out.
Also have, possess the operation LCD demonstration 103 that is used to carry out auxiliary operation in the camera head 1.
Also have, camera head 1 is power-supply battery 68 removable formations.Power-supply battery 68 is by chargeable secondary cell, and for example Ni-Cr battery, Ni-MH battery, lithium ion battery constitute.Power-supply battery 68 also can be by the primary cell of disposable type, and for example lithium battery, alkaline cell constitute.Power-supply battery 68 is contained in the not shown battery outer chamber, thereby is electrically connected with each circuit of camera head 1.
The 1st pinch is detected the 1st pinch of lens interval/convergence angle memory circuit 6 as the 2a of portion and the 2nd crowd of interval as interval/convergence angle that the 2b of portion constituted/convergence angle testing circuit 4a4b, memory gap/detected convergence angle of convergence angle testing circuit 4a4b as possessing respectively among the 2b of portion, carries out stretching and changing converge like the spokes of a wheel at the hub interval/convergence angle drive circuit 3a3b by rotation optical axis such as CD-ROM drive motor of base length as the 2a of portion, the 2nd crowd.
Also have, possess in the camera head 1 based on the detected interval/convergence angle of interval/convergence angle testing circuit 4a4b come the control interval/interval/convergence angle control circuit 5 of variation of convergence angle of convergence angle drive circuit 3a3b.
In order to make photoflash lamp 44a44b luminous and accept to supply with from the electric power of power-supply battery 68, the capacitor that not shown flash light emission is used charges the charging light emitting control 138a138b of portion respectively, and control photoflash lamp 44a44b's is luminous.
The charging light emitting control 138a138b of portion controls, make be taken into from host CPU 62, photometry range finding CPU19a19b release-push 5a5b half by under the situation by the signal of various signals such as operation signal, expression luminous quantity, luminous timing entirely, correspondingly carry out electric current supply control, obtain the luminous quantity of hope in the timing of hope photoflash lamp 44a44b.
It is the buttons that are used for indicating which kind of that take the photograph with vertical shooting and shake-hands grip to take that switching push button 40 is taken the photograph in vertical shootings/shake-hands grip.Vertical shootings/shake-hands grip is taken the photograph testing circuit 166 and is detected in taking the photograph with vertical shooting and shake-hands grip which kind of according to the state of this button and take.
Setting expression on 2D/3D mode switch sign 168 is the 2D pattern or the sign of 3D pattern.
Each freely is used for constituting as the 2a of portion and the 2nd pinch of same light-emitting diode (LED) that is replicated body irradiation focal point that is captured as the 2b of portion to the 1st pinch distance with light-emitting component 52a, 52b.
Distance is respectively to obtain the imaging apparatus that the range finding special use that is replicated body image of focal point is arranged with light-emitting component 52a, 52b irradiation by distance with imaging apparatus 51a, 51b.
Distance makes distance with the luminous timing of light-emitting component 52a, 52b with apart from using imaging apparatus 53a, the synchronous control of 53b with driving/control circuit 54.
Be transformed to DID by range finding with A/D transformation component 55a, 55b by distance respectively with the analog picture signal that scooping up of imaging apparatus 53a, 53b obtains as action, output in the range information treatment circuit 57.
Range information treatment circuit 57 is calculated the 1st pinch of distance that is replicated body that is captured as the 2b of portion as the 2a of portion and the 2nd pinch according to the DID of being imported.This is based on the principle of so-called range of triangle.The range information that range information treatment circuit 57 is calculated is stored in the range information memory circuit 58.
Detect in the view data of depositing in the view data that face test section 150 is deposited from buffering memory 32a or buffer storage 32b, the buffer storage 42 and comprise that to become the zone that the character facial that is replicated body divides be the face zone.
There is no particular limitation for the detection method in face zone, can adopt the whole bag of tricks, for example can be suitable for the applicant and open the technology that discloses in the flat 9-101579 communique the spy.This technology is whether the form and aspect of judging each pixel of the image that photographs are included in the scope of the colour of skin, it is cut apart by area of skin color and non-area of skin color, and classify each position in the image in the edge in the detected image by edge part or non-edge part.And, extraction by be positioned at area of skin color, constitute by the pixel of non-edge part classification and by the pixel that is judged as the edge part round the zone as the face candidate region, judge that the face candidate region of extracting is the zone of working as with personage's appearance, it is detected as the face zone based on this judged result.Also have, in addition, also can adopt the spy to open the 2003-209683 communique, the spy opens the method for putting down in writing in the 2002-199221 communique and detects the face zone.
In addition, multi-view image is not to obtain with above-mentioned compound eye imaging device 1, can obtain by the continuous shooting under the swing anaglyph that uses simple eye camera head yet.
Below, with reference to the flow chart of Fig. 3, illustrate that the general chart picture of being carried out by host CPU 62 shows the flow process of handling.The program of stipulating this processing leaves among the EEPROM146.According to indication, all multi-view image unifications of depositing in the storage card 38 are carried out from " thumbnail image guide look show " of operating portion 70.
In S1, read in the catalogue of storage card 38.
In S2, judge whether image file exists in this catalogue.The occasion that exists at image file enters S3.In the non-existent occasion end process of image file.
In S3, read the title division of image file.Then, judging whether to detect image file is the identifying information of having deposited the meaning of multi-view image.Enter S5 in the occasion of having deposited multi-view image, enter S8 in the occasion of having deposited common image.
In S5,, make thumbnail image according to the multi-view image of depositing in the image file.Its concrete gimmick is identical with the paragraph 0040 of patent documentation 2, can only make the left eye image according to the mode of rejecting data at interval, only makes the right eye image, perhaps makes the two.
In S6, the thumbnail image of made is implemented 3 dimension CG handle, thereby make 3 dimension thumbnail images.For example shown in Figure 3, carry out with image in mainly be replicated body and other the various objects that exist (Fig. 3 has represented teacup, but, also can be personage etc.) the polygon PG of corresponding 3 dimension shapes go up the texture mapping of the corresponding region of pasting thumbnail image th, obtain 3 dimension CG thumbnail image X.Its concrete gimmick is identical with patent documentation 1.Perhaps, also can implement 3 dimension CG and handle, it be rejected, thereby make 3 dimension CG thumbnail images, but, consider, preferably make thumbnail earlier from the load aspect of handling at original multi-view image.
In addition, also can be applicable to that thumbnail image makes 3 dimension CG thumbnail images to other 3D computer graphical gimmicks.
In S7, deposit the 3 dimension CG thumbnail images that are made in the guide look frame buffer of in buffer storage 41, guaranteeing.
In S8,, make thumbnail image according to the common image of the single view of depositing in the image file.Its concrete gimmick is identical with common thumbnail image, the single view image is rejected to make get final product.This thumbnail is called common thumbnail image.
In S9, deposit the common thumbnail image that is made in the guide look frame buffer of in buffer storage 41, guaranteeing.
In S10, judge whether to have made 3 dimension CG thumbnail images or common thumbnail image at all images of storage card 38.Become to enter S11 in the occasion of having made, get back to S3 making unclosed occasion, continue to make.
In S11,3 dimension CG thumbnail image and common thumbnail images of guide look display buffer memory 41 on LCD10.Name order, record date time sequencing that its ordering is made as the image file on the making basis of thumbnail image get final product.That is, 3 dimension CG thumbnail images and common thumbnail images are not made any distinction between, 3 dimension CG thumbnail images and the common thumbnail image arrangement that mixes in the picture of LCD10.
In addition, the processing of S3～S9 also can be carried out when each new record image.
Fig. 4 is illustrated in LCD10 and goes up 3 dimension CG thumbnail image X of guide look demonstration and the example of common thumbnail image Y.3 dimension CG thumbnail image X are implemented 3 dimension CG graphics process and given stereoeffect, and it knows at a glance that with the difference of common thumbnail Y even do not adopt the special device of realizing realistic gimmick, the beholder also can easily be familiar with the downscaled images of multi-view image.
＜the 2 execution mode 〉
Fig. 5 is that the general chart picture that expression is carried out by host CPU 62 shows another example of handling.The transposing of the thumbnail image of the multi-view image that has duplicated face is special sample here.
S21～S24 is identical with S1～S4.
In S25, face test section 150 is sent instruction, make and from side of multi-view image or both sides, to extract face.Face test section 150 is according to the instruction from host CPU 62, and the left eye of depositing from image file is used in the image with image or right eye, attempts the detection in face zone.
In S26, judge whether face test section 150 detects the face zone.Enter S27 in the occasion that detects the face zone, enter S29 in the occasion that does not detect the face zone.
In S27, from EEPROM146, take out the 3 dimension thumbnail images that the multi-view image that has duplicated face carried out the sample of presentation.
In S28, in the guide look frame buffer, deposit the sample that is taken out.3 dimension thumbnail images of sample be not must with the appearance that is replicated body of reality with, so long as the personage's of sample face gets final product, and given the 3D figure formed stereoeffect in advance to its face, health etc.
In S29, identical with above-mentioned S5.S30 deposits the thumbnail image that is made among the S29 in the guide look frame buffer.This thumbnail image is called the stereo-picture downscaled images, is distinguished with common thumbnail image.
S31～S34 is identical with above-mentioned S8～S11.
Fig. 6 represents to have carried out S34, and the result has a guide look of 3 dimension thumbnail image S of the sample that shows, the example of common thumbnail image Y on LCD10.
Sample thumbnail image S has been implemented 3 dimension CG graphics process, so know at a glance with the difference of common thumbnail Y, stereo-picture downscaled images Z, is easy to discern face as the stereo-picture that is replicated body.
＜the 3 execution mode 〉
Fig. 7 is that the general chart picture that expression is carried out by host CPU 62 shows another example of handling.The thumbnail image transposing of the multi-view image of flash of light shooting gained is special sample here.
S41～S44 is identical with S1～S4.
In S45, obtain flash light emission information from image file, the situation that switches on or off of the photoflash lamp 44a44b during the recognition image record.Flash light emission information for example is the information of depositing in the Flash mark of Exif file.
In S46, based on flash light emission information, the photoflash lamp 44a44b when judging shooting switches on or off.The occasion that photoflash lamp 44a44b connects when taking enters S47, and the occasion of cut-out enters S49.
In S47, from EEPROM146, take out and carry out 3 dimension thumbnail images of the sample of presentation belonging to the multi-view image this point.Identical with the 2nd execution mode, give stereoeffect by 3 dimension CG processing to this sample.In addition, the pattern of this sample is not the face that is necessary for specific personage.
In S48, in the guide look frame buffer, deposit 3 dimension thumbnail images of the sample that is taken out.3 dimension thumbnail images of sample are not must be same with the appearance that is replicated body of reality, just the thumbnail image that carries out 3D graphical display for the face this point of having duplicated the personage.
S49～S54 is identical with above-mentioned S29～S34.
Fig. 8 represents to have carried out S54, and the result has a guide look of 3 dimension thumbnail image W of the sample that shows, the example of common thumbnail image Y on LCD10.
Sample thumbnail image W has been implemented 3 dimension CG graphics process, so know at a glance with the difference of common thumbnail Y, stereo-picture downscaled images Z, is easy to discern face as the stereo-picture that is replicated body.
And, need not show up to being difficult to carry out sample till the dark image that effectively utilizes without exception as multi-view image, handle also efficient.
＜the 4 execution mode 〉
Fig. 9 is that the general chart picture that expression is carried out by host CPU 62 shows another example of handling.The transposing of the thumbnail image of the multi-view image that has duplicated face is tieed up the thumbnail image that CG handle gained for only implementing 3 at the face zone of this multi-view image here.
S61～S66 is identical with S21～S26., in S66, enter S67 being judged as the occasion of having the face herein, enter S69 in the occasion that is judged as no face.
In S67, from the multi-view image that has duplicated face, extract the face zone, 3 dimension CG are implemented in this face zone handle.This is for example to extract the face zone from the multi-view image that has duplicated face, and rejects this multi-view image with given rejecting rate.The suitable zone, face zone of setting on the downscaled images after rejecting and from former multi-view image, extracting.Then, make the peculiar polygon of face in this setting regions, carry out on this polygon, pasting the texture mapping of rejecting face zone, back, obtain only the thumbnail image that 3 dimension CG handle have been implemented in this face zone.This mode is to carry out 3 dimension CG to handle after the making of rejecting image.
Perhaps,, make this personage's the peculiar polygon of face, on this polygon, paste the texture mapping in face zone, be produced on the multi-view image and only the interim CG image that 3 dimension CG handle gained implemented in the face zone according to this detected face self.Then, reject this interim CG image, obtain only the thumbnail image that 3 dimension CG handle gained to be implemented in the face zone with given rejecting rate.This mode is to handle the back at 3 dimension CG to make the rejecting image.
Above-mentioned texture mapping mode need be from polygonal making, if make accurate polygon, it is very big then to handle load.As polygonal simple making method, can enumerate following method.At first, detect the component part of the faces such as profile, eye, nose, mouth of face from detected face zone, decide the polygon cut-off rule according to the position of the component part of detected face.For example, from two middle body (the glabella) and along the bridge of the nose reach the nose edge line, round two lines, round the profile of the line of mouth, face as the polygon cut-off rule.On this polygon, carry out texture mapping, just can obtain the part protuberance of mouth and nose, enclose hollow easy 3 dimension CG thumbnail images near the eyes.
In S68, in the guide look frame buffer, deposit the 3 dimension CG thumbnail images that are made into.
S69～S74 is identical with above-mentioned S29～S34.
Figure 10 represents to have carried out S74, and the result has a guide look of the pairing 3 dimension thumbnail image S of the multi-view image that detects face that shows, common thumbnail image Y, do not detect the example of the pairing thumbnail image Z of multi-view image of face on LCD10.
Sample thumbnail image U has only implemented 3 dimension CG graphics process at the face zone, so know at a glance with the difference of the downscaled images Z of common thumbnail Y, simple visual point image, is easy to discern face as the stereo-picture that is replicated body.And, adopt original image that 3 dimension CG graphics process have been implemented in the face zone, also can understand the summary of the face of original image three-dimensionally.
＜the 5 execution mode 〉
Figure 11 is that the general chart picture that expression is carried out by host CPU 62 shows another example of handling.Only carry out 3 dimension CG processing here, to taking the in-plant thumbnail image that is replicated the multi-view image of body gained.
S81～S84 is identical with S41～S44.
In S85, from image file, obtain range information, the camera head 1 of identification when image writes down is to the distance that is replicated body.Range information for example is the information of depositing in Exif file " SubjectDistance " mark, by 58 storages of range information memory circuit.
In S86,, judge whether be replicated body is positioned at position than given distance (D＞10m) for example based on range information D.Being replicated body is that remote occasion enters S89, is not that remote occasion is that in-plant occasion enters S87.
In S87, identical with S6, make 3 dimension CG thumbnail images.
In S88, deposit the 3 dimension CG thumbnail images that are made in the guide look frame buffer of in buffer storage 41, guaranteeing.
S89～S94 is identical with above-mentioned S69～S74.
Figure 12 represents to have carried out S94, and the result has a guide look of the 3 dimension thumbnail image T that show, common thumbnail image Y, do not implement the example that is replicated body thumbnail image V far away that 3 dimension CG handle on LCD10.
Sample thumbnail image T has been implemented 3 dimension CG and has handled, so know at a glance with common thumbnail Y, the difference that is replicated the thumbnail image V of body image far away.
And, only be in in-plant multi-view image and implement 3 dimension CG and handle at when taking, being replicated body, need not till implementing that this processings is also insignificant to be replicated body image far away, carry out 3 dimension CG processing without exception, handle efficient.
＜the 6 execution mode 〉
Figure 13 is the block diagram of personal computer 100.Personal computer 100 mainly comprises: the central processing unit (CPU) 102 of controlling the action of each inscape; Control program to device is deposited, the main storage 106 of the operating area when becoming the program execution; Hard disk unit 108, the various application software of the program that the picture editting that the mode that wherein deposit the operating system (OS) that comprises personal computer 100, is fit to by the device driver to the external equipment that is connected with personal computer 100 is edited visual point image handles, user's image etc.; CD-ROM device 110; Interim storage shows the display-memory 116 with data; Wait the monitoring arrangement 118 such as CRT monitor, LCD monitor of display image, literal etc. according to view data, lteral data from this display-memory 116; Keyboard 120; Mouse 122 as the position input equipment; The mouse controller 124 that detects the state of mouse 122 and the signal of the state of the position of the mouse pointer on the monitoring arrangement 118, mouse 122 etc. is exported to CPU102; The communication interface 132 that is connected with networks 30 such as the Internets; Has the wherein card 112 of the card insert opening of detachable memory cards 38; The bus 104 that connects above-mentioned each inscape; Carry out the camera head connection I/F134 that USB is connected with camera head 1.
The application software that leaves the picture editting's processing in the hard disk unit 108 in is wherein to have write down the CD-ROM of this application software by placing in the CD-ROM of personal computer 100 device 110, thereby is installed in the personal computer 100.
In hard disk unit 108, deposit by camera head and connect multi-view image and even the common image of I/F134 from compound eye imaging device 1 acceptance, perhaps multi-view image and even the common image that is taken into from storage card 38 by card I/F112.
CPU102 show to handle for the general chart picture that the multi-view image and the common image of storage in the hard disk unit 108 carries out above-mentioned the 1st～5 execution mode.This processing can be when depositing each image in hard disk unit 108, to carry out separately by each image at every turn, also can be according to the indication from keyboard 120, mouse 122, and all images unification of depositing in the hard disk unit 108 is carried out.
That is, general chart picture shows that handling is not to be undertaken by camera head 1, also can realize in other devices.
Claims (10)
1. image processing apparatus possesses:
Image input part, input comprise the multi-view image that photographs from the viewpoint more than 2 and the image of the single view image that photographs from single viewpoint;
Storage part is stored the image of having imported in the above-mentioned image input part;
Downscaled images preparing department makes the 1st downscaled images of dwindling the image gained of storing in the above-mentioned storage part;
Image recognition portion, discerning the image of storing in the above-mentioned storage part is multi-view image or single view image;
Image processing part is implemented 3 dimension computer graphicals to the 1st downscaled images corresponding with the image that is identified as multi-view image by above-mentioned image recognition portion and is handled, thereby makes the 2nd downscaled images; And
Display part, according to the indication of the image of storing in the above-mentioned storage part being had a guide look of demonstration, guide look shows the 1st downscaled images that is made into by above-mentioned downscaled images preparing department as the representative image that is identified as the image of single view image by above-mentioned image recognition portion, and guide look shows that the 2nd downscaled images that is made into by above-mentioned image processing part is as the representative image that is identified as the image of multi-view image by above-mentioned image recognition portion.
2. image processing apparatus possesses:
Image input part, input comprise the multi-view image that photographs from the viewpoint more than 2 and the image of the single view image that photographs from single viewpoint;
Storage part is stored the image of having imported in the above-mentioned image input part;
Downscaled images preparing department makes the downscaled images of dwindling the image gained of storing in the above-mentioned storage part;
Image recognition portion, discerning the image of storing in the above-mentioned storage part is multi-view image or single view image;
The face test section, the image for be identified as multi-view image by above-mentioned image recognition portion detects and is replicated having or not of body face zone; And
Display part, according to the indication of the image of storing in the above-mentioned storage part being had a guide look of demonstration, guide look shows that above-mentioned downscaled images is identified as multi-view image and is the representative image that is not detected the image that is replicated body face zone by above-mentioned face test section as the downscaled images of the image that is identified as the single view image by above-mentioned image recognition portion and by above-mentioned image recognition portion, and guide look shows and implemented given sample downscaled images that 3 dimension computer graphicals handle as being identified as multi-view image by above-mentioned image recognition portion and being the representative image that is detected the image that is replicated body face zone by above-mentioned face test section.
3. image processing apparatus possesses:
Image input part, input comprise the multi-view image that photographs from the viewpoint more than 2 and the image of the single view image that photographs from single viewpoint;
Storage part is stored the image of having imported in the above-mentioned image input part;
Downscaled images preparing department makes the downscaled images of dwindling the image gained of storing in the above-mentioned storage part;
Image recognition portion, discerning the image of storing in the above-mentioned storage part is multi-view image or single view image;
The flash light emission identification part, for the image that is identified as multi-view image by above-mentioned image recognition portion, the having or not of the flash light emission when identification is taken; And
Display part, according to the indication of the image of storing in the above-mentioned storage part being had a guide look of demonstration, guide look shows that above-mentioned downscaled images is as the representative image that is identified as the image of single view image by above-mentioned image recognition portion, guide look shows the downscaled images that is made into by above-mentioned downscaled images preparing department as being identified as multi-view image by above-mentioned image recognition portion and being the representative image that is identified as the image that does not have flash light emission by above-mentioned flash light emission identification part, and guide look shows has implemented given sample downscaled images that 3 dimension computer graphicals handle as being identified as multi-view image by above-mentioned image recognition portion and being the representative image that is identified as the image of flash light emission by above-mentioned flash light emission identification part.
4. image processing apparatus possesses:
Image input part, input comprise the multi-view image that photographs from the viewpoint more than 2 and the image of the single view image that photographs from single viewpoint;
Storage part is stored the image of having imported in the above-mentioned image input part;
Downscaled images preparing department makes the 1st downscaled images of dwindling the image gained of storing in the above-mentioned storage part;
Image recognition portion, discerning the image of storing in the above-mentioned storage part is multi-view image or single view image;
The face test section detects from the image that is identified as multi-view image by above-mentioned image recognition portion and is replicated having or not of body face zone;
Image processing part set and has been replicated after the corresponding zone, body face zone by above-mentioned face test section is detected above-mentioned the 1st downscaled images, 3 dimension computer graphicals processing carried out in the above-mentioned zone of having set, thereby made the 2nd downscaled images; And
Display part, according to the indication of the image of storing in the above-mentioned storage part being had a guide look of demonstration, guide look shows above-mentioned the 1st downscaled images as the image that is identified as the single view image by above-mentioned image recognition portion be identified as the representative image that is not detected the image that is replicated body face zone by above-mentioned face test section in the image of multi-view image by above-mentioned image recognition portion, and guide look shows that above-mentioned the 2nd downscaled images is as the representative image that is detected the image that is replicated body face zone by above-mentioned face test section.
5. image processing apparatus possesses:
Image input part, input comprise the multi-view image that photographs from the viewpoint more than 2 and the image of the single view image that photographs from single viewpoint;
Storage part is stored the image of having imported in the above-mentioned image input part;
Downscaled images preparing department makes the 1st downscaled images of dwindling the image gained of storing in the above-mentioned storage part;
Image recognition portion, discerning the image of storing in the above-mentioned storage part is multi-view image or single view image;
Apart from the identification part, the body that is replicated that identification is identified as by above-mentioned image recognition portion in the image of multi-view image is at a distance still closely;
Image processing part, to according to being identified as multi-view image by above-mentioned image recognition portion and being to be replicated body and to be in the 1st downscaled images that in-plant image is made into and to implement 3 dimension computer graphicals and handle, thereby make the 2nd downscaled images by above-mentioned being identified as apart from the identification part; And
Display part, according to the indication of the image of storing in the above-mentioned storage part being had a guide look of demonstration, guide look shows that the 1st downscaled images that is made into by above-mentioned downscaled images preparing department is as the representative image that is identified as the image of single view image by above-mentioned image recognition portion, guide look shows the 1st downscaled images that is made into by above-mentioned downscaled images preparing department as being identified as multi-view image by above-mentioned image recognition portion and being to be replicated the representative image that body is in remote image by above-mentioned being identified as apart from the identification part, and guide look shows that the 2nd downscaled images that is made into by above-mentioned image processing part is as being identified as multi-view image by above-mentioned image recognition portion and being to be replicated the representative image that body is in in-plant image by above-mentioned being identified as apart from the identification part.
6. image processing method may further comprise the steps:
Input comprises the multi-view image that photographs from the viewpoint more than 2 and the image of the single view image that photographs from single viewpoint;
The image that storage has been imported;
The 1st downscaled images of the image gained of being stored is dwindled in making;
The image that identification is stored is multi-view image or single view image;
1st downscaled images corresponding with the image that is identified as above-mentioned multi-view image implemented 3 dimension computer graphicals handle, thereby make the 2nd downscaled images; And
According to the indication of the above-mentioned image of storing being had a guide look of demonstration, guide look shows above-mentioned the 1st downscaled images as the above-mentioned representative image that is identified as the image of single view image, and guide look shows that above-mentioned the 2nd downscaled images is as the above-mentioned representative image that is identified as the image of multi-view image.
7. image processing method may further comprise the steps:
Input comprises the multi-view image that photographs from the viewpoint more than 2 and the image of the single view image that photographs from single viewpoint;
The image that storage has been imported;
The downscaled images of the above-mentioned image gained of storing is dwindled in making;
Discerning the above-mentioned image of storing is multi-view image or single view image;
For the above-mentioned image that is identified as multi-view image, detect and be replicated having or not of body face zone; And
According to the indication of the above-mentioned image of storing being had a guide look of demonstration, guide look shows above-mentioned downscaled images as the downscaled images of the above-mentioned image that is identified as the single view image and above-mentionedly be identified as multi-view image and be the above-mentioned representative image that does not detect the image that is replicated body face zone, and guide look shows that having implemented given sample downscaled images that 3 dimension computer graphicals handle is identified as multi-view image and is the above-mentioned representative image that detects the image that is replicated body face zone as above-mentioned.
8. image processing method may further comprise the steps:
Input comprises the multi-view image that photographs from the viewpoint more than 2 and the image of the single view image that photographs from single viewpoint;
Store the above-mentioned image of having imported;
The downscaled images of the above-mentioned image gained of storing is dwindled in making;
Discerning the above-mentioned image of storing is multi-view image or single view image;
For the above-mentioned image that is identified as multi-view image, the having or not of the flash light emission when identification is taken; And
According to the indication of the above-mentioned image of storing being had a guide look of demonstration, guide look shows that above-mentioned downscaled images is as the above-mentioned representative image that is identified as the image of single view image, guide look shows that above-mentioned downscaled images is identified as multi-view image and is the above-mentioned representative image that is identified as the image that does not have flash light emission as above-mentioned, and guide look shows that having implemented given sample downscaled images that 3 dimension computer graphicals handle is identified as multi-view image and is the above-mentioned representative image that is identified as the image of flash light emission as above-mentioned.
9. image processing method may further comprise the steps:
Input comprises the multi-view image that photographs from the viewpoint more than 2 and the image of the single view image that photographs from single viewpoint;
The image that storage has been imported;
The 1st downscaled images of the image gained of being stored is dwindled in making;
The image that identification is stored is multi-view image or single view image;
From the image that is identified as multi-view image, detect and be replicated having or not of body face zone;
Above-mentioned the 1st downscaled images set with above-mentioned be replicated after the corresponding zone, body face zone, 3 dimension computer graphicals processing have been carried out in the above-mentioned zone of having set, thereby made the 2nd downscaled images; And
According to the indication of the above-mentioned image of storing being had a guide look of demonstration, guide look shows above-mentioned the 1st downscaled images as the image that is identified as the single view image be identified as the representative image that does not detect the image that is replicated body face zone in the image of multi-view image, and guide look shows that above-mentioned the 2nd downscaled images is as the representative image that detects the image that is replicated body face zone.
10. image processing method may further comprise the steps:
Input comprises the multi-view image that photographs from the viewpoint more than 2 and the image of the single view image that photographs from single viewpoint;
Store the above-mentioned image of having imported;
The 1st downscaled images of the above-mentioned image gained of storing is dwindled in making;
Discerning the above-mentioned image of storing is multi-view image or single view image;
The body of discerning in the above-mentioned image that is identified as multi-view image that is replicated is at a distance still closely;
To being identified as multi-view image and being that above-mentioned being identified as is replicated body and is in the 1st downscaled images that in-plant image is made into and implements 3 dimension computer graphicals and handle, thereby make the 2nd downscaled images according to above-mentioned; And
According to the indication of the above-mentioned image of storing being had a guide look of demonstration, guide look shows that above-mentioned the 1st downscaled images is as the above-mentioned representative image that is identified as the image of single view image, guide look shows that above-mentioned the 1st downscaled images is identified as multi-view image and is that above-mentioned being identified as is replicated the representative image that body is in remote image as above-mentioned, and guide look shows that above-mentioned the 2nd downscaled images is identified as multi-view image and is that above-mentioned being identified as is replicated the representative image that body is in in-plant image as above-mentioned.
Applications Claiming Priority (4)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2007115824A JP4661824B2 (en) | 2007-04-25 | 2007-04-25 | Image processing apparatus, method, and program |
| JP2007-115824 | 2007-04-25 | ||
| JP2007-122642 | 2007-05-07 | ||
| JP2007122642A JP4632060B2 (en) | 2007-05-07 | 2007-05-07 | Image recording apparatus, method and program |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101296386A CN101296386A (en) | 2008-10-29 |
| CN101296386B true CN101296386B (en) | 2011-12-07 |
Family
ID=40054216
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200810095922XA Expired - Fee Related CN101296386B (en) | 2007-04-25 | 2008-04-25 | Picture processing apparatus and method |
Country Status (2)
| Country | Link |
|---|---|
| JP (1) | JP4661824B2 (en) |
| CN (1) | CN101296386B (en) |
Families Citing this family (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2448272A4 (en) | 2009-06-23 | 2017-04-26 | LG Electronics Inc. | 3d image providing device, display device, and method thereof |
| JP5552769B2 (en) * | 2009-07-29 | 2014-07-16 | ソニー株式会社 | Image editing apparatus, image editing method and program |
| JP2011059658A (en) * | 2009-08-11 | 2011-03-24 | Sony Corp | Display device, display method, and computer program |
| JP2011078029A (en) * | 2009-10-01 | 2011-04-14 | Sony Corp | Electronic equipment, and method and program for reproducing content |
| CN102780905A (en) * | 2012-05-31 | 2012-11-14 | 新奥特（北京）视频技术有限公司 | Method for color correction of 3D (three-dimensional) video |
| JP6462671B2 (en) | 2013-10-03 | 2019-01-30 | ユニバーシティー オブ デラウェア | X slit camera |
| JP6969157B2 (en) * | 2017-05-24 | 2021-11-24 | 富士フイルムビジネスイノベーション株式会社 | 3D shape data editing device and 3D shape data editing program |
| CN107967716B (en) * | 2017-11-01 | 2021-08-06 | 深圳依偎控股有限公司 | Thumbnail display control method and system based on three-dimensional picture |
| CN111292232B (en) * | 2018-12-06 | 2023-08-15 | 成都微晶景泰科技有限公司 | Lens array image stitching method, device and storage medium |
| WO2023282322A1 (en) * | 2021-07-07 | 2023-01-12 | えほんインク株式会社 | System for producing avatar-added product, program for producing avatar-added product, method for manufacturing avatar-added product, and system for creating picture book |
| CN115816833B (en) * | 2023-01-07 | 2023-06-30 | 深圳市创想三维科技股份有限公司 | Method and device for determining image correction data, electronic equipment and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2004120176A (en) * | 2002-09-25 | 2004-04-15 | Sharp Corp | Mobile phone |
| CN1497923A (en) * | 2002-09-25 | 2004-05-19 | ������������ʽ���� | Electronic equipment |
| CN1497987A (en) * | 2002-09-25 | 2004-05-19 | ������������ʽ���� | Image display device and method of displaying thumbnail based on three-D image data |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2000207575A (en) * | 1999-01-08 | 2000-07-28 | Nadeisu:Kk | Space fusing device and application devices adapting the same |
| JP2004163998A (en) * | 2002-09-25 | 2004-06-10 | Sharp Corp | Device and method for displaying picture image |
-
2007
- 2007-04-25 JP JP2007115824A patent/JP4661824B2/en not_active Expired - Fee Related
-
2008
- 2008-04-25 CN CN200810095922XA patent/CN101296386B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2004120176A (en) * | 2002-09-25 | 2004-04-15 | Sharp Corp | Mobile phone |
| CN1497923A (en) * | 2002-09-25 | 2004-05-19 | ������������ʽ���� | Electronic equipment |
| CN1497987A (en) * | 2002-09-25 | 2004-05-19 | ������������ʽ���� | Image display device and method of displaying thumbnail based on three-D image data |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101296386A (en) | 2008-10-29 |
| JP2008276301A (en) | 2008-11-13 |
| JP4661824B2 (en) | 2011-03-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101296386B (en) | Picture processing apparatus and method | |
| CN102308590B (en) | Three-dimensional image output device and three-dimensional image output method | |
| CN105580349B (en) | Image processing apparatus, photographic device and image processing method | |
| CN101959075B (en) | Stereoscopic image recording apparatus and method, stereoscopic image outputting apparatus and method, and stereoscopic image recording outputting system | |
| US8503764B2 (en) | Method for generating images of multi-views | |
| CN104782110B (en) | Image processing apparatus, photographic device and image processing method | |
| CN104871058B (en) | Image processing device, imaging device and image processing method | |
| CN102540689B (en) | Image pickup apparatus | |
| CN102388617B (en) | Compound eye imaging device and disparity adjustment method thereof and program | |
| Devernay et al. | Stereoscopic cinema | |
| CN101542536A (en) | System and method for compositing 3D images | |
| JP2006107213A (en) | Stereoscopic image printing system | |
| CN102972032A (en) | Three-dimensional image display device, three-dimensional image display method, three-dimensional image display program, and recording medium | |
| US8072487B2 (en) | Picture processing apparatus, picture recording apparatus, method and program thereof | |
| CN104755981B (en) | Image processor, image-capturing device and image processing method | |
| CN105051600B (en) | Image processing apparatus, camera device and image processing method | |
| WO2013108285A1 (en) | Image recording device, three-dimensional image reproduction device, image recording method, and three-dimensional image reproduction method | |
| CN104903769B (en) | Image processing apparatus, camera head and image processing method | |
| US20130155204A1 (en) | Imaging apparatus and movement controlling method thereof | |
| CN103155576A (en) | Three-dimensional image display device, and three-dimensional image display method | |
| CN101729916A (en) | Image processing device, and image recording device, method and program | |
| JP4632060B2 (en) | Image recording apparatus, method and program | |
| CN104737527B (en) | Image processing apparatus, photographic device and image processing method | |
| JP4184061B2 (en) | Stereoscopic image editing apparatus and stereoscopic image editing program | |
| JP4995966B2 (en) | Image processing apparatus, method, and program |
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 | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20111207 Termination date: 20140425 |