CN100527792C - Method and apparatus for taking pictures - Google Patents

Abstract

The invention provides a camera device which drives the hyperfine camera element by a low drive frequency for reading out whole pixel. The integral image is low image and the camera object is high image such that they are outputted as a system dynamic image. Control section divides the high image shot by camera device into a high resolution area and a low resolution area and makes the integral pixel group combined by a plurality of high resolution area and a low resolution area read out the predetermined correcting image of the pixel group number each second for frames and measures pixel group numbers of the correcting image, investigates whether the pixel group numbers of the correcting image can read out the predetermined pixel group numbers each second for frames. When the pixel group numbers of the correcting image is able to read out the predetermined pixel group numbers each second for frames, it outputs the image shot based on the pixel group number of the correcting image from the camera device.

Description

Image capture method and camera head

Technical field

The present invention relates to a kind of image capture method and camera head, relate in particular to and will utilize high meticulous imaging apparatus or the meticulous imaging apparatus of the superelevation resulting high-definition picture signaling zone of making a video recording to be divided into a plurality of high-resolution areas and low resolution zone, and image capture method and the camera head read with predetermined data output frequency.

Background technology

Recently, developed and to have broadcasted the meticulous imaging apparatus of superelevation that the above superelevation precise image of mode is photographed to HDTV (high definition television, high definition TV).On the other hand, for example,, need the driving frequency of 74MHz at least for the HDTV of horizontal direction 1920 pixels, vertical direction 1080 pixels is carried out both full-pixel output with (1/30 second) in the captured high precise image of high meticulous imaging apparatus is during 1 frame.Whole pixels of the meticulous imaging apparatus of superelevation that will have the above pixel count of HDTV image need higher driving frequency when exporting for 1 image duration.But, when driving frequency uprises, the electric power of imaging apparatus drive clock circuit and signal processing circuit increases, and the door number increase of signal processing circuit, and the therefore preferred a kind of high precise image that can realize imaging apparatus is taken is to hang down the camera that driving frequency is exported simply.

Therefore, existing known following image-input device: the high meticulous image area of imaging apparatus shooting is divided into the zone of the general image and the high-resolution parts of images of low resolution, with the two image with 30 outputs of per second (for example with reference to patent document 1).In this patent document 1 described image-input device, from the high-resolution solid-state imager that will dispose a plurality of photo detectors to high-density, the reading scan of the integral image of low resolution general image scanning efferent by having reduced resolution, the general image data of output low resolution, the output of high-resolution parts of images scanner section has kept the parts of images data of high-resolution little picture.And, by switching part the general image data and the high-resolution parts of images data of described low resolution were switched in the frame period of 30 of per seconds, exported the general image data and the high-resolution parts of images data of low resolution with the speed more than the video rate in turn.And the image cut handling part in the camera head determines the clipped position of the high-resolution parts of images data of next frame automatically according to the general image data of low resolution, and indication high-resolution parts of images scanner section.

Patent document 1: the spy opens the 2004-180240 communique

But, in patent document 1 disclosed image-input device, though can export the general image data of low resolution and the high-resolution parts of images data of little picture with the image that low driving frequency is taken from the high-resolution imaging apparatus as a whole, but the special-purpose separately output line that each pixel needs low resolution to use and high-resolution is used, and also need the switching part of the output of the general image data and the high-resolution parts of images data of low resolution/read, therefore device needs special-purpose structure.

And, in above-mentioned image-input device, show as a figure in a picture for the view data of the low resolution of filter process (addition/average) the circuit output that will use from low resolution with from the high resolution image data of high-resolution circuit output, volume image how no matter be taken, two districts that all are separated into low-resolution image viewing area and high-definition picture viewing area show, therefore, the difference of body owing to be taken, the problem on the border in these districts can be obviously seen in generation.And the selection of resolution can only be selected 2 kinds in a kind of a kind of low resolution and the high-resolution, and the resolution in the picture is changed reposefully, can not export to make the image that shows different resolution in a plurality of image-regions etc. change the abundant image data.

In addition, as the additive method of the driving frequency that is used to reduce imaging apparatus following method is arranged: imaging apparatus is not read both full-pixel when cmos image sensor, for example every 1 pixel or 4 pixels uniformly-spaced the method for reading.When 4 pixels are read, comparing with reading both full-pixel, driving frequency be its 1/4, but the whole pixel number tails off, and therefore can produce the problem of the image quality aggravations such as integral image roughening of demonstration.Especially always read at certain intervals in the zone of will photographing, and the image of demonstration in a picture and the image kind that will write down also reduces, so image quality aggravation.

Summary of the invention

The present invention is in view of the premises, purpose is to provide a kind of method for imaging and camera, can be from taking at one time the image pickup signal that obtains by the imaging apparatus of the general construction that does not have special-purpose output line or switching part, have the picture signal of a system in the image-region of a plurality of resolution and high-resolution desired image zone with predetermined low driving frequency output.

And, other purposes of the present invention are to provide a kind of image capture method and camera head, when the image region segmentation with a picture becomes more than 2 the zone, can set the resolution of its display image for predetermined resolution simply in each split image zone, and can make the selection of inconspicuous 2 the above resolution in border of the image-region that weighting changes.

To achieve these goals, first invention of the present invention provides a kind of image capture method, from whole photoelectric conversion pixels zone of camera watch region that can random-access imaging apparatus, select the photoelectric conversion pixel zone more than 2, from the photoelectric conversion pixel zone more than 2 of this selection, read picture element signal, obtain image pickup signal, have two-dimensional arrangements in the camera watch region of described imaging apparatus and become rectangular a plurality of photoelectric conversion pixels zone, this image capture method is characterised in that, may further comprise the steps: first step, make figure, this figure is divided into N split image zone with described camera watch region, and to described each split image zone, the detection information that obtains according to the shooting information of utilizing selected detection method from described imaging apparatus, set the weighting ratio of photoelectric conversion pixel number of regions with respect to the ratio of the whole photoelectric conversion pixel number of regions in this split image zone of reading of expression read output signal, wherein N is any natural number more than 1; Second step is according to the described imaging apparatus of described graphics driver; Third step, according to first picture element signal that read in the photoelectric conversion pixel zone of reading in described camera watch region, interpolation generates second picture element signal in the photoelectric conversion pixel zone of not reading in the described camera watch region; And the 4th step, synthetic described first picture element signal and described second picture element signal, and with predetermined frame frequency output image pickup signal.

To achieve these goals, second invention of the present invention provides a kind of image capture method, it is characterized in that, may further comprise the steps: first step, and the detection method of utilize selecting obtains the detection information that the shooting information according to described imaging apparatus obtains; Second step, according to each a plurality of figure of described detection information issuing and be stored in it in storage device in advance, described a plurality of figure has been set the weighting ratio of the ratio of the whole photoelectric conversion pixel number of regions in N the split image zone of reading the photoelectric conversion pixel number of regions and obtaining with respect to cutting apart described camera watch region of expression read output signal, according to the described detection information that in described first step, obtains according to described split image zone, with reference to described storage device, figure like from described a plurality of figures of storage, selecting recently, wherein N is any natural number more than 1; Third step, the described figure according to selecting in described second step drives described pixel element; The 4th step, according to first picture element signal that read in the photoelectric conversion pixel zone of reading in described camera watch region, interpolation generates second picture element signal in the photoelectric conversion pixel zone of not reading in the described camera watch region; And the 5th step, synthetic described first picture element signal and described second picture element signal, and with predetermined frame frequency output image pickup signal.

And to achieve these goals, the 3rd invention provides a kind of camera head, it is characterized in that, comprising: can random-access imaging apparatus, and have two-dimensional arrangements at its camera watch region and become rectangular a plurality of photoelectric conversion pixels zone; The graphic making unit, make figure, the detection information that obtains according to the shooting information of utilizing selected detection method from described imaging apparatus in this figure, described camera watch region is divided into N split image zone, and to described each split image zone, set the weighting ratio of photoelectric conversion pixel number of regions with respect to the ratio of the whole photoelectric conversion pixel number of regions in this split image zone of reading of expression read output signal, wherein N is any natural number more than 1; Driver element is according to the described imaging apparatus of described graphics driver; The interpolation unit, according to first picture element signal that read in the photoelectric conversion pixel zone of reading in described camera watch region, interpolation generates second picture element signal in the photoelectric conversion pixel zone of not reading in the described camera watch region; And image formation unit, synthetic described first picture element signal and described second picture element signal generate image pickup signal and supply to described graphic making unit with predetermined frame frequency.

And to achieve these goals, the 4th invention provides a kind of camera head, it is characterized in that, comprising: can random-access imaging apparatus, and have two-dimensional arrangements at its camera watch region and become rectangular a plurality of photoelectric conversion pixels zone; Detection information obtains the unit, and the detection method of utilize selecting obtains the detection information that the shooting information according to described imaging apparatus obtains; Storage device, store in advance according to each a plurality of figure of described detection information issuing and with it, described a plurality of figure has been set the weighting ratio of the ratio of the whole photoelectric conversion pixel number of regions in N the split image zone of reading the photoelectric conversion pixel number of regions and obtaining with respect to cutting apart described camera watch region of expression read output signal, and wherein N is any natural number more than 1; Selected cell, according to the described detection information that obtains according to described split image zone, with reference to described storage device, figure like the selection recently in described a plurality of figures of from this storage device, storing; Driver element, the described figure according to selecting drives described pixel element; The interpolation unit, according to first picture element signal that read in the photoelectric conversion pixel zone of reading in described camera watch region, interpolation generates second picture element signal in the photoelectric conversion pixel zone of not reading in the described camera watch region; And image formation unit, synthetic described first picture element signal and described second picture element signal, and with predetermined frame frequency output image pickup signal.

In the first and the 3rd invention, make figure, this figure is according to each split image zone, the detection information that obtains according to the shooting information of utilizing selected detection method from imaging apparatus, set the weighting ratio of photoelectric conversion pixel number of regions of reading of expression read output signal with respect to the ratio of the whole photoelectric conversion pixel number of regions in this split image zone, graphics driver imaging apparatus according to this making, therefore, the figure that utilization is made according to selected detection method can be read for the pixel that needs high-resolution split image zone in the shooting object and do not need high-resolution split image zone to carry out best weighting.

And, in the second and the 4th invention, based on the detection information that obtains according to each split image zone, with reference to detecting the storage device that information issuing has been set a plurality of figures of weighting ratio and stored in advance according to each, figure like selecting recently in a plurality of figures from be stored in this storage device, therefore, can read according to the pixel of the figure in the split image zone of high-resolution split image zone in the detection information shooting object that carried out best configuration respectively rapidly and simply and the mutually different a plurality of low resolution of resolution.

The invention effect

According to the present invention, the figure that utilization is made according to selected detection method, read for the image that needs high-resolution split image zone in the shooting object and do not need high-resolution split image zone to carry out optimum weighting, therefore, the drive signal that the picture signal of the system of taking at one time can be utilized low frequency is with predetermined frame frequency output, the picture signal of this system guarantees to make a video recording image section that object etc. needs resolution for high meticulous in the captured high-fineness image of the meticulous imaging apparatus of height, be low resolution with the image section of other parts.

And, according to the present invention, based on the detection information that obtains according to each separate image zone, with reference to detecting the storage device that information issuing has been set a plurality of figures of weighting ratio and stored in advance according to each, figure like selecting recently in a plurality of figures of from this storage device, storing, thus, can read according to the pixel of the figure in the split image zone of high-resolution split image zone in the detection information shooting object that carried out best configuration respectively rapidly and simply and the mutually different a plurality of low resolution of resolution, so not need to be used for the processing and the circuit of the complexity of graphic making.

And, according to the present invention, a plurality of split images zone of extracting the whole photoelectric conversion pixels zone in the split image zone at interval and reading, can set a plurality of resolution for, therefore can set in the specific region of picture is the high-resolution areas that both full-pixel is read, zone away from this specific region in picture is the low resolution zone, and resolution far away more is low more, at this moment change resolution is mild, and also can be configured to be provided with in picture the abundant resolution of variation of high-resolution areas that a plurality of both full-pixel read etc.

Description of drawings

Fig. 1 is the block diagram of the 1st execution mode of camera of the present invention.

Fig. 2 is the figure that the guide of expression photographic picture of the present invention is set example.

Fig. 3 is the block diagram of an example of the drive division of expression camera of the present invention.

Fig. 4 is the figure of cutting apart example in expression high-resolution areas of the present invention and low resolution zone.

Fig. 5 is the figure of configuration example of the pixel group of expression high-resolution areas of the present invention.

Fig. 6 is the figure of configuration example (one) of the pixel group in expression low resolution of the present invention zone.

Fig. 7 is the figure of configuration example (its two) of the pixel group in expression low resolution of the present invention zone.

Fig. 8 is the figure of configuration example (its three) of the pixel group in expression low resolution of the present invention zone.

Fig. 9 is the figure of kind of operable weighting pattern in first execution mode of expression image capture method of the present invention.

Figure 10 is the flow chart of action of each pattern that is used to illustrate first execution mode of image capture method of the present invention.

Figure 11 is the flow chart of action that is used to illustrate second execution mode of image capture method of the present invention.

Figure 12 is the block diagram of second execution mode of the present invention.

Figure 13 is the figure that is illustrated in an example of employed divided frame in image capture method of the present invention and the camera head.

Figure 14 is to the view data of each divided frame of second execution mode of image capture method of the present invention and camera head and the figure that processing describes thereof.

Figure 15 is expression and the figure of each example of the making figure figure table relatively of each divided frame of second execution mode of method for imaging of the present invention and camera.

Figure 16 is the circuit system figure of an example of imaging apparatus.

Figure 17 is the sequential chart of the imaging apparatus of Figure 16 when moving usually.

Figure 18 is that the imaging apparatus of Figure 16 is being read the sequential chart that figure is weighted an example of reading when action by pixel.

Figure 19 is the block diagram of the embodiment 1 of camera of the present invention.

Figure 20 is the figure of action of the major part of explanation camera shown in Figure 19.

Figure 21 is that expression pixel of the present invention is read the figure of each example of figure.

Figure 22 is that expression detection method of the present invention and pixel are read the figure of each example of figure.

Embodiment

Below, with reference to accompanying drawing embodiments of the present invention are described.

Fig. 1 represents the block diagram of first execution mode of camera of the present invention.Camera head 10 shown in Figure 1 is made of zoom control part 11, camera lens 12, imaging apparatus 13, A/D converter section 14, DV (Digital Video) recording portion 15, storage card 16, drive division 17, control part 18, operating portion 19 and DSP (Digital Signal Processor, digital signal processor) 20.Imaging apparatus 13 for example is the high meticulous imaging apparatus of horizontal direction 1600 pixels, vertical direction 1200 pixels, is made of cmos image sensor.

DSP 20 has the function of interpolator 21, line storage 22, image forming part 23, Y/C (Y=luminance signal, C=color difference signal) handling part 24, test section 25, FMC (field mode change) handling part 26.Test section 25 has the measuring ability of EIS (hand jitter correction) circuit 251, AF (automatic focus) circuit 252, face detection circuit 253.

And, shown in the block diagram of Fig. 3, drive division 17 is made of drive signal generator 171, frequency divider 172 and clock generator 173, and control part 18 is made of graphics controller 181, figure projection instrument (pattern map) 182, counter 183 and microcomputer (CPU) 189.

Secondly, the action to camera head 10 describes.In Fig. 1, at first, drive division 17 produces following drive signal: the body that is taken of shooting can carry out all pixel groups of opto-electronic conversion as the imaging apparatus 13 that can take high precise image (comprising the superelevation precise image), and (at this, so-called pixel group is meant the unit of the image-element that R, G, four pixels of G, B constitute.The below same record of use) picture signal, the picture signal that peripheral part extracts as every interval 1 pixel group is with predetermined frame frequency output photographic image data.

Imaging apparatus 13, to generate as whole view data of pixel groups via the photography target part (be also referred to as be taken body portion) of lens 12 in the body light image that is taken of not shown photoelectric conversion part imaging, and the part (being also referred to as periphery) beyond the part of the photography target in the body light image that will be taken generates as the view data of the pixel group that extracts at interval.A/D converter section 14 will convert data image signal to from the simulation image pickup signal of imaging apparatus 13 outputs.The interpolation portion 21 of DSP 20 is stored in data image signal in the line storage 22, generate to carry out interpolation as the part of the pixel group of the view data output of the pixel group that extracts at interval view data.Image forming part 23 will be stored in the built-in not shown frame memory from the view data of whole pixel groups of A/D converter section 14 output and the view data addition (synthetic chronologically) that obtains in 21 interpolations of interpolation portion.

Y/C handling part 24 carries out known matrix operation to R (red), G (green), the trichromatic view data of B (indigo plant) supplied with by image forming part 23, generates brightness signal Y, and generates two kinds color difference signal C by known computing.And then Y/C handling part 24 also carries out γ to be handled.FMC handling part 26 to brightness signal Y and the color difference signal C by Y/C handling part 24 output, carries out the amplification of photographs/dwindle and wait the electronic zoom processing.Dynamic image signal after DV recording portion 15 records are handled by the electronic zoom of FMC handling part 26 inputs, storage card 16 storages are by static image of the high-resolution of FMC handling part 26 inputs or the dynamic image of short time.

The EIS circuit 251 of test section 25, by the grade motion detection hand amount of jitter in the zone that comprises predetermined radio-frequency component of the profile portion that comprises in the view data that obtains in photography, shear in the view data from be stored in image forming part 23 and output has compensated the view data of hand amount of jitter.And then mobile lens 12 makes the volume image imaging on imaging apparatus 13 that is taken of the action of having proofreaied and correct hand shake composition.AF circuit 252 detects the radio-frequency component of the image of imaging, and lens 12 when mobile, are moved lens 12 on the position that could access big radio-frequency component on optical axis.Face detection circuit 253 detects colour of skin part according to the volume image that is taken, and detects the lower positions of brightness level such as whether eye is arranged in the colour of skin image-region, obtains if detect, and then this area of skin color is detected to be the part of face.EIS circuit 251, AF circuit 252 and face detection circuit 253 are all moved in the radio-frequency component that detects the picture signal that is obtained by Y/C handling part 24.

Control part 18 is used for obtaining doing according to the image-region that comprises detected radio-frequency component the control of the photographs of fineness.That is, control part 18 generates following control signal: be used for the both full-pixel group being photographed in the zone of the body portion that is taken that comprises more radio-frequency component interregional every extracting and photographing to periphery.And then, the microcomputer 189 of control part 18, judge whether that photographs from imaging apparatus 13 that these are regional exports with the scheduled time, when surpassing the scheduled time relevant, changes the high-resolution areas relevant and the ratio in the low resolution zone of being correlated with the image of the pixel group of interval extraction with both full-pixel group's image with frame frequency.

Drive division 17 generates and is used for photography is divided into the drive signal that the image of a plurality of resolution is photographed.The drive that imaging apparatus 13 origin are self-driven 17 with the picture signal of body portion as the both full-pixel group that be taken, as the image that is spaced apart extraction, is exported photographs with periphery.After, test section 25 detects the image-region that comprises radio-frequency component according to the picture signal of taking, and control part 18 decisions become the resolution in a plurality of zone with picture segmentation, and continue to take.Each one in the microcomputer 189 of Fig. 3 and each circuit part of camera head 10 and the control part 18 is connected, and controls its action.

Next, with reference to Fig. 2, the guide of photographic picture is provided with example describes.Fig. 2 (A) is the picture of landscape screening-mode, and Fig. 2 (B) is the picture of personage's photograph mode.Landscape photography pattern picture is with the picture of reading the low resolution that pixel shows every 1/2 pixel group after extracting in the both full-pixel of imaging apparatus 13 with picture integral body.Personage's photograph mode picture, picture middle section in the facial photography of supposition, being in this regional imaging apparatus 13 takes with high-resolution reading under both full-pixel group's pattern of both full-pixel, along with to the picture circumferential expansion, as 1/2 of 1/2 the pixel count of reading this regional full figure prime number extract the zone at interval, read this regional full figure prime number 1/4 pixel count 1/4 extract the zone at interval, export different low-resolution images respectively.In addition, in Fig. 2 (A), Fig. 2 (B), separately read the total identical of pixel.

These photograph modes, by the operating portion 19 of cameraman's application drawing 1, the regional graphics that appointment is provided with are in advance also set.When being set into automatic switchover mode,, move as the landscape photography pattern when the AF circuit 252 by test section 25 detects the distance of closing focus when being a distant place, when detect close focus distance for example in the 4m time, move by personage's pattern.And, when utilizing face detection circuit 253 in comparatively central zone, to detect face image, character image is obtained as high-definition picture by personage's pattern.

When the landscape photography pattern, from the 1/2 pixel group data that extract at interval of imaging apparatus 13 output per seconds 30 frames.When personage's pattern, the photographed images of high-resolution areas (both full-pixel group zone), 1/2 low resolution that extracts at interval and the photographs of 1/4 low resolution that extracts at interval are as the dynamic image output of per second 30 frames.For the deficiency that extracts at interval caused pixel count, after interpolation portion 21 carries out interpolation, carry out YC and handle, extract the driving of figure control imaging apparatus 13 at interval according to it, amplify or dwindle processing at FMC handling part 26.

Next, with reference to Fig. 3 drive division 17 and control part 18 are elaborated.The graphics controller 181 of control part 18 according to the information relevant with radio-frequency component by EIS circuit 251, AF circuit 252 and the face detection circuit 253 detected picture signals of test section 25, generates the relevant information of resolution with captured image.The 182 pairs of photographss of figure projection instrument zone generates with high-resolution areas (both full-pixel group zone) and by the figure described later that predefined extraction yield extracts and the low resolution zone of making a video recording is relevant video (pattern map).Extraction yield is appointed as by the specified ratio of mark beyond 1/2 or 1/4.183 pairs of numbers from the pixel group of the image pickup signal of imaging apparatus 13 outputs of counter are counted.

When in microcomputer 189, judging the view data output of formation when the number of the pixel group of the image pickup signal of imaging apparatus 13 output can not carry out that for example frame rate is 30Hz, generate and be used to reduce the size of high-resolution areas or the regional relevant adjustment signal of low resolution of change extraction yield.Graphics controller 181 changes the relevant information of resolution with the image of taking.In figure projection instrument 182, generate the figure reflection of having proofreaied and correct resolution as mentioned above.

The clock generator 173 of drive division 17 generates the clock signal that is used to drive DSP 20.This clock signal is fed into DSP 20.Be located at the image forming part 23 in the DSP 20, can store the both full-pixel group's who is taken by imaging apparatus 13 two field picture, DSP 20 carries out the interpolation processing of the view data of being stored, the treatment for correcting of picture characteristics etc.The clock frequency frequency division that frequency divider 172 will be generated by clock generator 173 is the clock of preset frequency.

Drive signal generator 171, the synchronous drive signal of clock frequency of generation and 172 frequency divisions of frequency divider, and supply to imaging apparatus 13, image pickup signal is read and exported to basis by the pixel of figure projection instrument 182 appointments from the both full-pixel of imaging apparatus 13.Promptly, drive signal generator 171, with respect to imaging apparatuss that element constituted 13 such as CMOS, generate the level relevant and the address signal of vertical direction with the location of pixels of output image pickup signal, and supply to imaging apparatus 13, and also supply with the readout clock signal that is used to export the pixel of reading by this address signal.In the pixel of not reading, reset according to (between each given period) reference signal.

Return Fig. 1 once more and describe, according to the readout clock signal from imaging apparatus 13 output image pickup signals (both full-pixel group's view data or be spaced apart the view data of the pixel group of extraction).A/D converter section 14 will convert digital signal to from the image pickup signal of imaging apparatus 13 output (view data of the pixel group that both full-pixel group's view data or be spaced apart has extracted).The image data storage that interpolation portion 21 will be extracted and export and is utilized the view data be stored in the line storage 22 in line storage 22, by linear interpolation or curve interpolating, the pixel group data that are not extracted and export are generated interpolation data.In image forming part 23, synthetic chronologically and store captured view data and interpolation data.The size of view data of storage is the image of the pixel data same size that can take with imaging apparatus 13.

In the present embodiment, imaging apparatus 13 for example is arranged in horizontal direction 1600 pixels, vertical direction 1200 pixels rectangular in its shooting face, for for example to read this both full-pixel in 1/60 second, need this quite high driving signal frequency of about 115MHz (=1600 * 1200 * 60), the consumption electric power that imaging apparatus drives with clock circuit and signal processing circuit becomes big, and when utilizing integrated circuit to realize these signal processing circuits, the door number of circuit increases.Therefore, in the present embodiment, be set to for example 28.75 (115/4) these lower frequencies of MHz from the frequency of the drive signal of drive signal generator 171 output.But, can't be for example to read the both full-pixel of the camera watch region of described imaging apparatus 13 in 1/60 second by this low frequency drive signal.

But, in the camera head of present embodiment, camera watch region is divided into a plurality of zones, wherein Qi Wang a cut zone is as high-resolution areas, read the both full-pixel of this cut zone, output both full-pixel group data, the cut zone of in addition other more than one or two is as the low resolution zone of the different resolution more than or two, read pixel with its resolution corresponding number, output is spaced apart the pixel group data of extraction, even therefore the drive signal of above-mentioned lower frequency for the body that is taken that moves, also can obtain the pixel group data of desired region as high-resolution areas with predetermined frame frequency.By said method, can realize a kind of camera head, it can be simply with the image section of high-resolution output by the high-resolution body that is taken of needs of imaging apparatus 13 shootings.

Secondly, with reference to Fig. 4 the figure reflection is described.Figure reflection shown in this figure, expression has the split image zone of the scope 1～scope 4 of 4 resolution.As the scope 1 in 2 split image zones of picture central portion, be the high-resolution areas of reading the both full-pixel group of captured image.Be provided with 2 scope 1 zones among the figure.The number of zone appointment is freely, and the shape of scope also can be appointed as arbitrary shape.As the scope 2 in the split image zone of encirclement scope 1, be that extraction yield is 4/6, promptly obtains the first low resolution zone of 4 pixel groups from 6 pixel groups at interval.As the scope 3 in the split image zone of encirclement scope 2 is that extraction yield is 2/4 the second low resolution zone at interval, is that extraction yield is 1/4 the 3rd low resolution zone at interval as the scope 4 in the split image zone of encirclement scope 3.Set in this example: except high-resolution areas, the mutually different low resolution of resolution zone has 3, and towards periphery, its resolution reduces from the picture central part.

Next, each view data to scope 1～scope 4 describes.The pixel group data that explanation is read in scope 1 with reference to Fig. 5.Fig. 5 (A) utilizes R, G, B to represent the pixel of 3 primary colors that the pixel group read with both full-pixel is relevant.Shown in Fig. 5 (B), 1 pixel group by the pixel of 1 R, 2 G and 1 B, amount to 4 pixels and constitute.The details that can cut apart each pixel group appointed area.

Secondly, with reference to Fig. 6 the pixel group data of reading in scope 2 are described.Fig. 6 (A) utilizes R, G, B to represent the pixel of reading with interval extraction yield 4/6.The configuration of the pixel that Fig. 6 (B) expression is disposed 6 pixel groups repeatedly as 1 unit.In these 6 pixel groups, the pixel group of the 1st row left end is read 4 pixels.The pixel group on its right side is read the pixel of G and B, and the pixel group on right side is read the pixel of R and G again.Similarly, the 2nd row are read 8 pixels shown in Fig. 6 (B).Shown in Fig. 6 (B), read 4 pixel group data in fact from 6 pixel groups.But the interpolation of carrying out in interpolation portion 21 to the pixel that extracts at interval needs to consider that pixel arrangement between homochromy concerns carries out.

Next, with reference to Fig. 7 the pixel group data of reading in scope 3 are described.The pixel of reading in the figure, is 4 pixels that constitute same pixel group.The alignment arrangements of pixel group becomes staggered, and extraction yield is 2/4 at interval.By the configuration of this staggered pixel group, suppress deterioration because of the resolution that extracts the horizontal direction that causes and vertical direction at interval.Also easily the pixel group data that extract are at interval carried out interpolation operation.

Secondly, with reference to Fig. 8 the pixel group data of reading in scope 4 are described.In the figure, the pixel group of reading is 1/2 of a scope 3.In this embodiment, the resolution of vertical direction is identical with scope 3, but the resolution of horizontal direction is 1/2 of scope 3, and extraction yield is 1/4 at interval thus.

More than the configuration example of the pixel group read in scope 1～4 is illustrated.The interval abstracting method of pixel it is also conceivable that the whole bag of tricks except that above-mentioned.Constitute the pixel group number of 1 unit and from this unit, extract at interval and the relation of the number of the pixel group of output when constituting fixedly, the interpolating method of can be unique definite interpolation portion 21, therefore, it is easy that interpolation operation becomes.

At this, camera head 10 has control part 18 as shown in Figure 3, this control part 18 has as the lower part: figure projection instrument (table) 182, be used for low resolution zone that capturing element 13 captured high precise images (comprising the superelevation precise image) are distinguished into high-resolution areas and can read with predetermined data read frequency (transfer rate), and generation correcting image, it makes divides into a plurality of described high-resolution areas and the regional whole pixel number that forms that merges of described low resolution, becomes the pixel count of reading with predetermined data output frequency (fixing); Measure the counter 183 of the pixel group number of described correcting image; And graphics controller 181, whether the pixel count of the described correcting image of investigation counter 183 meterings is the pixel count of reading with described tentation data output frequency, when being the pixel count of reading with described tentation data output frequency, carry out output based on the pixel count of described correcting image.

Next, the action in each pattern of the camera head 10 of present embodiment describes to this controller 18.As the example of the weighting pattern that can select in the present embodiment, as shown in Figure 9, comprising: carry out external schema 41, the zoom mode 42 of the weighting fixed by the cameraman and the image state pattern 43 of carrying out graphic making by image state.And, in zoom mode 42, comprise optical zoom pattern 44 and electronic zoom pattern 45.And, in image state pattern 43, comprise the motor pattern 46 that utilizes EIS circuit 251, the high resolution model 47 that utilizes AF circuit 252 and personage's pattern 48 of using face detection circuit 253.

At first, with reference to the action of the flowchart text external schema of Figure 10 (A).Control part 18, operation acceptance model selection input (step S11) by operating portion 19, when being judged as external schema (step S12), select the figure (step S13) used by the external schema of cameraman's appointment from the pre-prepd figure projection instrument (form) that stores multiple figure 182, and carrying out read (the step S14) of imaging apparatus 13 via drive division 17 according to this figure, the different resolution in number, shape and each split image zone in the split image zone of described multiple figure when 1 picture segmentation is become a plurality of image-region constitutes.The pixel group number that utilizes the specified figure of external schema and read from imaging apparatus 13 always can exported below the intended pixel group number of 1 two field picture with predetermined frame frequency.

, judge whether change pattern (step S15), be judged as under the situation of the pattern of not changing, continue reading of step S14 thereafter, under the situation that is judged as the change pattern, tenth skill.In this external schema, carry out the weighting fixed by the cameraman, carry out reading of imaging apparatus 13 by the figure that not influenced by other mechanisms or image state.

Secondly, the action to zoom mode describes with reference to the flow chart of Figure 10 (B).Control part 18, operation acceptance model selection input (step S21) according to operating portion 19, judging (step S22) when being optics/electronic zoom pattern, from pre-prepd figure projection instrument (form) 182, select and the corresponding figure of zoom ratio (step S23).At this moment, in pre-prepd a plurality of figures, zoom ratio high more (situation when getting over longshot), in the split image zone of picture central authorities, the high-resolution areas of reading as both full-pixel, and its area is less is in the split image zone of in addition periphery, by the split image zone of periphery, select the figure in the low resolution zone that extraction yield is big more at interval more.This point all is same in optical zoom and electronic zoom.

Thereafter, judge whether that in the zoom action (step S24) when being in the zoom action, reading until the zoom action of state stops (step S25) before keeping, when not being in the zoom action, carry out read (the step S26) of imaging apparatus 13 according to the figure of selecting among the step S23.The pixel group number that the figure of utilization appointment under zoom mode is read from imaging apparatus 13 can exported below the predetermined pixel group number of 1 two field picture with predetermined frame frequency.

Then, judge whether change pattern (step S27),, return step S24, judge whether to be in the zoom action judging when not changing pattern, when judging the change pattern, tenth skill.Under this zoom mode, zoom ratio hour, the bigger integral image in zone that is suitable for mid-resolution such as scope 2 shown in Figure 4 or scope 3 is photographed uniformly, when zoom ratio is big, therefore the depth of focus of photographs shoals, and is suitable for the high-resolution areas of scope 1 and the bigger photography of resolution of scope 4 as shown in Figure 4.

Next, the action to the image state pattern describes with reference to the flow chart of Figure 10 (C).Control part 18 according to the operation acceptance model selection input (step S31) of operating portion 19, when being judged as the image state pattern (step S32), is made and the corresponding figure of image state (step S33) by figure projection instrument (form) 182.In this case, when the pattern of selecting is motor pattern, testing result making figure based on the action vector of the be taken body relevant that detects in the EIS circuit 251 with the hand shake, when the pattern of selecting is high resolution model, the testing result of the radio-frequency component that obtains when the automatic focus photometry according to AF circuit 252 is made figure, when the pattern of selecting is personage's pattern, make figure based on the face image testing result of face detection circuit 253.

Then, with the figure of making compare with the pre-prepd figure projection instrument (form) 182 that stores multiple figure (step S34), in the multiple figure from this form, select and the most consistent figure of making (step S35) of figure, according to the figure of this selection, carry out read (the step S36) of imaging apparatus 13.

Then, judge whether change pattern (step S37).In this step S37, when not changing pattern, utilize 183 pairs in counter to count by the above-mentioned relevant valid pixel group number of number that outputs to the pixel group of A/D converter section 14 from imaging apparatus 13 of reading, the microcomputer 189 of control part 18 judges whether the valid pixel group number of counting gained can exported below the intended pixel group number of 1 two field picture with predetermined frame frequency.When judging the pixel group number and surpass predetermined pixel group and count, the area that making is corrected into the high-resolution areas of selected figure diminishes and the area in low resolution zone becomes the figure reflection of big figure, the figure of this making is compared with the figure of pre-prepd form, select the most consistent figure of weighting (step S33～S35).

On the other hand, judging valid pixel group number is that above-mentioned intended pixel group number is when following, read shot image data, but reading under the few situation of pixel count, the figure of making is compared with the figure of pre-prepd form, select weighting the most consistent, effectively read figure (the step S33～S35) of the approaching above-mentioned intended pixel group number of pixel count.And, when in step S37, being judged as the change pattern, tenth skill.

Like this, utilizing EIS circuit 251 to detect under the situation of motor pattern of mobile vector of the body that is taken, body when mobile very fast (body that is taken) in a big way when mobile is taken, be chosen in high-resolution areas and the bigger figure of its area that read as both full-pixel in the split image zone of picture central authorities, body when slow (body that is taken move) in than close limit when mobile is taken, select following figure: the high-resolution areas of reading for both full-pixel in the split image zone of picture central authorities, and its area is less; Split image zone in addition periphery is the low resolution zone, and by the split image zone of periphery, extraction yield is big more at interval more.

And, as the body information that is taken, when detecting the hand shake by EIS circuit 251, sell shake when bigger than set point in judgement, select full frame for for example 1/2 at interval extraction yield low resolution zone etc., to the figure of reading pixel count more (image information of the body that is taken of picture periphery is more) of picture periphery, be dithered as set point when following at hand, select following figure: in the split image district of picture central authorities is the high-resolution areas that both full-pixel is read; In in addition peripheral split image zone is the low resolution zone, and more by the split image zone of periphery, extraction yield is big more at interval.

And, select high resolution model, the radio-frequency component that AF circuit 252 obtains when detecting the automatic focus photometry, in general, whether the automatic focus photometry region is by having the scope that radio-frequency component is determined the zone at the middle section that is positioned at predetermined picture central authorities, when in this middle section, not having radio-frequency component, further enlarge this middle section, similarly judge whether have radio-frequency component in the middle section.At this moment, prepare the figure of resolution a plurality of and the middle section interlock that is used for AF, select following figure: the high-resolution areas of reading for both full-pixel in the split image zone of the picture central authorities corresponding with the middle section that is used for AF of decision, partition graph zone in addition neighboring area is the low resolution zone, and by the split image zone of periphery, extraction yield is big more at interval more.Under the situation of having dwindled the middle section that is used for AF on the contrary, also carry out and above similar figure selecting.

Like this, in the present embodiment, can select the multiple figure that different resolution constituted pre-prepd, split image number of regions, shape and each split image zone when a picture segmentation is become a plurality of image-region by the cameraman, also can select to be suitable for most the figure of the captured body that is taken, generate drive signal, in addition, the combination of selected figure prepares to make the combination of the fixing figure of the pixel count read.By pixel count is maintained fixed, can from imaging apparatus 13, read the image that remains on predetermined quality with predetermined driving frequency.Read in the method for sum of pixel count in management, when sense data surpasses predetermined number, need setting regions figure again, in contrast to this, select to pre-determine the regional graphics of sum of all pixels and when reading, saved the sum management of reading pixel count.

As mentioned above, according to the imaging apparatus shown in the present embodiment 10, have following special structure: imaging apparatus 13 has a plurality of photoelectric conversion pixels zone of two-dimensional arrangements in length and breadth; Drive division 17, a plurality of photoelectric conversion pixels zone do not carried out extracting at interval first image-region directly read as first picture element signal on ground and carry out extracting at interval from a plurality of photoelectric conversion pixels zone and during second image-region read as second picture element signal being divided into, read first picture element signal from first image-region, read second picture element signal from second image-region; Interpolation portion 21 utilizes described first or second picture element signal of reading that interpolation is carried out in the photoelectric conversion pixel zone of extracting at interval, generates the 3rd picture element signal; And image forming part 23, synthetic first picture element signal and described the 3rd picture element signal, form 1 frame or 1 's image, therefore, can be achieved as follows camera, it makes the image section of the resolution that requires photography target etc. keep high meticulous image by the high-fineness picture signal of being taken by the meticulous imaging apparatus of height, can be with it as exporting with the picture signal of predetermined frame frequency output.

(the 2nd execution mode)

Next, second execution mode of the present invention is described.In above-mentioned the 1st execution mode, the camera watch region (a plurality of photoelectric conversion pixels zone of two-dimensional arrangements in length and breadth) of imaging apparatus is not cut apart or is divided into zone more than 2, make both full-pixel zone or each cut zone ratio of reading pixel count (extraction yield or weighting at interval) with respect to the full figure prime number in separately zone, can change according to the detection information of be taken body information or photographic information etc., and the 2nd execution mode constitutes, utilize and detect the pre-prepd form of Information Selection, utilize the output of form easily described interval extraction yield and weighting to be carried out variable setting.Basic structure and Fig. 1 of the camera head of the 2nd execution mode are same.

Figure 11 represents the flow chart of second execution mode of image capture method of the present invention.In the figure, the imaging apparatus of present embodiment can arbitrary access, and as mentioned above, it for example is cmos image sensor with camera watch region of the high meticulous image pickup that can carry out horizontal direction 1600 pixels, vertical direction 1200 pixels, extract the both full-pixel of this camera watch region at interval and read with less interval extraction yield, obtain general image (step S100).Extracting at interval and read pixel, is for the above-mentioned camera watch region of taking high precise image from the imaging apparatus of present embodiment, utilizes the drive signal of this lower frequency of 28.75MHz for example reading whole picture signal in 1/60 second.

Then, the control part of camera head (be equivalent to Fig. 1 18), obtain the appointed information (step S101) that the cameraman utilizes which kind of detection method (photograph mode) to photograph, when this designates the information as motor pattern, the mobile vector of selecting EIS circuit (be equivalent to Fig. 1 251) to carry out detects (step S102), when high resolution model, the radio-frequency component of selecting AF circuit (be equivalent to Fig. 1 252) to carry out detects (step S103), when the face detection pattern, the face detection (step S104) of selecting face detection circuit (be equivalent to Fig. 1 253) to carry out obtains the detection information (step S105) of above-mentioned selection thereafter.

Then, the control part of camera head utilizes the information that detects in step S105, with reference to the database corresponding with specified image pickup mode, selects and the only form of detected information (step S106) in a plurality of forms of storage in advance from this database.At this, above table represents that the pixel of the camera watch region of imaging apparatus reads figure.This pixel is read figure, the configuration of each cut zone (comprise and not cutting apart) of the camera watch region of expression imaging apparatus and the ratio of reading pixel count with respect to the full figure prime number of this each cut zone (extraction yield or weighting at interval) utilize this form can carry out the variable setting of interval extraction yield or weighting simply.In addition, the object lesson that this pixel is read figure illustrates later.

Then, the control part of camera head is stored in the represented pixel of selecting in the microcomputer of form and reads graphical information (step S107).Microcomputer is read graphical information according to the pixel of being stored, the drive division of controlling and driving imaging apparatus (be equivalent to Fig. 1 17) (step S108), from the both full-pixel of imaging apparatus, only be appointed as the pixel of reading pixel from read figure by pixel, read picture element signal (step S109).

Thus, according to present embodiment, by the high-fineness picture signal of utilizing high meticulous capturing element to take, to export according to the high-resolution picture signal that photography target requires the image-region of resolution to read as full images, with in addition image-region as the picture signal of the predetermined frame frequency that can export, the picture signal output of the low resolution after the extraction processing (perhaps weighted) at interval, this can set by the utilization open-and-shut processing that body information etc. selects that is taken, and the image-region that can set low resolution for is made of a plurality of image-regions that resolution gradually changes along with the periphery of close picture, so change resolution is milder.

(embodiment 1)

Next, the embodiment 1 to second execution mode describes.This embodiment 1 is the embodiment of cameraman when having specified above-mentioned high resolution model in advance.The block diagram of the embodiment 1 of the camera head of the present invention when Figure 12 has represented to specify high resolution model.In Figure 12, from light scioptics 31 optically focused of the body that is taken, imaging and carry out opto-electronic conversion on the shooting face of imaging apparatus 32.This imaging apparatus 32, with the imaging apparatus 13 of Fig. 1 similarly, be for example have can shooting level direction 1600 pixels, the cmos image sensor of the camera watch region of the high precise image of vertical direction 1200 pixels, the drive signal that is used to this lower frequency of for example 28.75MHz of self-driven 39 drives, for example with 1/60 second frame frequency, extract the both full-pixel of the camera watch region of imaging apparatus 32 at interval with less interval extraction yield, read by reading that pixel is carried out opto-electronic conversion and the picture signal (the step S100 that is equivalent to Figure 11) that obtains.

Image pickup signal from imaging apparatus 32 outputs, after AFE (Analog Front End) circuit 33 has carried out that gain is adjusted etc., supply to back image processing part 34, after carrying out gamma correction etc., supply to FMC handling part 36, carry out at this after electronic zoom such as the amplification of photographs/dwindle handles, to not shown demonstration receiver output.

And owing to specified high resolution model, therefore the picture signal of taking out from the part of image processing part 34 is fed into AF circuit 35, detects radio-frequency component.AF circuit 35 at first will be eliminated by band pass filter (BPF) or high pass filter (HPF) 351 from the unwanted low-frequency component of the picture signal of image processing part 34, and AF be controlled required preset frequency composition take out, and supply to ABS circuit 352.In addition, utilizing band pass filter (BPF) or high pass filter (HPF) 351 to carry out the preset frequency composition that frequency is selected, is predefined a plurality of frequency content, exports side by side according to each frequency content.

ABS circuit 352, for from each of a plurality of preset frequency compositions of the picture signal of above-mentioned BFP (or HPF) 351 output, carry out the absolute value signal that obtains being supplied to horizontal peak test section 353 with turn back signal and obtain the processing of absolute value of a certain fixing level.Horizontal peak test section 353, based on divided frame information from divided frame input part 355, according to each divided frame and each horizontal line, the peak value of the absolute value signal of the preset frequency composition of detected image signal.At this, described divided frame information for example is expressed as partition with 1 picture and is set to 5 row, 3 row and amounts to 15 divided frames (split image zone) for example as shown in figure 13.Certainly, the number of divided frame and configuration are not limited to the example of Figure 13.This divided frame can utilize microcomputer 37 to preestablish.

Vertical addition portion 354 is based on the divided frame information from divided frame input part 355, to each of each divided frame, to from each of a plurality of preset frequency compositions of horizontal peak test section 353, in vertical direction with the horizontal peak addition.According to each divided frame each, and each vertical summation result who carries out of each preset frequency composition, as variable storage in register 356.Circuit from above-mentioned BPF (or HPF) to register 356 constitutes AF circuit 35.

The operation result (variable) of storage in this register 356, be fed into the comparing section 371 in the microcomputer 37, this with 1 frame before and the operation result (variable) in past before 2 frames compare, carry out the detection of frequency content and the appointment of peak for each of each divided frame.Directivity judging part 372 is according to comparing the frequency content of each divided frame that obtains and the testing result of peak in comparing section 371, judge that peak becomes big direction, move the position of control lens 31 by motor driver 38 to the direction of this judgement.

And, be stored in the operation result (variable) in the register 356, be fed into the operational part 373 in the microcomputer 37, be used for the computing of figure selecting, according to its operation result, from for example in the graphics table storage part 374 that constitutes by memory in advance a plurality of pixels of storage read the graphics table and to select a pixel to read graphics table.Promptly, in this embodiment 1, in AF circuit 35, in divided frame, compare with past data (variable), carry out the detection of frequency content or the appointment of peak thus, therefore, at this AF circuit 35, the testing result of the frequency content that is obtained by the volume image that is taken is read the parameter of the combination of figure etc. and is used as pixel, similarly, use the information of the peak that is obtained by the operation result starting position of reading the zone as both full-pixel.

The information that the pixel of selecting from graphics table storage part 374 is read figure (comprise both full-pixel read the starting position in zone) is output, supply to TG drive pattern selection portion 375, control outputs to the drive signal of imaging apparatus 32 from drive division 39, reads reading of figure with the pixel of selecting.

Next, the example that above-mentioned pixel is read the system of selection of graphics table is described in detail.Each view data of the divided frame that 5 row 3 as shown in figure 13 that obtain at AF circuit 35 are listed as is shown in Figure 14 (A), by the variable that detects and temporarily be stored in each divided frame that frequency content in the register 356 and peak value obtain in view of the above shown in Figure 14 (B), at this moment, operational part 373 at first is standardized as " 1.00 " with the maximum in this variable " 200 ", obtains the variable of each divided frame shown in Figure 14 (C).

Then, operational part 373 according to the bigger order of variable intermediate value of each divided frame shown in Figure 14 (C), is compared the kind of reading pixel count for example as shown in table 1 and the form of ratio, determines the kind of reading of each divided frame.

Table 1

The kind of reading (explanation)	Ratio
		4/4 (reading 4 pixels in 4 pixels)	1.00
4/8 (reading 4 pixels in 8 pixels)	0.50
		4/16 (reading 4 pixels in 16 pixels)	0.25
4/32 (reading 4 pixels in 32 pixels)	0.125
		4/64 (reading 4 pixels in 64 pixels)	0.0625
4/128 (reading 4 pixels in 128 pixels)	0.03125

Table 1 expression is read both full-pixel (reading 4 pixels in 4 pixels) and is scaled 1.00 (100%) the ratio of reading pixel count.

At this, shown in Figure 14 (C), the variable of the Main subrack of the 3rd row the 2nd row is " 1.00 " in 15 divided frames, is maximum, therefore, as shown in Table 1, read (that is, both full-pixel is read) as 4/4, the variable " 0.25 " that then will be worth the divided frame of second largest the 1st row the 1st row is compared with table 1, the ratio of option table 1 is (read 4 pixels in 16 pixels, 12/16 the extracts at interval) ratio of reading as this divided frame of 4/16 reading of " 0.25 ".Below with the ratio of reading of above similarly selecting each divided frame.But, when the variable after the standardization of each divided frame of Figure 14 (C) all is identical value, adjust the kind of reading pixel count, carry out uniform weighting, so that the whole pixel count of reading becomes predetermined fixed value.

Consequently, make variable and the comparative result of table 1 after the standardization of each divided frame shown in Figure 14 (C), the blank that will become to read the figure of kind shown in Figure 14 (D) according to the substitution of variable after each each divided frame standardization.

Then, shown in Figure 14 (E) signal, with the multiple graphics table contrast of storage in the storage part 374 of the blank of the figure of Figure 14 (D) and figure, adopt weighting nearest like graphics table.As multiple graphics table, for example comprise the form shown in Figure 15 (A), (B), (C).In the above-mentioned contrast, at first read the value of the identical divided frame of the maximum Main subrack of pixel count (4/4) position in the blank with figure and compare, with unanimity or read the immediate graphics table of pixel count as candidate with graphics table.Thus, in the graphics table of Figure 15 (A)～(C), Figure 15 (A) and (B) these two graphics tables as candidate.

Then, in the blank of figure, be conceived to read the kind of reading pixel count " 4/16 " of divided frame of the 1st row the 1st row of pixel count more than second, compare with the pixel count of reading of the divided frame of the same position of the graphics table candidate of Figure 15 (A), (B), comparatively a Jin Si graphics table is as candidate.Consequently, final decision and adopt the graphics table shown in Figure 15 (A).

Imaging apparatus 32 carries out pixel according to the graphics table shown in this Figure 15 (A) and reads.Like this,, only,, select only graphics table, can read the pixel of weighting simply according to the variable of each divided frame that obtains from AF circuit 35 by from pre-prepd a plurality of graphics tables according to present embodiment.

Next, the pixel of imaging apparatus 32 is read describe.Figure 16 represents the circuit system figure of an example of imaging apparatus 32.As shown in the drawing, imaging apparatus 32 is made of as the lower part one of a pixel: photodiode 321; Buffer amplifier 322, temporary transient storage carry out opto-electronic conversion by photodiode 321 and the electric charge that obtains, and as the signal output of voltage etc.; And V switch 323, will output on the output line of assortment in vertical direction from the signal of buffer amplifier 322.This imaging apparatus 32 is the rectangular cmos image sensors that its pixel arrangement become 4 row, 5 row.In addition, Figure 16 has only represented 4 row, 5 row totally 20 pixels for the ease of diagram, but in fact as indicated above, pixel count is a lot.

The V switch 323 of each pixel, select signal V_SW_1～V_SW_4 according to switch from vertical register 324, via transversely arranged holding wire, be switched on or switched off V switch 323 simultaneously, and the pixel column of the V switch of connecting 323 switches in turn according to direction from the top down with the pixel of delegation.And the V switch 323 of each pixel of same row is connected with known correlated-double-sampling (CDS) circuit 325 via the output line of longitudinal direction, and then is connected with H switch 326 via CDS circuit 325.

H switch 326 is located in every row of pixel group, by selecting signal H_SW_1～H_SW_5 to be switched on from the switch of horizontal register 327 or disconnect, and for example is switched into from left to right connection.Vertical register 324 and horizontal register 327 are controlled by the control signal from the outside.And H switch 326 is connected with output amplifier 328 jointly via holding wire, come free switch to select signal V_SW_1～V_SW_4 and switch to select the signal of the selected pixel of signal H_SW_1～HSW_5 to be exaggerated, and output to the outside as image pickup signal at output amplifier 328.

At this, imaging apparatus 32 carries out common both full-pixel when reading, above-mentioned switch selects signal V_SW_1～V_SW_4 shown in Figure 17 (A)～(D), switch selects signal H_SW_1～HSW_5 shown in Figure 17 (F)～(J), (M)～(Q), timesharing is made as the L level (time is cut apart), is made as connection with V switch 323, H switch 326 corresponding during the L level.Thus, for example, during selecting signal V_SW_3 to be the L level as Figure 17 (E) switch that is shown in, in each pixel of 1 row of shown in Figure 17 (K), connecting, come the switch shown in free Figure 17 (F)～(J) to select the signal of each pixel of signal H_SW_1～H_SW_5 selection to be output in turn by V_SW_3.And, shown in Figure 17 (L), during switch selects signal V_SW_4 to be the L level, in each pixel of 1 row of connecting by the V_SW_4 shown in Figure 17 (R), come the switch shown in free Figure 17 (M)～(Q) to select the signal of each pixel that signal H_SW_1～HSW5 selects to be output in turn.

To this, as present embodiment, be weighted when reading, during for example the switch shown in Figure 18 (A) selects signal V_SW_3 to be the L level, in each pixel of 1 row of connecting by V_SW_3, select signal H_SW_1～HSW_5 by the switch shown in this figure (B)～(F), begin to select the 1st, the 2nd pixel from the left side, the pairing H switch 326 of the 3rd, the 4th pixel keeps disconnecting and not selected, the 5th pixel is selected, is output in turn from the signal of selected pixel.

At this, the H switch that is connected with horizontal direction the 5th pixel that horizontal direction the 2nd pixel of ining succession is read out from Figure 18 (C), (F) as can be seen, when the transversal switch that is connected with horizontal direction the 2nd pixel is switched to firm disconnection, is switched on.Consequently, from 15 pixels of going of being selected signal V_SW_3 selection by switch, shown in Figure 18 (G), extract the output from 2 pixels at interval, output is from the signal of remaining 3 pixels.

Switch shown in Figure 18 (H) select signal V_SW_4 be the L level during page or leaf with above similar, select H_SW_3 and H_SW_5 among signal H_SW_1～HSW_5 to become the L level in turn by being controlled to be the switch that makes shown in this figure (I)～(M), thus from 15 pixels of going of selecting signal V_SW_4 selection by switch, shown in Figure 18 (N), extract output at interval, and output is from the signal of 2 pixels of the residue corresponding with H_SW_3 and H_SW_5 from 3 pixels.In addition, above-mentioned switch selects signal V_SW_1～V_SW_4 and switch to select the output timing of signal H_SW_1～HSW_5, controls based on the pixel read output signal corresponding with the graphics table of exporting from TG drive pattern selection portion 375.

In addition, in the present embodiment,, do not need the mechanical shutter that is used for reading continuously regardless of pattern, and, read pixel count and fix, therefore driving frequency is fixed.

(embodiment 2)

Secondly, the embodiment 2 to the 2nd execution mode describes.This embodiment 2 is that the cameraman specifies above-mentioned motor pattern in advance, and EIS circuit (be equivalent to Fig. 1 251) carries out the embodiment of mobile vector when detecting.The block diagram of the embodiment 2 of the camera head of the present invention when Figure 19 has represented to specify motor pattern.In the figure, the part with the identical formation of Figure 12 is marked with same numeral, omits its explanation.

In Figure 19, image processing part 34 has interpolating circuit 341, Y/C processing/gamma correction portion 342, and the brightness signal Y of exporting from Y/C processing/gamma correction portion 342 is fed into EIS circuit 40.EIS circuit 40 by prefilter (BPF) 401, carries out frequency to the preset frequency composition of brightness signal Y and selects, and supplies to extraction circuit 402 at interval.Interval extraction circuit 402, in each picture area corresponding with each divided frame of described for example 5 row 3 row, in order in the representative point memory 403 of limited memory span, to store representative point, after the interval that the data of the preset frequency composition of input luminance signal are used to cut down data volume is extracted and is handled, to each point in a plurality of points (representative point) predetermined in each divided frame, with the storage of this point (representative point) in representative point memory 403.At this, above-mentioned point (representative point) for example uniformly-spaced is provided with 48 respectively with what 8 row 6 were gone regularly in each divided frame.

Then, EIS circuit 40, in comparison circuit 404, the data of a plurality of points of a certain vertical scanning period (n-1) V in the representative point memory 403 will be stored in, compare with the data compare 1 nV after the vertical scanning period with it in from a plurality of points of interval extraction circuit 402, according to this comparative result, in consistent address efferent 405, obtain monochrome information and which point which to put the mobile variable quantity (directions X, Y direction) and the Vector Message of change direction to from, as the output of EIS circuit 40, supply to microcomputer 37.

At this, further describe the action of above-mentioned comparison circuit 404 and consistent address efferent 405 with reference to Figure 20.Comparison circuit 404 is in each divided frame, for the variation of the value of the data before and after a plurality of points (representative point) in each divided frame output 1V.Consistent address comparison circuit 405 for these a plurality of points (representative point), respectively according to the variation of the value of the data before and after the 1V, calculates the direction and the value of the mobile vector of monochrome information.

For example, point (representative point) during during (n-1) V, shown in Figure 20 (A), be positioned at the position shown in the coordinate (3,4), during next among the nV, when representative point changes on the position shown in the coordinate (5,6) shown in this figure (B), consistent address efferent 405, comparative result according to the representative point of importing, obtain by computing: the mobile vector of the representative point shown in 45 among Figure 20 (B) (directions X, Y direction) is (2,2) (=(5-3,6-4)), its variable quantity is $\sqrt{8}=\left(\sqrt{2^2+2^2}\right)$ Information.Also handle equally for other divided frames.

At this, the direction of above-mentioned mobile vector or variable quantity comprise the information and because the information that cameraman's hand shake is produced that are produced by moving of the body that is taken.Hand shake is in picture on the representative point more than more than half, as identical direction vector and identical change amount and the information that produces and obtaining, therewith relatively, the information produce owing to moving of the body that is taken does not have this specific relation, therefore microcomputer 37 carries out computing that the data of whole representative points are averaged at operational part 376, and in integrating circuit 377, obtain the data of every 1V being carried out the representative point of integration according to each divided frame, it is supplied to operational part 376, get its difference.Thus, remove the information of hand shake by the direction of above-mentioned mobile vector or the information of variable quantity, mainly the mobile message (mobile vector direction, variable quantity) that moves each divided frame that information constituted that is produced by the body that is taken supplies to figure comparing part 379 from operational part 378.

Figure comparing part 379 is from the mobile message of each divided frame of importing, to the maximum of mobile vector (promptly, the mobile maximum body that is taken) distributes the first figure selecting method of weight greatly, or in the second graph system of selection of the big weight of body distribution that is taken that does not move, move by predefined any figure selecting method.When utilizing the action of the first figure selecting method, figure comparing part 379 is from the mobile message of each divided frame of input, obtain and to get the mobile vector maximum (promptly, move the maximum body that is taken) the variable standardization of divided frame be the variable of each divided frame after 1.00 after, below with the explanation of embodiment 1 similarly, select to be stored in the graphics table in the graphics table storage part 374.

On the other hand, when utilizing second graph system of selection action, for example to be the cameraman stay the state (state that be taken body and camera are always photographed over against ground) that the mode of the same position of picture is photographed with the body that is taken that always will move to the shooting state, in the mobile message of each divided frame that figure contrast Figure 37 9 will export, with the minimum value of getting mobile vector (promptly, move hardly, with camera over against the body that is taken) the variable standardization of divided frame be 1.00, carry out the processing that deducts the value of the variable in each divided frame from 1.00, after obtaining the variable of each divided frame, below with the explanation of embodiment 1 similarly, select to be stored in the graphics table in the graphics table storage part 374.

Like this, according to present embodiment, only from pre-prepd a plurality of graphics tables, according to the variable of each divided frame that obtains from EIS circuit 40, select only graphics table, the pixel that can be weighted is simply read thus.And present embodiment also is regardless of pattern, does not all need the mechanical shutter that is used for reading continuously, and then reads pixel count and fix, and therefore driving frequency is fixed.

Next, each example of pixel of the present invention being read figure with reference to Figure 21 describes.Figure 21 (A)～(C), shown in Fig. 5 (B), with 1 red light pixel R, be configured in locational 2 green light pixel G at mutual team diagonal angle, a blue light is arranged in totally 4 pixels of 2 row, 2 row with pixel B, be made as 1 pixel group as the unit of reading, and reading pixel count, to be 152 pixel this point be identical, but Figure 21 (A) is the example of weighting (reading pixel count) along with diminish before the neighboring area and then gradually from the central area (extraction yield becomes big gradually at interval), Figure 21 (B) expression does not use 4/6 pixel among Figure 21 (A) to read the example in zone, Figure 21 (C) expression is along with advancing to lower right area from top left region, tail off the gradually example of (extraction yield becomes big gradually at interval) of weighting (reading pixel count).

Promptly, among Figure 21 (A), the 1st zone at center (vertical address 9-12, horizontal address 9-12) reads the zone for both full-pixel, the 2nd zone (vertical address 7-14 (except 9-12) with its encirclement, horizontal address 7-14 (except 9-12)), be to read the zone with 4/8 pixel that the ratio that per 8 pixel groups are read 4 pixel groups is read, surround the 3rd zone (vertical address 5-16 (except 7-14) in the 2nd zone, horizontal address 5-16 (except 7-14)), be to read the zone with 4/6 pixel that per 6 pixel groups are read 4 pixel groups, the 4th remaining zone that surrounds the 3rd zone is to read the zone with 4/16 pixel that per 16 pixels are read 4 pixel groups.

And Figure 21 (B) expression is not used the 3rd zone of Figure 21 (A) and is read the figure of reading that the zone constitutes by 3.And, among Figure 21 (C), (vertically address 17-20, horizontal address 1-4) is that both full-pixel is read the zone in the 1st upper left zone, (vertically address 13-16, horizontal address 5-8) is that 4/8 pixel is read the zone in the 2nd zone of its bottom right, the 3rd zone of the bottom right in the 2nd zone (vertically address 9-12, horizontal address 9-12) is that 4/6 pixel is read the zone, and the 4th zone of the bottom right in the 3rd zone is that 4/16 pixel is read the zone.

Secondly, other execution modes that detection method and pixel are read figure describe.When having selected external setting-up, Figure 22 (A) expression cameraman reads the figure example.Externally setting is the irrespectively pre-prepd a plurality of figure projection instruments of image (graphics table) of any selection of cameraman and the body that is taken, the pattern that figure is read in decision, can consider the neighboring area in the split image zone as the low resolution zone, and the closer to the low more example of peripheral resolution: shown in Figure 22 (A1), split image not, carrying out 4/8 pixel from the both full-pixel zone reads, or shown in this figure (A2), the central area of picture is carried out 4/6 pixel and is read, its neighboring area is carried out 4/8 pixel and is read, or shown in this figure (A3), picture segmentation is become the central area, surround the 2nd zone of this central area, and the 3rd zone that surrounds the 2nd zone, wherein the central area is carried out both full-pixel and is read, the 2nd zone is carried out 4/8 pixel and is read, and the 3rd zone is carried out 4/16 and read etc.

Figure 22 (B) expression selects to read the example of figure based on optical zoom.When being taken body information detection optical zoom (zoom ratio), under the lower wide-angle state of zoom ratio, in order to make picture whole is equal resolution, make the ratio of reading average, can consider shown in Figure 22 (B1), divided frame not, carrying out 4/6 pixel from the both full-pixel zone reads, use under the state of optical zoom in the higher distally of multiplying power, shown in Figure 22 (B2), the high-resolution areas that read as both full-pixel in the split image zone of picture central authorities, and its area is less, in the split image zone of in addition periphery, select 4/16 pixel in the low resolution zone that extraction yield is bigger at interval to read examples such as figure.This is also identical for electronic zoom.

What Figure 22 (C) expression embodiment 2 was illustrated selects to read the example of figure based on the output of EIS circuit.Shown in the explanation of embodiment 2, when the maximum that goes out mobile vector by the EIS electric circuit inspection is picture central authorities, shown in Figure 22 (C1), the high-resolution areas of reading as both full-pixel with the split image zone of picture central authorities, and its area is less, and the split image zone of periphery in addition is along with advancing to periphery, read zone, 4/16 pixel according to 4/6 pixel and read the zone this becomes the zone of low resolution in proper order, select to read figure thus.

And, when the maximum that goes out mobile vector by the EIS electric circuit inspection is upper left regional of picture, shown in Figure 22 (C2), with Figure 21 (C) similarly, the high-resolution areas that read as both full-pixel in the 1st upper left zone of picture, and its area is less, and the split image zone of periphery in addition is along with advancing to the picture bottom right, read zone, 4/16 pixel according to 4/6 pixel and read the zone this becomes the zone of low resolution in proper order, select to read figure thus.

Figure 22 (D) expression as embodiment be 1 described, read the example of figure based on the output selection of AF circuit.As described in embodiment 1, frequency content of each divided frame (split image zone) and the variate-value that peak determined are when the maximum of the zone of picture central authorities, shown in Figure 22 (D1), the high-resolution areas that read as both full-pixel in the split image zone of picture central authorities, and its area is less, the split image zone of periphery in addition is along with advancing to periphery, read zone, 4/16 pixel according to 4/8 pixel and read the zone this becomes the zone of low resolution in proper order, select to read figure thus.

And, frequency content of each divided frame (split image zone) and the variate-value that peak determined are in the upper left zone of picture when maximum, shown in Figure 22 (D2), the high-resolution areas that read as both full-pixel in the 1st upper left zone of picture, and its area is less, the split image zone of periphery in addition is along with advancing to the picture bottom right, reads zone, 4/16 pixel according to 4/8 pixel and reads the zone this becomes the zone of low resolution in proper order, selects to read figure thus.

Figure 22 (E) expression expression utilizes the face detection circuit such as face detection circuit 253 of Fig. 1 to detect colour of skin part by the volume image that is taken, detect and whether to exist in the zone of broca scale picture and the relevant lower part of luminance level such as eyes, detecting when existing, with this area of skin color detection is the part of face, when the zone of detected face part is positioned at the bottom right of picture, shown in Figure 22 (E1), the high-resolution that the picture lower right area is read as both full-pixel, and its area is the size that comprises about detected face part, the split image zone of periphery in addition is along with advancing to picture is upper left, read the zone according to 4/6 pixel, 4/8 pixel is read the zone, 4/16 pixel is read the zone, and this becomes the zone of low resolution in proper order, selects to read figure thus.

And, when the zone of the detected face of face detection circuit part is positioned at the lower-left of picture or bottom right, shown in Figure 22 (E3), the high-resolution that read as both full-pixel in each zone of picture lower-left and picture bottom right, and its area is the size that comprises about detected face part, the split image zone of periphery in addition is along with advancing to picture is upper right and upper left, read zone, 4/16 pixel according to 4/8 pixel and read the zone this becomes the zone of low resolution in proper order, select to read figure thus.The pixel count of reading of reading figure shown in Figure 22 (E1)～(E3) all is identical.

In addition, be illustrated for the element that can take high precise image or superelevation precise image with imaging apparatus.The pixel count of imaging apparatus also can be the pixel count beyond high-precision thin image or the superelevation precise image, also can use the imaging apparatus of other pixel counts to constitute the camera head with above-mentioned effect.

Claims (8)

1. image capture method, from whole photoelectric conversion pixels zone of camera watch region that can random-access imaging apparatus, select the photoelectric conversion pixel zone more than 2, from the photoelectric conversion pixel zone more than 2 of this selection, read picture element signal, obtain image pickup signal, have two-dimensional arrangements in the camera watch region of described imaging apparatus and become rectangular a plurality of photoelectric conversion pixels zone, this image capture method is characterised in that, may further comprise the steps:

First step, make figure, this figure is divided into N split image zone with described camera watch region, and to described each split image zone, the detection information that obtains according to the shooting information of utilizing selected detection method from described imaging apparatus, set the weighting ratio of photoelectric conversion pixel number of regions with respect to the ratio of the whole photoelectric conversion pixel number of regions in this split image zone of reading of expression read output signal, wherein N is any natural number more than 1;

Second step is according to the described imaging apparatus of described graphics driver;

Third step, according to first picture element signal that read in the photoelectric conversion pixel zone of reading in described camera watch region, interpolation generates second picture element signal in the photoelectric conversion pixel zone of not reading in the described camera watch region; And

The 4th step, synthetic described first picture element signal and described second picture element signal, and with predetermined frame frequency output image pickup signal.

2. image capture method, from whole photoelectric conversion pixels zone of camera watch region that can random-access imaging apparatus, select the photoelectric conversion pixel zone more than 2, from the photoelectric conversion pixel zone more than 2 of this selection, read picture element signal, obtain image pickup signal, have two-dimensional arrangements in the camera watch region of described imaging apparatus and become rectangular a plurality of photoelectric conversion pixels zone, this image capture method is characterised in that, may further comprise the steps:

First step, the detection method of utilize selecting obtains the detection information that the shooting information according to described imaging apparatus obtains;

Second step, according to each a plurality of figure of described detection information issuing and be stored in it in storage device in advance, described a plurality of figure has been set the weighting ratio of the ratio of the whole photoelectric conversion pixel number of regions in N the split image zone of reading the photoelectric conversion pixel number of regions and obtaining with respect to cutting apart described camera watch region of expression read output signal, according to the described detection information that in described first step, obtains according to described split image zone, with reference to described storage device, figure like from described a plurality of figures of storage, selecting recently, wherein N is any natural number more than 1;

Third step, the described figure according to selecting in described second step drives described pixel element;

The 4th step, according to first picture element signal that read in the photoelectric conversion pixel zone of reading in described camera watch region, interpolation generates second picture element signal in the photoelectric conversion pixel zone of not reading in the described camera watch region; And

The 5th step, synthetic described first picture element signal and described second picture element signal, and with predetermined frame frequency output image pickup signal.

3. image capture method according to claim 1 and 2 is characterized in that,

The described figure of making or selecting, it is the figure that constitutes by zone of first split image more than 1 and the second split image zone more than 2, the described first image segmentation zone more than 1 is carried out the both full-pixel of the whole opto-electronic conversion image-region read output signals in this split image zone and is read, reading of a part of photoelectric conversion pixel zone extracted with mutually different weighting proportional spacing the whole photoelectric conversion pixels zone in this split image zone in the described second split image zone more than 2, the sum in the photoelectric conversion pixel zone that the photoelectric conversion pixel zone in the described first split image zone and the carrying out in the described second split image zone more than 2 are respectively read, irrelevant with the figure of making, be the predetermined fixed value that the number than whole photoelectric conversion pixels zone of described camera watch region lacks.

4. image capture method according to claim 1 and 2 is characterized in that,

Described detection information is any in the following information: the mobile message of the body that is taken behind the hand jitter correction; Zoom ratio information; Mobile message based on the mobile vector of the body that is taken; Static information based on the mobile vector of the described body that is taken; The characteristic frequency information of the image pickup signal when focusing is consistent automatically; And be considered as from face's information of the face that described image pickup signal obtains.

5. a camera head is characterized in that, comprising:

Can random-access imaging apparatus, have two-dimensional arrangements at its camera watch region and become rectangular a plurality of photoelectric conversion pixels zone;

The graphic making unit, make figure, the detection information that obtains according to the shooting information of utilizing selected detection method from described imaging apparatus in this figure, described camera watch region is divided into N split image zone, and to described each split image zone, set the weighting ratio of photoelectric conversion pixel number of regions with respect to the ratio of the whole photoelectric conversion pixel number of regions in this split image zone of reading of expression read output signal, wherein N is any natural number more than 1;

Driver element is according to the described imaging apparatus of described graphics driver;

The interpolation unit, according to first picture element signal that read in the photoelectric conversion pixel zone of reading in described camera watch region, interpolation generates second picture element signal in the photoelectric conversion pixel zone of not reading in the described camera watch region; And

Image formation unit, synthetic described first picture element signal and described second picture element signal generate image pickup signal and supply to described graphic making unit with predetermined frame frequency.

6. a camera head is characterized in that, comprising:

Can random-access imaging apparatus, have two-dimensional arrangements at its camera watch region and become rectangular a plurality of photoelectric conversion pixels zone;

Detection information obtains the unit, and the detection method of utilize selecting obtains the detection information that the shooting information according to described imaging apparatus obtains;

Storage device, store in advance according to each a plurality of figure of described detection information issuing and with it, described a plurality of figure has been set the weighting ratio of the ratio of the whole photoelectric conversion pixel number of regions in N the split image zone of reading the photoelectric conversion pixel number of regions and obtaining with respect to cutting apart described camera watch region of expression read output signal, and wherein N is any natural number more than 1;

Selected cell, according to the described detection information that obtains according to described split image zone, with reference to described storage device, figure like the selection recently in described a plurality of figures of from this storage device, storing;

Driver element, the described figure according to selecting drives described pixel element;

The interpolation unit, according to first picture element signal that read in the photoelectric conversion pixel zone of reading in described camera watch region, interpolation generates second picture element signal in the photoelectric conversion pixel zone of not reading in the described camera watch region; And

Image formation unit, synthetic described first picture element signal and described second picture element signal, and with predetermined frame frequency output image pickup signal.

7. according to claim 5 or 6 described camera heads, it is characterized in that,

The described figure of making or selecting, it is the figure that constitutes by zone of first split image more than 1 and the second split image zone more than 2, the described first image segmentation zone more than 1 is carried out the both full-pixel of the whole opto-electronic conversion image-region read output signals in this split image zone and is read, reading of a part of photoelectric conversion pixel zone extracted with mutually different weighting proportional spacing the whole photoelectric conversion pixels zone in this split image zone in the described second split image zone more than 2, the sum in the photoelectric conversion pixel zone that the photoelectric conversion pixel zone in the described first split image zone and the carrying out in the described second split image zone more than 2 are respectively read, irrelevant with the figure of making, be the predetermined fixed value that the number than whole photoelectric conversion pixels zone of described camera watch region lacks.

8. camera head according to claim 6 is characterized in that,

It is in the following information any that described detection information obtains the obtained described detection information in unit: the mobile message of the body that is taken behind the hand jitter correction; Zoom ratio information; Mobile message based on the mobile vector of the body that is taken; Static information based on the mobile vector of the described body that is taken; The characteristic frequency information of the image pickup signal when focusing is consistent automatically; And the face's information that is considered as face that obtains from described image pickup signal.

Images