CN100425062C - Spatial information detector - Google Patents

Abstract

A spatial information determining apparatus capable of precisely determining information of a target space. This apparatus comprises a plurality of photoelectric converting parts that receive a reflected light from a space to which a flash light is irradiated; a charge storing part formed, on each photoelectric converting part, by applying a control voltage to a plurality of electrodes provided on each photoelectric converting part; control means for controlling the number of electrodes, to which the control voltage is to be applied, such that the area of the charge storing part is different based on the flash period of the flash light; and an amplitude image generating part for generating an amplitude image having pixel values each of which is the difference between the charge stored by a charge storing part during the light up of the flash light and the charge stored by another charge storing part during the light out of the flash light.

Description

Spatial information detecting apparatus

Technical field

The present invention relates to spatial information detecting apparatus, be used for the relevant information that detects object space from the reverberation of object space by receiving, wherein just will inject this object space through the illumination of intensity modulated.

Background technology

In recent years, practice has been used by illumination being injected object space and receiving existence or the profile of object and the equipment of the spatial information the shape that detects object in object space from the reverberation of object space in various technical fields (for example, realize in factory automation quality control, in the safety system and the family expenses TV PABX Telephone System at airport and station).

For example, Japanese Patent Laid Open Publication [kokai] No.2001-148808 discloses a kind of solid-state image pickup apparatus, and it has the ability that the picture rich in detail of object is provided by the influence of eliminating surround lighting.This solid-state image pickup apparatus has two CCD, is used to store the electric charge about a photodiode.In each CCD, be stored in the electric charge that receives in the light period of luminous component and the dark period, the influence of using the difference signal between them to come the cancellation surround lighting then.

In addition, international open No.2004/090994 discloses a kind of spatial information detecting apparatus, and it has excellent S/N ratio and the aperture ratio that improves.This spatial information detecting apparatus is equipped with: optical-electrical converter is used for generating and the corresponding quantity of electric charge of reception light intensity; The electric charge that forms in the optical-electrical converter is selected the zone, is applied to electrode on the optical-electrical converter by controlling voltage, is chosen in the electric charge that generates in the optical-electrical converter; Electric charge is discharged (ejecting) parts, is used to export self charge to select the electric charge in zone; And sensitivity controller, be used for control control voltage, select the size in zone to change electric charge.

Yet, receiving these viewpoints that influences that light is exported the operational stability of checkout equipment under the saturated particular surroundings and simplified the labyrinth of conventional detection devices and eliminate surround lighting from the detection efficiency, the method for development information that effective evaluation detects, the raising that improve spatial information, above-mentioned traditional equipment also has very big room for improvement.

Summary of the invention

Therefore, main contents of the present invention provide a kind of spatial information detecting apparatus, and it has by using and existing simple device structure is accurately detected the ability of spatial information, and can reduce the influence of surround lighting effectively.

That is, spatial information detecting apparatus of the present invention comprises:

At least two optical-electrical converters are configured to receive the light from object space, and the illumination that wherein just will carry out intensity modulated with the modulation signal with preset frequency injects this object space, and generate and receive the corresponding electricity output of light intensity;

At least one electrode is formed on each optical-electrical converter;

The charge storage parts that form in each optical-electrical converter by apply control voltage at least one electrode are used for accumulating at least a portion of the electric charge that optical-electrical converter generates;

Controller, be configured to control the control voltage that applies at least one electrode, thereby the area of charge storage parts is different with the area of charge storage parts in another phase cycle at modulation signal in one of them of two different phase cycles of modulation signal;

Electric charge is discharged parts, is configured to export the electric charge of assembling in the charge storage parts; And

Evaluation unit, be configured to according to the electric charge of in of two different phase cycles, assembling by the charge storage parts that in of at least two optical-electrical converters, form, and the electric charge in another phase cycle, assembled by the charge storage parts that in another of at least two optical-electrical converters, form between difference, come the evaluation objective space.

According to the present invention, because can change the area of the charge storage parts that in optical-electrical converter, form by the control voltage that control applies to electrode, so compare with the situation that aperture incides the light quantity on the optical-electrical converter with adjustment with using shutter, might provide simple relatively structure to spatial information detecting apparatus.In addition, with by using single optical-electrical converter, carry out estimating then and the situation of assembling electric charge in each of two different phase cycles of modulation signal is compared, might carry out the detection and the evaluation of spatial information more efficiently, this is because can disposablely provide, also estimate then the electric charge of assembling in two different phase cycles of modulation signal.

In a preferred embodiment of the invention, at least two optical-electrical converters receive the light from an object space, wherein just passage of scintillation light are being shone into this object space.The control voltage that controller control applies at least one electrode, thus the area of charge storage parts is different from the area of charge storage parts in the dark period of passage of scintillation light in the light period of passage of scintillation light.Evaluation unit is by using by the electric charge of assembling in the light period of charge storage parts in passage of scintillation light that forms in of optical-electrical converter, and by the difference between the electric charge of assembling in the dark period of described charge storage parts in passage of scintillation light that forms, coming the evaluation objective space in another of optical-electrical converter.In the case, preferred controller control control voltage makes the area of the charge storage parts that form in each optical-electrical converter, with the scintillation time synchronous change of passage of scintillation light.

In addition, preferably, controller control control voltage, make in of optical-electrical converter the area of the charge storage parts that form, bigger in dark period in the light period internal ratio, and, the area of the charge storage parts that form in another of optical-electrical converter is bigger in light period in the dark period internal ratio.

In addition, preferably, controller control control voltage makes the area of the charge storage parts that form in light period, in an optical-electrical converter, equals the area of the charge storage parts that form in dark period, in another optical-electrical converter.

The evaluation unit of preferred aforesaid space information detecting apparatus comprises the magnitude image maker, is configured to generate the magnitude image with pixel value, and wherein each pixel value is provided by above-mentioned difference.In the case, might obtain to have emphasized the image of object, especially not comprise the image of the object of background with respect to background.Therefore, the shape and size of object have effectively been discerned.

Except that the magnitude image maker, spatial information detecting apparatus preferably also comprises the gray level image maker, the gray level image maker, be configured to generate gray level image with pixel value, wherein each pixel value is provided by following: the quantity of electric charge that described charge storage parts are assembled in the light period of passage of scintillation light and the one-period in the dark period, or the mean value of the quantity of electric charge of being assembled in the two at light period and dark period by described charge storage parts.

According to this structure, except that magnitude image, also might obtain to have considered the gray level image of the light quantity that provides from object space.In addition because each pixel value in the magnitude image is all relevant with the same position in the object space with each pixel value in the gray level image, all might eliminate object area, and easily generate the image only have powerful connections.In addition because can generate gray level image by the reception light output of discharging parts and provide from electric charge is provided, with the acquisition magnitude image, thus might obtain efficiently gray level image and magnitude image the two.

The spatial information detecting apparatus of another preferred embodiment also comprises according to the present invention: the Characteristic Extraction parts are configured to according to the characteristic quantity that is extracted the object in the object space by the magnitude image of magnitude image maker generation; The similitude calculating unit is configured to calculate similarity degree by characteristic quantity is compared with the previous template of preparing; And the target identification parts, be configured to when similarity degree is not less than predetermined value, object identification is the target object corresponding to template.

When object to be detected is face, preferably, spatial information detecting apparatus also comprises the template stores parts, be configured to store according to the characteristic quantity of face and pre-prepd face template, and when the similitude between the characteristic quantity that is extracted by the Characteristic Extraction parts and the face template of storing in the template stores parts was not less than predetermined value, the object identification parts were people corresponding to face template with this face recognition.In the case, following advantage is arranged: carry out face recognition reliably, and do not have the influence of surround lighting.

The spatial information detecting apparatus of another preferred embodiment also comprises according to the present invention: saturated definite parts are configured to predetermined threshold is compared with the quantity of electric charge of being assembled at least one of two of modulation signal different phase cycles by the charge storage parts; And output regulates parts, is configured to regulate electricity output corresponding to the receiving light power degree according to comparative result.

According to this structure, when determining the quantity of electric charge greater than threshold value, promptly receive light output when saturated by saturated definite parts, the electricity output that parts reduce optical-electrical converter is regulated in output, receives light output to reduce.Estimating under the situation of spatial information according to the difference that receives light output, receiving light and export when saturated when one that is used for determining difference, becoming is difficult to obtain important spatial information.Especially, when the output of two reception light was all saturated, difference was 0.Structure, target can not differentiate from background.Yet, according to the present invention, even when using this spatial information detecting apparatus under when the surround lighting that has increase, as outdoor environment, also determine to receive the saturation condition of light output, and parts prevent to receive light output according to saturated definite result saturated generation is regulated in output by saturated definite parts.Therefore, even taking place also might stably to extract important spatial information under the saturated environmental condition of reception light output from the difference that receives light output.As hereinafter described, at least one in component controls optical-electrical converter, light source and the light-receiving optical system regulated in especially preferred output, to prevent to receive the saturated of light output.

In addition, preferred spatial information detecting apparatus of the present invention comprises saturated definite parts, be configured to predetermined threshold is compared with the quantity of electric charge of being assembled at least one of described two different phase cycles of modulation signal by the charge storage parts, and, wherein when the quantity of electric charge during greater than threshold value, evaluation unit replaces described difference to come the evaluation objective space by using preset difference value.In the case, the maximum by the difference that use to allow might differentiate the zone of saturation of the output of the reception light in the magnitude image as preset difference value from background.

In addition, preferred spatial information detecting apparatus of the present invention comprises: saturated definite parts, be configured to predetermined threshold with comparing corresponding to the quantity of electric charge of assembling in each of described two different phase cycles on cycle memory time of a plurality of circulations of described modulation signal, described modulation signal; And output adjusting parts, be configured to according to comparative result, regulate corresponding to the electricity of receiving light power degree by change cycle memory time and export.In the case, when receiving the light quantity increase, might reduce cycle memory time, and therefore obtain improved response.On the other hand, when receiving the light quantity minimizing, prolong cycle memory time, but it effectively suppresses noise level.Cycle memory time equals the product of the duration of each phase cycle and the quantity of circulation (quantity of accumulation operation).Replacedly, preferably, the quantity of electric charge that saturated definite parts also are configured to assemble in each of two different phase cycles in the circulation of predetermined threshold and modulation signal, modulation signal compares; And output regulates parts, is configured to according to comparative result, regulates electricity output corresponding to the receiving light power degree by at least one the duration that changes two different phase cycles.

From following the preferred embodiments of the present invention, will be expressly understood other features of the present invention and effect.

Description of drawings

Fig. 1 is the block diagram according to the spatial information detecting apparatus of the first embodiment of the present invention.

Fig. 2 A and 2B are the operating instruction figure of the light receiving element that uses in spatial information detecting apparatus.

Fig. 3 A and 3B are flash cycle (flash cycle) that light source is shown respectively and the figure that receives the catoptrical cycle.

Fig. 4 A and 4B are the schematic diagrames that is illustrated in the light period of light source and the charge storage parts in the dark period.

The figure of the output of light receiving element when Fig. 5 is the Strength Changes that illustrates when surround lighting.

Fig. 6 A and 6B are the figure that the flash cycle of light source is shown respectively and receives the catoptrical cycle.

Fig. 7 A and 7B be second modification according to first embodiment is shown, at the schematic diagram of the light period of light source and the charge storage parts in the dark period.

Fig. 8 A and 8B be the 3rd modification according to first embodiment is shown, at the schematic diagram of the light period of light source and the charge storage parts in the dark period.

Fig. 9 is the block diagram of spatial information detecting apparatus according to a second embodiment of the present invention.

Figure 10 A is the example of gray level image, and Figure 10 b is the magnitude image (amplitude image) of Figure 10 A, and Figure 10 C is the amplitude error image of Figure 10 A, and Figure 10 D is the amplitude gradient direction image of Figure 10 A.

Figure 11 A and 11B illustrate the explanatory of determining the method for amplitude gradient direction image from magnitude image.

Figure 12 A is that the explanation view of extraction as the method for the characteristic points of faces of examined object is shown to 12C.

Figure 13 A is that the explanation view that extracts the method for characteristic quantity from amplitude gradient direction image is shown to 13C.

Figure 14 is the block diagram according to the spatial information detecting apparatus of third embodiment of the invention.

Figure 15 A is the schematic diagram that the image that is obtained by the spatial information detecting apparatus of the 3rd embodiment is shown to 15D.

Figure 16 is the flow chart of operation that the spatial information detecting apparatus of the 3rd embodiment is shown.

Embodiment

＜the first embodiment 〉

As shown in Figure 1, the spatial information detecting apparatus of first embodiment of the invention mainly comprises: light source 1 is used for flash irradiation is advanced object space; Imageing sensor 2 forms by arrange a plurality of light receiving elements 20 that are used to receive from the light of object space on semiconductor substrate; Control circuit 3 is used to control light source 1 and imageing sensor 2; Evaluation unit 4 is used for coming the evaluation objective space according to the output of imageing sensor 2, and this evaluation unit 4 comprises and is used to the gray level image maker 42 that generates the magnitude image maker 40 of magnitude image and be used to generate gray level image.In the accompanying drawings, Reference numeral 5 expression light-receiving optical systems by this system, will incide on the imageing sensor 2 from the light that object space provides.

For example, might use by arranging the array that a plurality of light-emitting diodes (LED) form or the combination of semiconductor laser and divergent lens in the plane and be used as light source 1.Can use infrared light or visible light to be used as from the light of light source 1 irradiation.Under the situation of using infrared light, because can use light source 1 without being noticed, so it is applicable to the purpose that stops crime, as security cameras at night.On the other hand, when using visible light, might obtain to approach the image of eye-observation.There is the modulation signal that has from the preset frequency of control circuit 3 output to operate light source 1.In the present embodiment, square wave is used as modulation signal.Scope from 10 to 100kHz is selected frequency.Duty ratio is 50%.Therefore, light source 1 provides a kind of flash of light, and its light period has the identical duration with dark period, and repeats them with the cycle alternation of 10 to 100 μ s.The short circulation that this flash period can not be discerned corresponding to human eye.Frequency and duty ratio are not limited to above-mentioned value.According to the type of spatial information to be detected and light source 1, can suitably determine frequency and duty ratio.

Each light receiving element 20 of imageing sensor 2 has: optical-electrical converter 22 is used to receive light from object space, and generates and corresponding electric output of institute's light intensity that receives; A plurality of electrodes 24 are formed on the optical-electrical converter 22; And by controlling that voltage is applied to electrode 24 and the charge storage parts 26 of in optical-electrical converter, inducting, so that accumulate at least a portion of the electric charge that generates in the optical-electrical converter; And electric charge discharge parts 28, be used for from charge storage parts 26 output charges.For example, shown in Fig. 2 A and 2B, each light receiving element 20 is equipped with: semiconductor layer 11, make by solid material such as doped silicon; Dielectric film 12 such as oxide-film, forms on whole general (general) surface of semiconductor layer 11; And electrode 24, form on semiconductor layer 11 by dielectric film 12.Known this type of light receiving element is the MIS device.Yet the light receiving element 20 of present embodiment is different from traditional MIS equipment part and is to form a plurality of electrodes 24 in the zone that act as single light receiving element 20.Dielectric film 12 and electrode 24 are made by the material with light transmission.When light incides on the semiconductor layer 11 by dielectric film 12, in semiconductor layer 11, generate electric charge.That is the optical receiving surface of light receiving element 20, is provided by the general surface (upper surface) of the semiconductor layer 11 shown in Fig. 2 A and the 2B.In the figure, the conduction type of semiconductor layer 11 is the n type.Therefore, the electric charge that is generated by rayed is an electronics.In above-mentioned light receiving element 20, when with positive controling voltage (+when V) being applied to electrode 24, in semiconductor layer 11,, form to assemble the electromotive force well 14 (depletion layer) of electronics in location corresponding to electrode 24.Promptly, in the time will controlling voltage and be applied to electrode 24 and light shine semiconductor layer 11 under with the situation that forms the electromotive force well, near the portions of electronics that generates electromotive force well 14 is caught in the electromotive force well 14, and since with the reconfiguring of the hole of semiconductor layer 11 depths, make the balance that generates electronics disappear.The electronics that generates in the position away from electromotive force well 14 is also owing to the reconfiguring of depths of semiconductor layer 11 disappears.Thus, when irradiates light, semiconductor layer 11 act as the optical-electrical converter 22 that generates electric charge, and electromotive force well 14 act as the charge storage parts 26 of assembling and keeping electric charge.

In addition, because forming electromotive force wells 14, so might control by the quantity that change is applied in the electrode of control voltage along the size of the electromotive force well 14 on the general surface of semiconductor layer 11 with the corresponding zone of the electrode 24 that is applied in control voltage.When the size (area) of electromotive force well 14 became big, the electric charge of assembling in the electromotive force well 14 also increased with respect to the ratio of the electric charge that generates in the semiconductor layer 11.In the present invention, as hereinafter described, because use the electric charge of in electromotive force well 14, assembling, so can reach the raising of sensitivity by the size that increases electromotive force well 14.In brief, can control size, thereby the sensitivity of optical-electrical converter 22 can be adjusted by changing the voltage that applies to electrode 24 as the electromotive force well 14 of charge storage parts 26.

For example, have in use under the situation of light receiving element 20 of five electrodes 24, when shown in Fig. 2 A, to three electrodes 24 of the inside apply control voltage (+when V) and not to the outside two electrodes 24 do not apply (0V), along the size of the electromotive force well 14 (being the charge storage parts) of optical receiving surface with shown in Fig. 2 B, (+the situation that V) and do not apply (0V) to four electrodes 24 of residue is compared, and becomes bigger only to apply control voltage to the electrode 24 of centre.Therefore, when shining identical light quantity, the quantity of electric charge of assembling in the electromotive force well shown in Fig. 2 A is bigger than the quantity of electric charge of assembling in the electromotive force well shown in Fig. 2 B.This means that optical-electrical converter has higher sensitivity.

Can all have the light receiving element 20 of said structure by on the lattice point of the two-dimensional square lattice pattern of preparing on single semiconductor substrate, arranging each, form imageing sensor 2.For example, can use the matrix array of 100 * 100 light receiving elements 20.In addition, in each row, integrally and continuously form semiconductor layer 11 in the vertical direction of the matrix array of light receiving element 20, and when arranging electrode 24 in vertical direction, semiconductor layer 11 can be used as the transfering channel of transfer charge in vertical direction with adjoining each other.Imageing sensor 2 also has the horizontal transfer parts that form in semiconductor substrate, and comprises CCD, is used for receiving electric charge, transfer charge in the horizontal direction then from the end of each row of extending in the vertical direction of semiconductor layer 11.

For the electromotive force well 14 from each light receiving element 20 (being charge storage parts 26) output charge, might use the technology identical with traditional C CD.That is, in electromotive force well 14, assemble after the electric charge, apply the mode of controlling voltage to electrode 24, shift the electric charge of assembling in the electromotive force well 14, thereby can export this electric charge from another electrode (not shown) that semiconductor layer 11 forms by control.The parts that are used for transfer charge can have and frame transfer CCD or the substantially the same structure of interline transfer CCD.For example, when it has the structure identical with frame transfer CCD, can be by the shape that changes electromotive force well 14 on the left side of Fig. 2 A or 2B or right transfer charge.In addition, when it has the structure identical with interline transfer CCD, can be along a left side and the right-hand CCD that always forms of Fig. 2 A or 2B.Electric charge is sent to CCD from electromotive force well 14, on the left side of Fig. 2 A or 2B and right, shift this electric charge by CCD then.

In addition, when electric charge discharge parts 28 have the structure identical with frame transfer CCD, come transfer charge to the mode that electrode 24 applies control voltage by control.In the case, may command is to the voltage that electrode 24 applies, and makes the electric charge of in electric charge tap-off cycle (it is different from the charge integration period of assembling electric charge in charge storage parts 26) output charge memory unit 26.Therefore, in the present embodiment, also can use electrode 24 to come transfer charge, and semiconductor layer 11 is also discharged parts 28 as electric charge.

In addition, be the imageing sensor 2 of easily understanding present embodiment, in Fig. 1, discharge parts, show the function of light receiving element 20 individually by optical-electrical converter, charge storage parts and electric charge.In addition, electric charge discharge parts 28 shown in Figure 1 comprise above-mentioned semiconductor layer 11 and horizontal transfer parts.In addition, discharge the common electrode 24 that uses in the parts 28 at optical-electrical converter 22, charge storage parts 26 and electric charge.

Control circuit 3 generates the control voltage to electrode 24 to be applied, and control the mode that applies control voltage to electrode, thereby adjust the timing of sensitivity (promptly the electric charge of assembling is with respect to the ratio of the electric charge that generates owing to illumination), formation charge storage parts 26 and discharge the timing of parts 28 in charge storage parts 26 in optical-electrical converter 22 from charge storage parts 26 output charges by electric charge.Promptly, apply the mode of control voltage and the timing that changes this mode by control to electrode, might be provided at the charge integration period of assembling electric charge in the charge storage parts 26, and discharge parts 28 by electric charge and discharge electric charges, output to the electric charge tap-off cycle of evaluation unit 4 as receiving light output then, and should the cycle define being different from the cycle of charge integration period from charge storage parts 26.

Form light-receiving optical system 5, on each light receiving element 20 that object space is incident upon imageing sensor 2.That is,, will be mapped to corresponding to the three dimensions of object space on the two dimensional surface that the arrangement by the light receiving element 20 of imageing sensor 2 provides by light-receiving optical system 5.Therefore, by light-receiving optical system 5, can will interrelate at object M and light receiving element 20 in the visual field of the unilateral observation of imageing sensor 2.

The gray level image maker 42 of evaluation unit 4 generates gray level image, and it comprises object M and background (for example, shown in Figure 10 A).On the other hand, the magnitude image maker generates magnitude image, emphasizes object M (for example, shown in Figure 10 b) with respect to background therein.Can use these images to check the shape and size of object M.In addition, the reflection of determining object M is available.Can determine according to the structure of evaluation unit 4 will be by the type of the spatial information that uses these images to detect.

The operation of the spatial information detecting apparatus of present embodiment then, is described.Light source 1 advances object space with flash irradiation, thereby alternately repeats light period and dark period, as shown in Figure 3A.In this embodiment, duration " Ta " of light period is set to equal the duration " Tb " of dark period.That is, the duty ratio of light modulated is 50%.Advance the light of object space and the object M reflection from object space from light source 1 irradiation, incide on the light receiving element 20 with the time of delay " Td " of the distance that depends on object M.In the case, because time of delay " Td " than fluorescent lifetime " Ta " or short many of fluorescent lifetime " Tb " not, so it can be ignored.

In the present embodiment, because export " Ab ", so the set of two adjacent light receiving elements 20 is counted as a pixel to the evaluation unit 4 disposable reception light output " Aa " of light period " Ta " and the reception light of dark period " Tb " of providing.That is, in light period " Ta ", be set to highly sensitive state,, and provide and receive light output so that in charge storage parts 26, assemble electric charge as one of optical-electrical converter of adjacent two light receiving elements 20 of a pixel.At this moment, by adjusting the control voltage that applies to electrode 24, another optical-electrical converter is set to the muting sensitivity state.On the contrary, at non-light period " Tb ", another optical-electrical converter of adjacent two light receiving elements 20 is set to the high sensitivity state, so that assemble electric charge in charge storage parts 26, and provides the output of reception light.At this moment, the control voltage that is applied to electrode (24) by adjustment is set to the muting sensitivity state with one of described optical-electrical converter.According to this mode, in the different charge storage parts 26 that in imageing sensor 2, form, remain on respectively with the corresponding charge integration period of light period " Ta " in the electric charge assembled and with the corresponding charge integration period of dark period " Tb " in the electric charge of gathering, and the disposable evaluation unit 4 that sends to of the electric charge that in single electric charge tap-off cycle, will in two charge storage parts, keep.

In the following description, each that is used as in two light receiving elements 20 of a pixel all has three electrodes.In addition, distinguish between the electrode for the electrode of a light receiving element therein and another light receiving element, specify these electrodes 24 by Reference numeral (1) to (6), shown in Fig. 4 A and 4B.That is, one of them of two light receiving elements 20 has electrode (1) to (3), and another light receiving element 20 has electrode (4) to (6).In the case, be preferably formed with as the relevant spill and leakage of the light receiving element 20 of a pixel.

Be the sensitivity of control optical-electrical converter 22, control has been applied in the quantity of the electrode 24 of control voltage, to change the area of electromotive force well 14 on the optical receiving surface.That is, in the light period " Ta " of light source 1, to all electrodes (1) to (3) apply positive controling voltage (+V), to obtain and the area of electrode (1), shown in Fig. 4 A to (3) corresponding electromotive force well 14.On the other hand, in this light period, only to the electrode placed in the middle (5) of the electrode (4) to (6) of another light receiving element 20 apply positive controling voltage (+V), to reduce the area of electromotive force well 14.In the case, provide the high sensitivity state of optical-electrical converter 22, and provide the muting sensitivity state of optical-electrical converter 22 with the corresponding zone of electrode (4) to (6) with the corresponding zone of electrode (1) to (3).Therefore, the amount by the electric charge (electronics " e ") newly-generated receiving light with electrode (4) to (6) corresponding zone is far fewer than the amount at the electric charge that generates to (3) corresponding zone with electrode (1).As a result, with receive light output " Aa " corresponding electric charge be collected at electrode (1) in the electromotive force well 14 in (3) corresponding zone.

In the dark period " Tb " of light source 1, to all electrodes (4) to (6) apply positive controling voltage (+V), to obtain and the area of electrode (4), shown in Fig. 4 B to (6) corresponding electromotive force well 14.In this dark period, only to the electrode placed in the middle (2) of the electrode (1) to (3) of another light receiving element 20 apply positive controling voltage (+V), to reduce the area of electromotive force well 14.In the case, provide the high sensitivity state of optical-electrical converter 22, and provide the muting sensitivity state of optical-electrical converter 22 with the corresponding zone of electrode (1) to (3) with the corresponding zone of electrode (4) to (6).Therefore, the amount by the electric charge (electronics " e ") newly-generated receiving light with electrode (1) to (3) corresponding zone is far fewer than the amount at the electric charge that generates to (6) corresponding zone with electrode (4).As a result, with receive light output " Ab " corresponding electric charge be collected at electrode (4) in the electromotive force well 14 in (6) corresponding zone.

As mentioned above, can under the situation shown in Fig. 4 A, assemble, and can under the situation shown in Fig. 4 B, assemble with the corresponding electric charge of dark period " Tb " with the corresponding electric charge of light period " Ta ".Thereby obtain a circulation of the situation of Fig. 4 A and 4B by the mode that control applies control voltage to electrode, the reception light that might obtain light period " Ta " and dark period " Tb " export (Aa, Ab).Yet, when the light quantity that sufficiently obtains in the circulation at only a light period " Ta " and dark period " Tb " to incide on the light receiving element 20, worry to cause that owing to the influence of the shot noise that takes place in the light receiving element receiving light exports (Aa, S/N ratio reduction Ab).In the case, preferably in the circulation of the situation that repeatedly repeats Fig. 4 A and 4B with after in charge storage parts 26, sufficiently assembling electric charge, by electric charge discharge parts 28 outputs receive light output (Aa, Ab).

After the electric charge of in the charge storage parts 26 of two light receiving elements 20 that are used as a pixel, having assembled the electric charge of generation in light period " Ta " respectively and in dark period " Tb ", having assembled, electric charge is discharged parts 28 and in the electric charge tap-off cycle two classes reception light is exported (Aa, Ab) the disposable evaluation unit 4 that sends to.This means and to pass through three cycles, promptly in light period " Ta ", assemble the integration period of electric charge, in dark period " Tb ", assemble the integration period of electric charge and the tap-off cycle of exporting the electric charge of in these integration period, assembling, obtain two classes receive light output (Aa, Ab).For preventing in the time cycle of the area of minimizing, to mix non-this situation of electric charge of wishing electric charge, form optical screen film in the cycle between can be at this moment, to cover the adjacent part with charge storage parts 26 corresponding electrodes 24 with charge storage parts 26.

In each of the situation of Fig. 4 of present embodiment A and 4B, the control voltage that will apply to three electrodes (1) to (3) or (4) to (6) (+control voltage that V) is set to equal to apply to electrode placed in the middle (2) or (5) (+V).Therefore, even when the area of electromotive force well 14 changes, the degree of depth of electromotive force well 14 also can keep constant substantially.Therefore, the electric charge of the adjacent portion office of the shielding between electromotive force well 14 (barrier) generation flows in the adjacent electromotive force well 14 equably.

By this mode, even when the area of electromotive force well is reduced to the muting sensitivity state that obtains optical-electrical converter 22 and shifts the electric charge of assembling in electromotive force well 14, also be mapped to light receiving element 20 and in charge storage parts 26, assemble electric charge by illumination.That is, in charge storage parts 26, be blended in the electronics that generates in time cycle of high sensitivity state of non-optical-electrical converter 22.In the present embodiment, can with the influence of unwanted electric charge (its in charge storage parts 26, generate) at the time durations of the electric charge of electric charge that keeps light period " Ta " and dark period " Tb " and with illustrate later, determining that the corresponding composition of surround lighting in the difference gets rid of.During when maintenance in charge storage parts 26 and from charge storage parts 26 transfer charges, optical-electrical converter 22 is in the muting sensitivity state.At this moment, because reduced the area of charge storage parts 26, the growing amount of unwanted electric charge diminishes.

In order to be easy to explain, suppose that the area of the quantity of electric charge of gathering in electromotive force well 14 and electromotive force well 14 is proportional.In addition, the area of electromotive force well 14 is big three times at the muting sensitivity state of optical-electrical converter 22 at the high sensitivity state ratio of optical-electrical converter 22.This means that the quantity of electric charge of assembling is bigger three times than the quantity of electric charge of assembling in the muting sensitivity state in the high sensitivity state.

When by (S) expression according to the light that provides from light source 1 with the corresponding electromotive force well 14 of single electrode when the quantity of electric charge assembled and the quantity of electric charge assembled according to surround lighting by (N) expression, by (S+N) be illustrated in the corresponding electromotive force well 14 of single electrode (5) in the total amount of electric charge assembled, shown in Fig. 4 A.On the other hand, by (3S+3N) be illustrated in the corresponding electromotive force well 14 of electrode (1) to (3) in the gathering the quantity of electric charge.Because the situation of Fig. 4 B is corresponding to dark period " Tb ", so the electric charge that the light that does not have basis to provide from light source 1 is assembled.Therefore, with the corresponding electromotive force well 14 of electrode (2) in the quantity of electric charge assembled equal the quantity of electric charge (N) assembled according to surround lighting.Similarly, with the corresponding electromotive force well 14 of electrode (4) to (6) in the quantity of electric charge assembled only be (3N).

For in magnitude image maker 40, generating magnitude image, as mentioned above, calculating by at the high sensitivity state (shown in Fig. 4 A, its with the corresponding zone of electrode (1) to (3) on form) electromotive force well 14 in assemble the reception light that electric charge provided output " Aa ", with by the high sensitivity state (shown in Fig. 4 B, its and the corresponding zone of electrode (4) to (6) on form) electromotive force well 14 in the reception light that electric charge provided assembled export difference between " Ab ".By (3S+4N) expression with receive light output " Aa " corresponding quantity of electric charge, its be the electric charge (3S+3N) light period " Ta " in, assembled with in the situation of Fig. 4 B with the corresponding electromotive force well 14 of electrode (2) in the unwanted electric charge (N) assembled with.On the other hand, by (S+4N) expression with receive light output " Ab " corresponding quantity of electric charge, its be the electric charge (3N) dark period " Tb " in, assembled with in the situation of Fig. 4 A with the corresponding electromotive force well 14 of electrode (5) in the unwanted electric charge (S+N) assembled with.(Aa, the Ab) difference between promptly, (3S+4N)-(S+4N)=2S, might be got rid of the influence of the unwanted electric charge (N) that is caused by surround lighting, and generate magnitude image by determining these to receive light output.

By this mode, shown in the curve E of Fig. 5, under the situation that hypothesis environment light intensity changes in time, this curve E is corresponding to the light intensity that incides in the dark period " Tb " of light source 1 on the optical-electrical converter 22.As a result, it is corresponding to the output of the reception light in dark period " Tb " " Ab ".Thus, because the reception light of the dark period of light source 1 " Tb " output " Ab " is corresponding to the height of curve E, so the reception light of the light period of light source 1 " Ta " output " Aa " becomes higher than curve E.That is, when alternately repeating light period and dark period, in light period " Ta ", the reception light of imageing sensor 2 is exported " Aa ", " Ab " than curve E height, and in dark period, equals the height of curve E.Provide and the corresponding reception light output of the light that shines object space from light source 1 by the zone that is higher than curve E.Therefore, by determining to export " Aa " and the difference (Aa-Ab) between the reception light output " Ab " that dark period " Tb " obtains at the reception light that light period " Ta " obtains, might only extract the composition that shines the light of object space from light source 1, and not have the influence of surround lighting.This difference (Aa-Ab) is relevant with the position of each light receiving element 1, thereby obtains magnitude image.

Export to determine difference (Aa-Ab) (among Fig. 5, be right after at light period " Ta " dark period " Tb " is provided afterwards) from the reception light of the reception light of light period " Ta " output " Aa " and the dark period " Ta " adjacent with light period " Ta ".In the case, suppose determine in needed light period of difference (Aa-Ab) " Ta " and dark period " Tb " are totally, the intensity of surround lighting E do not have material alterations.Therefore, in light period " Ta " based on the output of the reception light of surround lighting and dark period " Tb " in the reception light output based on surround lighting cancelled out each other, therefore only obtain with by exporting from the corresponding reception light of the reverberation of object M reflection to the object space irradiates light from light source 1.As a result, might obtain magnitude image, it is the image of emphasizing of object M.

On the other hand, gray level image maker 42 generates image, the reception light of the reception light of light period " Ta " output " Aa ", dark period " Tb " output " Ab " therein or these receive light output, and (Aa, mean value Ab) the position with each light receiving element 20 are relevant.In the case, because use the catoptrical light that is different from from object M to generate this image, might obtain to comprise the comprehensive gray level image of background and object M.By using the reception light output " Aa " that in the light period " Ta " of light source 1, obtains to generate the gray level image of present embodiment.

By this mode, need not go up optical-electrical converter 22 is maintained the high sensitivity state, as shown in Figure 3A in the whole cycle of light period " Ta " or dark period " Tb ".According to technological concept of the present invention, optical-electrical converter 22 can be set in light period " Ta " partly or dark period " Tb ", be in the high sensitivity state.In the case, assemble under the condition of electric charge in the constant time cycle, the duty ratio of light period " Ta " and dark period " Tb " may be different from 50%.In addition, when known to object M apart from the time, can consider to depend on the time of delay " Td " of the distance of object M and determine light period " Ta " and dark period " Tb ", shown in Fig. 6 A and 6B.That is, can after process time of delay (, opening and closing light source 1), light period " Ta " and dark period " Tb " be set, respectively so that in the optical-electrical converter 22 of high sensitivity state, assemble electric charge from this moment.In the case, compare with the control mode shown in the 3B, according to having reduced the quantity of electric charge of in charge storage parts 26, assembling time of delay " Td " with Fig. 3 A.Yet (Aa Ab), therefore gets rid of the influence of surround lighting reliably might to obtain the reception light output accurately corresponding with the light quantity that receives in light period " Ta " and dark period " Tb ".

In addition, when each light period " Ta " and dark period " Tb " being set to the relative short time cycle, and only carry out a circulation time of light period " Ta " and dark period " Tb ", be difficult to obtain enough reception light output required in the evaluation unit 4.Under these circumstances, electric charge of preferably being assembled by light period " Ta " repeatedly by charge storage parts 26 and the electric charge assembled by dark period " Tb " repeatedly by charge storage parts 26 similarly are used separately as and are received light output.As mentioned above, can adjust according to the control voltage that applies to electrode 24 and be used for assembling the integration period of electric charges and being used for discharging parts 28 are assembled electric charges to evaluation unit 4 outputs from charge storage parts 26 tap-off cycle by electric charge at charge storage parts 26.In the present embodiment, receive light output by obtaining two classes from following three time cycles: light period " Ta " in, assemble the integration period of electric charge, the integration period of gathering electric charge and the tap-off cycle of exporting the electric charge of in this two cycles, assembling dark period " Tb " in as two light receiving elements 20 of a pixel.As the modification of present embodiment, can be with single light receiving element as a pixel.In the case, preferably apply the mode of control voltage to electrode 24 by control circuit 3 controls, thereby the electric charge that repeats to assemble under the high sensitivity state by the area that increases charge storage parts 26 in the light period " Ta " of light source 1 is provided as the tap-off cycle to the reception light output of evaluation unit 4, and the electric charge that will assemble under the high sensitivity state by the area that increases charge storage parts 26 in the dark period " Tb " of light source 1 is provided as the tap-off cycle to the reception light output of evaluation unit 4.Therefore, by four different cycles, promptly, integration period, the output institute of assembling electric charge in light period " Ta " assembles the tap-off cycle of electric charge, assembles the integration period of electric charge and the tap-off cycle of electric charge that discharge is assembled dark period " Tb " in, obtains two classes reception light and exports.

As the modification of present embodiment, the area of electromotive force well 14 and the degree of depth can change.For example, shown in Fig. 7 A and 7B, preferably only be set to be higher than the control voltage (+V2, for example 3V) that applies to single electrode (2) or (5) to (3) or (4) to the control voltage (+V1, for example 7V) of (6) with being applied to three electrodes (1) simultaneously.In the case, the degree of depth of the depth ratio with large-area electromotive force well 14 with electromotive force well 14 of small size wants big.

By this mode, with do not apply the electric charge that generates on the electrode (1), (3) or (4), (6) corresponding zone of control voltage and trend towards flowing to electromotive force well 14 to it.In the case, believe that electric charge easily flows to the electromotive force well when electromotive force well 14 is darker.This means that a large amount of electric charges flow to the electromotive force well 14 of high sensitivity state.As a result, might be reduced in electrode (1), (3) or (4), (6) corresponding zone on the unwanted electric charge that generates flow to the probability of the electromotive force well 14 of muting sensitivity state.In brief, might further reduce the amount of the unwanted electric charge that flows to the electromotive force well 14 that is used to keep electric charge.

The 3rd modification as present embodiment, shown in Fig. 8 A and 8B, preferably the control voltage that applies to the electrode placed in the middle (5) of the electrode placed in the middle (2) of three electrodes (1) to (3) of each light receiving element 20 or three electrodes (4) to (6) is higher than the control voltage that electrode (1), (3) or (4), (6) to both sides apply, and goes up arranged light screened film 25 at electrode (2) placed in the middle or (5).In the case, because the existence of optical screen film 25, may generate electric charge hardly with electrode (2) or (5) corresponding zone, but allow with the electrode (1) of light receiving element 20, (3) or (4), (6) corresponding zone on the electric charge that generates flow to and electrode (2) or (5) corresponding zone.In addition, because partly covered light receiving element 20, so, generate electric charge hardly in the part of the electromotive force well 14 that is covered by optical screen film 15 by optical screen film 15.Therefore, under the condition that reduces the electromotive force well 14 that keeps electric charge, generate electric charge hardly, thereby reduce the probability that unwanted electric charge mixes as noise contribution significantly.

Thus, when forming with step-wise manner under the situation in dark zone and shallow zone, generating electric charge with electrode (1), (3) or (4), (6) corresponding location one in electromotive force well 14, they just move to and electrode (2) or (5) corresponding zone.Therefore, even when in several nanoseconds (ns) or short cycle still less, switching cycle that generates electric charge and the cycle that keeps electric charge, also might be reduced in the possibility of hybrid charge between the electromotive force well 14 that forms in the adjacent light receiving element 20, thereby reach the minimizing of noise contribution.Can use the method for the form that is different from optical screen film 15 to be used as forming the technology of electromotive force well 14 with step-wise manner.

In present embodiment and above-mentioned modification, the quantity of the electrode that forms on each light receiving element 20 is not limited to 3.In addition, applied the quantity that controls the electrode of voltage, can arbitrarily determine with high sensitivity state or muting sensitivity state.The imageing sensor 2 of present embodiment is made of the two-dimensional array of light receiving element 20.Replacedly, can use the one-dimensional array of light receiving element.Described in first modification, only using, a light receiving element 20 is not impossible.Also available is that the distance according to from electrode 24 beginnings along optical receiving surface forms the concentrated distribution of impurity in the semiconductor layer 11 of light receiving element 20, and can change the area of charge storage parts 26 by the voltage that control applies to electrode 24.

＜the second embodiment 〉

The spatial information detecting apparatus of present embodiment is characterised in that: from the clarification of objective amount of magnitude image extraction such as face; Compare with the template of previous preparation, with recognition objective accurately; And it is substantially the same with first embodiment except following assembly.Therefore, omit the repetition of explanation of same components.

In the spatial information detecting apparatus of this embodiment, as shown in Figure 9, Characteristic Extraction parts 50 extract the characteristic quantity of the object (people Ob) in the object space according to the magnitude image that is generated by magnitude image maker 40.On the other hand, in template preparation/memory unit 56, store face template.Face image by obtaining object to be identified (people Ob) in advance, extract characteristic quantity by the Characteristic Extraction parts from face image then, prepare this face template.The characteristic quantity of the object that the magnitude image that will be generated from actual measurement by Characteristic Extraction parts 50 is extracted compares with the face template of storage in the template preparations/memory unit 56, so that by the similarity degree of similitude calculating unit 52 calculating between them.In object identification parts 54, when the similarity degree that is calculated by similitude calculating unit 52 is not less than predetermined value, the object identification that is detected is the people corresponding to face template.Can switch the operator scheme of preparing template and the operator scheme of carrying out actual measurement by operator scheme selector (not shown).

The Characteristic Extraction parts 50 of this embodiment comprise: amplitude difference image composer (not shown), be used to generate the amplitude differential image that has as the amplitude difference value of pixel value, this pixel value is the difference intensity level according to each definite pixel of the amplitude of magnitude image; And image processing parts (not shown), being used to generate output image, it is by obtaining amplitude difference image binaryzation with predetermined threshold.Extract the characteristic quantity of object (that is face) from the output image that generates by the image processing parts.Figure 10 A is the example of gray level image.Figure 10 b is the magnitude image corresponding to this gray level image.Figure 10 C is the amplitude differential image of preparing from this magnitude image.

In the coordinate system that is used for magnitude image shown in Figure 10 b, the top-left position of magnitude image is an initial point, and the positive direction of x axle (x direction) is corresponding to level direction to the right, and the positive direction of y axle (y direction) is corresponding to vertically downward direction.In Characteristic Extraction parts 50, by the whole pixels of magnitude image being used Suo Beier (Sobel) filter (hx) of x directions and the Suo Beier filter (hy) of y direction (its each all have mask (mask) size of 3 * 3 pixels shown in Figure 10 b), carry out local space differentiation, thereby generate the amplitude differential image that has as the amplitude difference value of pixel, it is the difference intensity level according to each definite pixel of the amplitude of magnitude image.Can be from extracting the characteristic quantity of the face of people Ob by the amplitude differential image being carried out the output image that binaryzation obtains with threshold value.Arrangement in the weight factor on magnitude image shown in each Suo Beier filter of Figure 10 b.

When by | G (u, v) | (u, during the difference intensity level of v) locating, it is the value of determining around the pixel value (amplitude in the present embodiment) of 8 pixels of center pixel by in the magnitude image to the pixel of expression magnitude image.When shown in Figure 11 A, when using (a) to (i) expression to arrange the pixel value of the pixel (p1) to (p9) in the regional area (rectangular area) that provides respectively by 3 * 3 pixels with center pixel (p5), shown in Figure 11 B, come the differential expression intensity level by the difference value (dx) on the following use x direction and the equation (1) of the difference value (dy) on the y direction | G (u, v) |.

|G(u，v)|＝{(dx ²(u，v)+dy ²(u，v))} ^1/2 (1)

Wherein,

dx(u，v)＝(c+2f+i)-(a+2d+g) (2)

dy(u，v)＝(g+2h+i)-(a+2b+c) (3)

In having the amplitude differential image of being determined by equation (1) as the amplitude difference value of pixel value, when the amplitude difference in the magnitude image increased, it is big that the amplitude difference value also becomes.Generate parts 50 by with threshold value the amplitude difference value being carried out after binaryzation generates output image at characteristic quantity, the template of storing in the characteristic quantity of the object Ob that similitude calculating unit 52 will be extracted by Characteristic Extraction parts 50 and the template preparation/memory unit 56 compares, thereby calculates the similarity degree between them.When the similarity degree that provides from similitude calculating unit 52 was not less than threshold value, object identification parts 54 were identified as face corresponding to the people of face template with the object Ob that is detected.In object identification parts 54, can use traditional face detection.For example, can suitably adopt traditional converter technique such as affine transformation.

In the present embodiment,,, might reduce data volume (amount of information), and therefore carry out face recognition effectively and handle so compare with the situation of using magnitude image because used by the amplitude differential image is carried out the output image that binaryzation obtains.In addition, also have another accurately advantage of recognition object, this is because can easily identify the big zone apart from the change rate that has such as a step (astep) from the zone with small distance change rate.

With under illumination light, prepare the to have pixel value gray difference image of (its each be the difference intensity level of each pixel of gray level image), and use by the situation that the gray difference image of being prepared carries out the image that binaryzation obtains being compared with predetermined threshold, under identical illumination light, prepare the to have pixel value amplitude differential image of (its each be the difference intensity level of each pixel of magnitude image), and use by the amplitude differential image of being prepared being carried out under the situation of the image that binaryzation obtains with predetermined threshold, might minimize the influence of surround lighting simultaneously by the accurate reliably recognition object of object identification parts 54.

As first modification of present embodiment, can concern according to the position of characteristic quantity and carry out identification, replace carrying out identification to whole face.For example, from the output image shown in the magnitude image set-up dirgram 12B shown in Figure 12 A (that is image) by amplitude difference image binaryzation is obtained.Under situation to amplitude difference image binaryzation, corresponding to eye, nose and mouthful the zone with compare corresponding to the zone of cheek and forehead, have bigger difference intensity level.Therefore, shown in Figure 12 C, can with corresponding to eye, nose and mouthful the boundary point (end point) in zone be extracted as characteristic points of faces C.Because determine to have the candidate of the face template of higher coupling reliability according to the position of these characteristic points relation, then by using candidate preferentially to calculate similarity degree, so might save the required time of discerning.

As second modification of present embodiment, replace the amplitude differential image, can extract characteristic quantity from the amplitude gradient direction image shown in Figure 10 D.Promptly, Characteristic Extraction parts 50 comprise amplitude gradient direction image composer (not shown), are used for generating having the amplitude gradient direction image of amplitude gradient direction value as pixel value (it is the difference direction value from each definite pixel of the amplitude of magnitude image).Extract characteristic quantity from amplitude gradient direction image.In the case, when the reverberation from object Ob changes, the advantage of influence of the change of the object that reduces in the object space and the distance between the light source is arranged, thereby even in when, illumination taking place fluctuating when also recognition object accurately.

For generating amplitude gradient direction image, as described in top embodiment, to all pixels of magnitude image by using the Suo Beier filter (hx of each mask size that all has 3 * 3 pixels, hy), carry out local space differentiation, have the amplitude gradient direction image of amplitude gradient direction value as pixel value (it is the difference direction value from each definite pixel of the amplitude of magnitude image) so that generate.Express amplitude gradient direction value θ by the equation (4) of the above-mentioned equation of following use (2) and (3).

θ＝tan ^-1{dx/dy} (4)

The amplitude gradient direction image of Figure 10 D has the image of lumped values as pixel value (it is to change from amplitude gradient direction value θ by using yardstick), thereby when the amplitude gradient direction value θ that is determined by equation (4) increases in the scope of 0 to 359 degree, lumped values also progressively increases (when amplitude gradient direction value θ is 0 when spending, it is minimum that lumped values becomes, and when amplitude gradient direction value θ be 359 when spending, it is maximum that lumped values becomes).Under the situation of using amplitude gradient direction value, carry out identification by similitude calculating unit 52 and object identification parts 54, just as the situation of using the amplitude differential image to object.

For in Characteristic Extraction parts 50, extracting characteristic quantity, for example, four direction E1, E2, E3 and E4 are carried out four values of amplitude gradient direction value and handle, as shown in FIG. 13A, so that prepare amplitude gradient direction image (in the case, four value images), shown in Figure 11 B.Then, by all pixels being used the filter (h1) of mask size with 7 * 7 pixels shown in Figure 11 B, respectively to the four direction that comprises in the filter (h1) (E1, E2, E3, quantity E4) is counted.When each institute's count number of each the institute's count number of direction E1 and direction E3 is not less than threshold value (for example 14), this position (that is the center pixel of 7 * 7 pixels) regarded as valid pixel.When its during less than threshold value, regard this position as inactive pixels.As a result, obtain the mask image, wherein show valid pixel and show inactive pixels, shown in Figure 13 C by black region by white portion.Handle by using this mask image to carry out mask, can only extract for the special characteristic of single body such as eye, nose and mouth.Therefore, can only be that the part of being extracted is carried out face recognition by using.In the case, compare, might save the required time of identification with the situation of carrying out the identification processing by the image that uses whole face.

In the superincumbent Characteristic Extraction parts 50, be positioned at center pixel (p5) 8 pixels (p1) to (p4), the pixel value (a) to (d) of (p6) to (p9), (f) to (i) (seeing Figure 11 A) on every side by use and calculate dx (u, v) and dy (u, v), to determine amplitude gradient direction value θ.Replacedly, 4 pixels around it can be positioned at by use or the pixel value of 16 pixels calculate.In addition, use arctan (tan in the present embodiment ^-1) as from dx (u, v) and dy (u v) determines the function of amplitude gradient direction value θ.Other functions are also available.For example, wish to use a kind of function, by use this function can revise dx (u, v) and dy (u, non-linear ratio v) is to obtain linear amplitude gradient direction value θ.

＜the three embodiment 〉

The spatial information detecting apparatus of present embodiment is characterised in that to carry out and is used to prevent that the saturated adjusting that receives light output from handling, and stably generating important magnitude image, and except following assembly, itself and first embodiment are basic identical.Therefore, omission is to the repetition of explanation of same components.

In the spatial information detecting apparatus of present embodiment, as shown in figure 14, light source 1 provides the flash of light with flicker cycle 10 to 100kHz.Imageing sensor 2 has the ability of each pixel being adjusted optical receiver sensitivity.By receive the light from object space in each light period in flash of light source 1 and dark period, the electric charge that imageing sensor 1 output is assembled in the modulation signal circulation of predetermined quantity is exported as receiving light.For example, when the frequency of modulation signal was 30Hz, exportable 60 of imageing sensor 2 per 1 second (that is, 30 light periods and 30 dark periods) received light output.In the following description, the unit definition that the reception light that at every turn provides from imageing sensor 2 is exported is a frame.Therefore, when output in each light period and dark period receives light output, obtain the data of two frames.

By A/D converter 6 output of imageing sensor 2 is converted to digital signal, is stored in the frame memory 7 then temporarily.Frame memory 7 can have the capacity of at least 2 frames, thereby can be stored in the group of received light output that obtains in light period and the dark period.Suppose the not fluctuation of light quantity of the surround lighting (sunlight or illumination light) that imageing sensor 2 receives in a circulation of modulation signal, difference between the output of the reception light that then obtains in light period and dark period is relevant with the catoptrical amplitude of object (Ob) reflection from object space (from its radiation signal light of light source 1 forward).When from imageing sensor 2 output reception light outputs, it is stored in the frame memory 7, and when assembling electric charges by imageing sensor 2, reads the data of being stored from frame memory 7.

Evaluation unit 4 comprises difference arithmetic unit 44, be used for calculating and the corresponding reception light output of two frame data of frame memory 7 storages between difference.With this difference value input magnitude image maker 40, to obtain error image (that is, magnitude image), it has for the difference of each pixel between light period and dark period, with as pixel value.Each pixel value of this error image is ideally corresponding to the receiving light power degree of flashlight.Therefore, might obtain the relevant error image of amplitude with the flashlight that receives by imageing sensor 2.For example, when in light period, obtaining with the corresponding reception light output of the image P1 of Figure 15 A and in dark period, obtaining, obtain the magnitude image P3 shown in Figure 15 C with the corresponding reception light output of the image P2 of Figure 15 B.In this magnitude image, detect the background of the object (Ob) in object space, be different from reflected signal light, thereby have only object (Ob) to be present in the magnitude image (pixel value that is different from object (Ob) is 0, that is, black picture element).In addition, by only being delivered in the reception light output that obtains in light period and the dark period, and uncertain difference between them, might obtain to comprise the gray level image of background.

In Figure 14, saturated definite parts that Reference numeral 8 indications are equipped with between frame memory 7 and evaluation unit 4.In these saturated definite parts 8, reception light output that will obtain in light period and dark period and the saturation threshold that each pixel is scheduled to compare.For example, when the reception light of each pixel that obtains in light period is output as (Aa) and the reception light of each pixel of obtaining when being output as (Ab) in dark period, the overall relation (Aa＞Ab) that realizes in short cycle (change of surround lighting does not wherein take place basically).Therefore, when reception light output (Aa) is no more than predetermined saturation threshold (Th1), receives light output (Ab) and also be no more than saturation threshold (Th1).In saturated definite parts 8, the reception light output (Aa) and the saturation threshold (Th1) that will obtain in light period compare, shown in the step among Figure 16 (S32).Saturation threshold (Th1) is to be used for checking the saturated threshold value that receives light output " Aa ".When reception light output (Aa) exceeded saturation threshold (Th1), it is saturated that saturated definite parts 8 determine to receive light outputs (Aa).If necessary, also can carry out the comparison between saturation threshold (Th1) and the reception light output (Ab).

Saturated reception light output (Aa) does not provide spatial information.In other words, even determined difference DELTA A, can not generate magnitude image by evaluation unit 4.Therefore, when saturated definite parts 8 detect when saturated, send the order that output valve is changed to predetermined value to difference arithmetic unit 44.That is, when confirming when saturated, the output of difference arithmetic unit 44 is become predetermined value.Might use the maximum of allowed band of output valve of difference arithmetic unit 44 or median or, be used as predetermined value usually not from the particular value of difference arithmetic unit 44 outputs.For example, when by 8 promptly 255 grades represent to receive light output (Aa, in the time of Ab), can be with maximum 255 as predetermined value.When using maximum to determine the output valve of parts 44, might will provide the pixel of saturated reception light output (Aa) to differentiate from background as difference.On the other hand, when the median of the allowed band of the output valve of using difference arithmetic unit 44, might obtain not have the magnitude image of jamais vu, this is because the difference between saturated pixel value and the surrounding pixel value becomes less relatively.

In addition, when by the value that on another pixel, does not occur (for example, when in using 255 grades 254 grades were represented difference DELTA A, value 255 can be as difference DELTA A appearance) when predetermined value is provided, can regard the pixel of predetermined value as inactive pixels, it shows that reception light output (Aa) is saturated.Thus, because saturated pixel is differentiated from other pixels as inactive pixels, so can the value interpolation of surrounding pixel be advanced inactive pixels by magnitude image maker 40.When using the interpolation magnitude image, the pixel with undesired value shows hardly, therefore can obtain to have the image of the jamais vu of reduction.

By this mode, the only pseudo-value of predetermined value be used for avoiding jamais vu when showing magnitude image, and saturated reception light is exported not reflection space information.Therefore, need to carry out and regulate, make the reception light output of next time in frame memory 7, storing not exceed saturation threshold.In addition, as mentioned above, use the purpose of magnitude image to be that acquisition only is the image of object (Ob), and do not have background.Therefore, when exporting when unsaturated the magnitude image that might obtain to want with the reception light of the corresponding pixel of object (Ob).

When for the reception light outgoing inspection of imageing sensor 2 to the saturated appearance in the appointed area, and confirm when this appointed area reception light output is saturated, preferably, reduce reception light output by imageing sensor 2 by emissive porwer, transmissivity that reduces the light-receiving optical system that for example reduces light source 1 or the cycle that shortens the gathering electric charge.In the case, can prevent to receive the saturated of light output effectively.

In the present embodiment, because export the output of difference arithmetic unit 44 when saturated and be set to predetermined value when receiving light, thus might prevent the appearance of jamais vu in the magnitude image, and this predetermined value is used for regulating.Use corresponding to the zone of object (Ob) as the appointed area.Can determine this appointed area by the output of using difference arithmetic unit 44 by regional specified parts 60.That is, the zone that regional specified parts 60 definite difference DELTA A that export from difference arithmetic unit 44 therein exceed effective threshold value (Th2) is arrived shown in (S9) as step among Figure 16 (S7) as the appointed area.In the method, can determine the appointed area shown in Figure 15 D (Da) from the magnitude image (P3) shown in Figure 15 C.In addition, when using this equipment, pre-determine effective threshold value (Th2), therefore be extracted in the people's face that requires to exist in the distance range from imageing sensor 2 as object (Ob) as the camera of call gate system.

Yet, there is a kind of situation, even wherein set in advance in the scope that is used for object (Ob) above-mentioned, difference also is no more than effective threshold value (Th2), this is because of the reflectivity difference that is caused by irregular (replace receiving the saturated position of light output with predetermined value, thereby exceed effective threshold value (Th2)) or apart from difference.In the case, extension process is carried out in the zone that exceeds effective threshold value (Th2).By this processing, even when exceeding effective threshold value (Th2) on the regional integration but sub-fraction that should the zone when not exceeding effective threshold value (Th2), this fraction can be included in the appointed area.For example, when object (Ob) when being face, eyes, eyebrow and hair place reflectivity are low.In these zonules, the difference that exists in the acquisition of difference arithmetic unit 44 places becomes and wants little possibility than effective threshold value (Th2).Yet, as whole face, there is the pixel of significant quantity to surpass effective threshold value (Th2), might eyes, eyebrow and hair be thought of as the appointed area by extension process.

In addition, by brightness detection part 62 definite mean values of exporting by the reception light of pixel in the appointed area of regional specified parts 60 detections.By using this mean value, regulate the regulated quantity that parts 9 are identified for controlled plant by output.That is, in brightness detection part 62, set in advance with magnitude image in the corresponding reference value of enough brightness of object (Ob), and the mean value above determining and as the error between the reference value of desired value.In addition, output is regulated parts 9 according to the error that provides from brightness detection part 62, determines the regulated quantity at controlled plant.

Regulate in the parts 9 in output,, determine regulated quantity on the direction that receives light output reducing when the mean value of the output of the reception light in the appointed area during greater than reference value.On the contrary, when the mean value that receives light output during, determine regulated quantity on the direction that receives light output increasing less than reference value.Controlled plant has admissible adjustable range.Therefore, when regulated quantity reaches the upper limit or following the prescribing a time limit that can allow adjustable range, keep current state.

Regulating under the situation of a plurality of controlled plants, pre-determining the priority orders of controlled plant, and reaching the upper limit that can allow adjustable range or down in limited time, other controlled plants of scalable when one of them controlled plant.In addition, the amount of the responsible error of change rate of preferred regulated quantity changes.In addition, preferred regulated quantity once can change fixing change rate, and when error was big, several changed regulated quantity, with near reference value.In the case, might prevent to change the fluctuation of the magnitude image that causes owing to moment in response to light quantity.In addition, preferred reference value has a scope, and when mean value that the reception light in the appointed area is exported is in the scope of reference value, does not change regulated quantity.In the case, might prevent unwanted operation to controlled plant.Wish to use the upper limit that can allow adjustable range to be used as the initial value of controlled plant when beginning to operate.

As mentioned above, parts 9 are identified for controlled plant according to the result of difference arithmetic unit 44 regulated quantity is regulated in regional specified parts 60, brightness detection part 62 and output.In Figure 14, can realize parts except that light source 1 and imageing sensor 2 by in microcomputer, carrying out proper procedure.

Operation with reference to the flow chart brief description present embodiment of Figure 16.When the beginning operating equipment, at first obtain the image (S1) of object space by imageing sensor 2.As mentioned above, carry out imaging synchronously with the light period and the dark period of light source 1.When in light period, obtaining to receive light output (Aa) (S2), itself and saturation threshold (Th1) are compared (S3) by saturated definite parts 8.When reception light output (Aa) is not more than saturation threshold (Th1), mean to receive also unsaturation of light output.Therefore, determine to receive light output (Aa, Ab) the difference DELTA A between (S4) by difference arithmetic unit 44.On the other hand, when reception light output (Aa) exceeded saturation threshold (Th1), difference DELTA A was set to predetermined value (S5).Thus obtained difference DELTA A is exported as magnitude image (S6) by magnitude image maker 40.Difference arithmetic unit 44 also sends to regional specified parts 60 with the difference DELTA A that obtains.In regional specified parts 60, difference DELTA A is compared (S7) with effective threshold value (Th2).The pixel that exceeds effective threshold value (Th2) is stored as the pixel (S8) of appointed area.All pixels of magnitude image are carried out this processing (S9).When obtaining that wherein difference DELTA A exceeds the appointed area of effective threshold value (Th2), determine the mean value (S10) of difference DELTA A by the pixel in 62 pairs of appointed areas of brightness detection part.

Regulate in the parts 9 in output, mean value and the reference value of the difference DELTA A in the appointed area compared (S11).When mean value during greater than reference value, controlled plant carried out regulate, receive light output (S13) to reduce.On the contrary, when the mean value of difference DELTA A is not more than reference value, controlled plant is carried out adjusting, receive light output (S15) to increase.Because there is upper and lower limit in the effective regulated quantity of controlled plant, so on regulated quantity reduce to receive the direction of light output, becoming minimum (S12) or increasing when becoming maximum (S13) on the direction that receives light output, carry out next image-forming step (S1), and do not change regulated quantity.

In the superincumbent operation, output is regulated parts 9 and provide order to controlled plant when the difference that each reception light that obtains between light period and the dark period is exported.Replacedly, when the mean value of the reception light output of being determined by brightness detection part 62 has repeated pre-determined number continuously greater than the situation of reference value or the mean value of the reception light output determined by brightness detection part 62 when having repeated pre-determined number continuously less than the situation of reference value, provide order to controlled plant.

For preventing swing (hunting) phenomenon at the controlled plant place, parts 9 are regulated in output can have hysteresis characteristic.For example, thereby become greater than reference value when controlled plant provides the order that reduce to receive light output when mean value,, also do not provide to increase and accept next order that light is exported even when causing mean value to become less than reference value when error takes place.Only when the absolute value of error is not less than predetermined value, just provide next order.On the contrary, thereby become less than reference value when controlled plant provides the order that increase to receive light output when mean value,, also do not provide to reduce and receive next order that light is exported even when causing mean value to become greater than reference value when error takes place.Only when the absolute value of error is not less than predetermined value, just provide next order.Replacement determines to receive the mean values of light output by brightness detection part 62, also preferably counts receiving the quantity that light exceeds the pixel of reference value in the appointed area, and determines regulated quantity according to the pixel count of being counted.

In the superincumbent explanation, mean value and the single reference value that receives light output in the appointed area compared.Replacedly, a plurality of reference values can be set, make it possible to change the adjusting quantity that is used for controlled plant with step-wise manner.In addition, when in each of a plurality of grades, reference value being set,, might count, and definite regulated quantity, with level corresponding to the counting maximum of pixel to the quantity of pixel for every grade.

Extracting under the situation of appointed area by regional specified parts 60, also preferably will multiply by result that a coefficient obtains by the difference that is obtained by difference arithmetic unit 44 compares with effective threshold value (Th2), and this coefficient that is used to multiply each other changes with step-wise manner according to the mean value of being exported by brightness detection part 62 definite reception light.Coefficient is set, thereby when mean value was big more, this coefficient was more little.When the little mean value that receives light output in owing to the appointed area causes the image blackening,, increase the zone that is extracted as the appointed area by difference being multiply by a bigger coefficient.For determining this coefficient, use mean value in this embodiment as the typical value of the reception light output of determining by brightness detection part 62.Replacedly, might use other values, as maximum and pattern as typical value.

When the amplitude of flashlight is very little, for example object (Ob) and light source separate the reflection of a big distance or object (Ob) when very low, by increasing the appointed area, have higher possibility that this object (Ob) is included in the appointed area.By this operation, might basically the pixel value corresponding to object (Ob) be remained in the preset range, and have nothing to do, thereby have the magnitude image of the object (Ob) of better visuality from 40 acquisitions of magnitude image maker with the change of the distance that arrives object (Ob) or the reflection of object (Ob).

In the superincumbent structure, be used for being not limited to therefrom extract the zone of object (Ob) by the zone that brightness detection part 62 determines to receive the mean value quantity of the pixel of required condition (or satisfy) of light output.In regional specified parts 60, all pixels in can image are set to the appointed area.In addition, presumptive area that can partly define in picture centre or the desired region in the image are set to the appointed area.Thus, how the appointed area being set can arbitrarily determine according to purpose.Pre-determining under the situation of appointed area, it can be independent of the difference that is obtained by difference arithmetic unit 44 and be determined by regional specified parts 60.

In frame memory 7, the corresponding reception light output of a circulation of storage and light period and dark period, and, the difference between the corresponding reception light output of difference arithmetic unit 44 circulations definite and light period and dark period.Replacedly, corresponding to the reception light output of light period repeatedly with corresponding to the repeatedly reception light output of dark period, can be stored in the frame memory 7 with append mode.In the case, because use additive value by a plurality of circulations acquisitions of light period and dark period, to carry out the processing in difference arithmetic unit 44, saturated definite parts 8 and brightness detection part 62, so might suppress to receive the influence of the noise contribution that comprises in the light output.In addition, might use the mean value that obtains by the number of times that additive value is exported divided by output reception light.In addition, that can use the difference that obtains by difference arithmetic unit 44 and or the mean value of difference.

As mentioned above, when use receives the mean value of light output or the mean value of the difference that reception light is exported, and when changing one of the number of times of addition and divisor, expectation occurs and the situation confers similar advantages that is used for being determined by regional specified parts 60 coefficient of appointed areas in change.In addition, when use receives the mean value of the mean value of light output or the difference that reception light is exported, compare with the situation of the corresponding reception light output of a circulation of light period and dark period with use, frame rate reduces, but can reduce the amount of the shot noise that causes at imageing sensor 2 places.Thus, also preferably reduce the number of times of addition down, and on the other hand, under outdoor environment (wherein the amount of surround lighting is totally bigger), increase the number of times of addition, so that The noise minimizes in indoor environment (wherein the amount of surround lighting is totally less).

The modulation signal that is used to control flasher is not limited to square wave.For example, can use sine wave, triangular wave or sawtooth waveforms.In the case, because can receive light output by synchronously extracting, and determine difference between the reception light output of these two phase cycles from imageing sensor 2 with two of modulation signal different phase cycles, eliminate the surround lighting composition, so might obtain to be similar to the image of magnitude image.Can arbitrarily be identified for extracting the duration in the cycle that receives light output from imageing sensor 2.When the differing of these phase cycle phase mutual deviations 180 degree, it is substantially equal to determine the situation of the difference between light period and the dark period.Differing between the phase cycle can be different from 180 degree.

In addition, various compositions can be used for light-receiving optical system 5.Under the situation of using iris ring, preferably can adjust aperture diameter according to external signal.In addition, preferably except that iris ring and lens, also prepare a plurality of neutral-density filters, and can select to determine suitable filter according to external signal with different transmissivities.Replacedly, might use the neutral-density filter that utilizes liquid crystal, it has the ability that changes transmissivity according to external signal.

Industrial applicability

Such as what understand from top explanation, spatial information detecting apparatus of the present invention has for effectively Eliminate surround lighting impact the purification apparatus structure and be used for preventing from receiving the light output saturation by carrying out Output regulate to process and stably obtain the ability of important magnitude image. Therefore, wish that the present invention will Be used to wide, wish to detect reliably in the application of information of object space therein.

Claims (10)

1. spatial information detecting apparatus comprises:

At least two optical-electrical converters are configured to receive light from object space, and generate and receive the corresponding electricity output of light intensity, and the passage of scintillation light that wherein just will carry out intensity modulated with the modulation signal with preset frequency is shone into this object space;

A plurality of electrodes are formed on each described optical-electrical converter;

By the charge storage parts that potential energy well provides, it is formed in each described optical-electrical converter by apply control voltage at least one described electrode, is used for accumulating in the part of charge that described optical-electrical converter generates;

Controller, be configured to control described control voltage, thereby, in the light period of described passage of scintillation light, in of described optical-electrical converter, form and have large-area first potential energy well, and in another of described optical-electrical converter, form second potential energy well with small size, and in the dark period of described passage of scintillation light, form described first potential energy well in another the described of described optical-electrical converter, and in described of described optical-electrical converter, form described second potential energy well;

Electric charge is discharged parts, is configured to export the electric charge of assembling in described charge storage parts; And

Evaluation unit, be configured to electric charge that the electric charge assembled in described light period according to described first potential energy well and described second potential energy well assemble in described dark period and, the electric charge in described dark period, assembled with the electric charge in described light period, assembled of described second potential energy well and described first potential energy well and between difference, detect spatial information about object space

Wherein said evaluation unit comprises the magnitude image maker, is configured to generate the magnitude image with pixel value, and wherein each pixel value is provided by described difference, and

Described controller is controlled described control voltage, make described first potential energy well that in described light period, forms and described second potential energy well area and, equal described first potential energy well that in described dark period, forms and described second potential energy well area and.

2. spatial information detecting apparatus as claimed in claim 1, wherein, described controller control is applied in the quantity of the described electrode of described control voltage.

3. spatial information detecting apparatus as claimed in claim 1, also comprise: the gray level image maker, be configured to generate gray level image with pixel value, wherein each pixel value is provided by one of the following: the mean value of the quantity of electric charge that described charge storage parts are assembled in the described light period of described passage of scintillation light and described dark period one and the quantity of electric charge assembled in the two at described light period and described dark period by described charge storage parts.

4. spatial information detecting apparatus as claimed in claim 1 also comprises: the Characteristic Extraction parts are configured to according to the characteristic quantity that is extracted the object in the described object space by the described magnitude image of described magnitude image maker generation; The similitude calculating unit is configured to calculate similarity degree by described characteristic quantity is compared with the previous template of preparing; And the target identification parts, be configured to when similarity degree is not less than predetermined value, described object identification be the target object corresponding to described template.

5. spatial information detecting apparatus as claimed in claim 4, wherein, described object to be detected is a face, and this spatial information detecting apparatus also comprises the template stores parts, be configured to store according to the characteristic quantity of described face and pre-prepd face template, and when the similitude between the described characteristic quantity that is extracted by described Characteristic Extraction parts and the described face template of storing in described template stores parts was not less than predetermined value, described object identification parts were people corresponding to described face template with described face recognition.

6. spatial information detecting apparatus as claimed in claim 1, also comprise: saturated definite parts are configured to predetermined threshold is compared with the quantity of electric charge of being assembled at least one of described two different phase cycles of described modulation signal by described charge storage parts; And output regulates parts, is configured to regulate electricity output corresponding to the receiving light power degree according to comparative result.

7. spatial information detecting apparatus as claimed in claim 6, wherein, when the quantity of electric charge during greater than threshold value, the electricity output that parts reduce described optical-electrical converter is regulated in described output.

8. spatial information detecting apparatus as claimed in claim 1, also comprise saturated definite parts, be configured to predetermined threshold is compared with the quantity of electric charge of being assembled at least one of described two different phase cycles of described modulation signal by described charge storage parts, and, wherein when the quantity of electric charge during greater than threshold value, described evaluation unit replaces described difference to come the evaluation objective space by using preset difference value.

9. spatial information detecting apparatus as claimed in claim 1, also comprise: saturated definite parts, be configured to predetermined threshold with comparing corresponding to the quantity of electric charge of assembling in each of described two different phase cycles on cycle memory time of a plurality of circulations of described modulation signal, described modulation signal; And output adjusting parts, be configured to according to comparative result, regulate corresponding to the electricity of receiving light power degree by change cycle memory time and export.

10. spatial information detecting apparatus as claimed in claim 1, also comprise: saturated definite parts, the quantity of electric charge that is configured to assemble in each of described two different phase cycles in the circulation of predetermined threshold and described modulation signal, described modulation signal compares; And output regulates parts, is configured to according to comparative result, regulates electricity output corresponding to the receiving light power degree by at least one the duration that changes described two different phase cycles.

Images