JP2011038847A - Illumination information detector, illumination information signal converter and illumination information signal conversion method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a small-scale, precise illumination information detector having fewer components and provides a precise illumination spectral distribution even with a small number of bands. <P>SOLUTION: The illumination information detector includes: a k band detection part 2 for detecting illumination light and outputting a k band signal; and a signal conversion part 1 for converting the k band signal to an n band signal. The signal conversion part 1 has: an illumination characteristic data calculation part 3 for calculating illumination characteristic data by estimating the type of an illumination light source from the k band signal; and an illumination spectrum estimation part 4 for setting a conversion matrix for converting the k band signal to an n band signal according to the illumination characteristic data. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to an illumination information detection device, an illumination information signal conversion device, and an illumination information signal conversion method.

  With the recent development of digital cameras, there is an increasing demand for monitor display of captured images with good color reproducibility. For this purpose, the color reproducibility is enhanced by using the characteristics of the illumination light source that irradiates the subject at the time of photographing (see, for example, Patent Document 1), or the color characteristics of the monitor used for image display and the illumination characteristics around the monitor are taken into consideration. Various methods and apparatuses for color correction have been proposed.

  For example, an apparatus is known that detects illumination light, monitor color, and spectral distribution characteristics of a printer output pattern using a diffraction grating and a sensor of about 128 pixels. Since this apparatus uses a diffraction grating, the spectral distribution can be measured with high accuracy, and the spectral distribution characteristics of an illumination light source (such as a fluorescent lamp) having a bright line spectrum can also be detected. However, the diffraction grating itself has a problem that it is expensive and the assembly position accuracy of the optical system to be constructed is also necessary.

  On the other hand, a method for detecting a spectral distribution with a small number of bands using an optical filter has also been proposed (see, for example, Patent Document 2). This method using an optical filter is inexpensive and does not require much assembly accuracy. However, since this method uses an optical filter, the bandwidth is wide and the spectrum in a predetermined range is averaged. As a result, there is a drawback that the bright line spectrum such as a fluorescent lamp cannot be detected with high accuracy. .

Japanese Patent Laid-Open No. 11-089552 JP-T-2004-527729

  Therefore, an object of the present invention is to provide an apparatus that can acquire a spectral distribution with the same degree of accuracy as a multi-band sensor even when a sufficient number of bands cannot be secured as in the case of using an optical filter.

  The above-described problem of the present invention includes a k-band detection unit that detects illumination light and outputs a k-band signal, and a signal conversion unit that converts the k-band signal into an n-band signal satisfying k <n. An illumination characteristic data calculation unit that calculates the illumination characteristic data by estimating the type of illumination light source from the k band signal; and for converting the k band signal into the n band signal according to the illumination characteristic data. This is solved by an illumination information detection apparatus having an illumination spectrum estimation unit for setting a transformation matrix. That is, what kind of light source information (light source type, ratio, etc.) is discriminated from the k-band signal, and characteristic data used to estimate the lighting spell using this light source information is calculated / selected and used. At this time, k is considered to be about 4 to 16 bands, and n is considered to be about 64 to 256 bands.

  The k-band detector includes m optical long-pass filters having different cutoff wavelengths in steps, detects the illumination light through the m optical long-pass filters, and detects an m-channel signal. It is preferable to obtain a k-band signal with m> k based on the difference between channel signals. Furthermore, it is preferable that the m kinds of different optical long-pass filters are configured as mosaic filters, and m-channel detection is performed by a matrix type detection element close to the mosaic filter. By configuring in this way, the configuration of the detector is small and the number of parts can be reduced.

  The illumination characteristic data calculation unit estimates the type of the illumination light source using a ratio of an intensity of at least one band signal to an intensity of at least another band signal among the k band signals as a light source evaluation value. . Further, when the illumination characteristic data calculation unit acquires a plurality of the light source evaluation values estimated to be used, the illumination characteristic data calculation unit mixes the types of illumination light sources based on the light source evaluation values normalized by the plurality of light source evaluation values. Estimate the ratio.

  The detection band of the k-band signal is preferably configured to include at least one of 436 nm, 544 nm, and 612 nm. That is, it is preferable to include a spectral distribution corresponding to the emission line spectrum of the fluorescent lamp.

  It is conceivable to wirelessly transmit a k-band signal between the k-band detection unit and the signal conversion unit. At this time, application examples such as a configuration in which a k-band detection unit is provided in a remote control or a configuration in which a k-band detection unit is provided in a camera-equipped mobile phone are conceivable.

  The signal conversion unit is preferably provided in a digital camera and is also used as a video processing unit of the digital camera. In general, a digital camera includes a video processing unit / image processing unit such as a so-called image engine. Therefore, by using the signal conversion unit also as the video processing unit or built in the video processing unit, the number of parts can be reduced and the processing can be quickly used for processing such as auto white balance which is an application of the present invention.

  As another aspect of the present invention, an illumination characteristic data calculation unit that calculates the illumination characteristic data by estimating the type of the illumination light source from the k band signal having illumination light information, and the k band according to the illumination characteristic data There is a signal conversion device including an illumination spectrum estimation unit that converts a signal into an n-band signal with k <n.

  In addition, a first illumination characteristic data calculation unit that calculates the first illumination characteristic data by estimating the type of the illumination light source from the k band signal having illumination light information, and the k band signal according to the first illumination characteristic data. A first illumination spectrum estimation unit that converts k into a k ′ band signal with k <k ′, and a second illumination characteristic data calculation unit that calculates the second illumination characteristic data by estimating the type of the illumination light source from the k ′ band signal A second illumination spectrum estimation unit for converting the k ′ band signal into a k ″ band signal satisfying k ′ <k ″ in accordance with the second illumination characteristic data, and the type of the illumination light source from the k ″ band signal. A third illumination characteristic data calculation unit that calculates third illumination characteristic data and a third illumination that converts the k ″ band signal into an n band signal satisfying k ″ <n in accordance with the third illumination characteristic data. Spectral guess In particular, the first illumination characteristic data calculation unit estimates a large classification of the illumination light source, the second illumination characteristic data calculation unit estimates a medium classification of the illumination light source, and It is desirable that the three illumination characteristic data calculation unit estimates a small classification of the illumination light source.

Also, an input stage for inputting a _ki band signal having illumination light information, an illumination characteristic data calculation unit for calculating the illumination characteristic data by estimating the type of the illumination light source from the input _ki band signal, and the illumination depending on the characteristic data comprises said k _i band signal illumination spectrum estimation unit that converts the k _{i <k} i + ₁ becomes k _{i + 1} band signal, and a feedback circuit from the output stage of the illumination spectrum estimation unit to said input stage It is also possible to obtain an n-band signal by repeating the conversion from the _ki band signal to the ki _{+ 1} band signal a plurality of times.

  According to the present invention, since a highly accurate illumination spectral distribution can be obtained with a small number of bands, a small and highly accurate illumination information detection apparatus with a small number of components can be provided.

It is a block diagram of the illumination information detection apparatus by implementation of this invention. It is a block diagram of the illumination determination part by implementation of this invention. It is a block diagram of the data calculation part by implementation of this invention. It is a spectral distribution figure of the spectrum data of a fluorescent lamp. It is a spectral distribution figure of the spectral data of a halogen lamp. It is a spectral distribution figure of the spectrum data of LED illumination. It is a block diagram of the illumination spectrum estimation part by implementation of this invention. It is a figure which shows the spectrum data of a k band signal and an n band signal. It is a block diagram of the k band detection part by implementation of this invention. It is the schematic which shows the structural example of the k band detection part by implementation of this invention. It is a figure which shows the combination of the filter of the k band detection part by implementation of this invention. It is a figure which shows the spectral distribution characteristic of k band detection by implementation of this invention. It is a figure which shows the spectral distribution characteristic of the spectral distribution of a fluorescent lamp, and k band detection. It is a figure which shows the spectral distribution characteristic of the spectral distribution of a halogen lamp, and k band detection. It is a figure which shows the spectral distribution of LED illumination, and the spectral distribution characteristic of k band detection. It is a block diagram which shows 2nd Example of this invention. It is a figure which shows the structure of an illumination light source spectrum database. It is a block diagram which shows 3rd Example of this invention. It is the schematic which shows the Example which applied this invention to the digital camera. It is the schematic which shows the Example which applied this invention to the auto white balance. It is the schematic which shows the Example which applied this invention to the remote control of a camera. It is the schematic which shows the Example which applied this invention to the mobile phone with a camera.

  Hereinafter, the present invention will be described in detail with reference to the drawings.

  FIG. 1 is a block diagram of an illumination information detection apparatus according to an embodiment of the present invention. 1 includes a k-band detector 2 that detects illumination light and outputs a k-band signal, and a signal converter that outputs an n-band signal (k <n) from the input k-band signal. The signal conversion unit 1 further includes an illumination characteristic data calculation unit 3 that calculates illumination characteristic data and an illumination spectrum estimation unit 4 that calculates an n-band signal according to the illumination characteristic data. The illumination characteristic data calculation unit 3 calculates the illumination characteristic data using the k band signal acquired from the k band detection unit 2, and the illumination spectrum estimation unit 4 estimates the illumination spectrum (from the k band signal using the illumination characteristic data). conversion to an n-band signal).

  Furthermore, the illumination characteristic data calculation unit 3 includes an illumination determination unit 5 and a data calculation unit 6. The illumination determination unit 5 determines the type and ratio of the illumination light source from the k-band signal, and the data calculation unit 6 calculates illumination characteristic data from the determined illumination information data. Hereinafter, the illumination determination unit 5 and the data calculation unit 6 will be described in more detail with reference to FIGS. 2 and 3.

As shown in FIG. 2, the illumination determination unit 5 includes a specific light source component detection unit 7 and an illumination mixture ratio determination unit 8. The specific light source component detection unit 7 calculates a light source evaluation value using the k band signal acquired from the k band detection unit 2. Here, E ₁ , E ₂ , E _{3 and the} like described later can be used as the light source evaluation value, and the type of illumination light source can be specified by this light source evaluation value. The illumination mixture ratio determination unit 8 determines the mixture ratio of the specified illumination light source based on the light source evaluation value obtained by the specific light source component detection unit 7. Therefore, the illumination information data that is the output from the illumination determination unit 5 includes the type and ratio of the illumination light source.

  The data calculation unit 6 illustrated in FIG. 3 includes a target data reading unit 9, an illumination light source spectrum database 10, and a principal component analysis unit 11. The illumination light source spectrum database 10 stores spectrum data of various light sources as shown in FIGS. Further, the illumination light source spectrum database 10 can be provided with a natural light database 12 and an artificial light database 13 separately. The target data reading unit 9 reads the corresponding spectrum data from the illumination light source spectrum database 10 according to the illumination information data (illumination light source and its ratio), and sends these spectrum data to the principal component analysis unit 11. The principal component analysis unit 11 performs principal component analysis using all of the transmitted spectral data, and sends the result to the illumination spectrum estimation unit 4 as illumination characteristic data.

  Next, the illumination spectrum estimation unit 4 will be described in more detail using the block diagram of FIG.

  The illumination spectrum estimation unit 4 shown in FIG. 7 includes an estimation matrix calculation unit 14 and an estimation calculation unit 15. The estimation matrix calculation unit 14 calculates a spectrum estimation matrix T (n × k matrix) from the illumination characteristic data obtained from the data calculation unit 6, and the estimation calculation unit 15 uses the spectrum estimation matrix T to calculate k-band. Conversion from a signal to an n-band signal is performed.

  According to the above configuration, since the conversion matrix is set according to the type of illumination light source, it is possible to convert from a k-band signal to an n-band signal satisfying k <n with higher accuracy than in the past. For example, even if the spectrum includes bright lines as shown in FIG. 8, the spectrum can be estimated with high accuracy from a detection signal with a small number of bands. In particular, since the illumination characteristic data is generated according to the magnitude of the mixing ratio, estimation that reflects the magnitude of contribution is possible. In addition, since the information of the light source with little contribution is used, the characteristics of the light source with little contribution are reflected in the calculation without being ignored. Here, the data read from the spectrum database is changed according to the ratio of the mixing ratio. However, the illumination characteristic data is obtained by calculating in advance the main components of each light source and multiplying the weight according to the mixing ratio. Needless to say.

  Next, the configuration of the k-band signal acquisition part (k-band detection unit 2) and the acquisition method according to the embodiment of the present invention will be described.

  FIG. 9 is a block diagram of a k-band detection unit according to an embodiment of the present invention. As shown in the figure, the k-band detection unit 2 of this embodiment includes an optical system 16, a filter 17, a sensor unit 18, an A / D converter 19, a signal correction unit 20, a difference calculation unit 21, and a signal storage unit 22. And detecting illumination light and outputting a k-band signal.

  The k band detection unit 2 can be configured as shown in FIG. 10, for example. The configuration example shown in FIG. 10 includes an integrating sphere 23 as the optical system 16, a 16-color mosaic filter 24 as the filter 17, and a matrix detection element 25 as the sensor unit 18. Furthermore, the signal correction unit 20, the difference calculation unit 21, and the signal storage unit 22 can be provided in the substrate unit 26, or can be separately provided outside the substrate unit.

The k-band detection unit 2 performs m-division spectroscopy on the illumination light detected by the optical system 16 by the filter 17 (m = 16 in the example of FIG. 10), and detects m channels by the sensor unit 18. At this time, as shown in FIG. 11, the spectral characteristics of the filter are combined with optical long-pass filters having different cut-on wavelengths in stages. The m-channel signal detected by the sensor unit 18 is converted into a digital signal by the A / D converter 19, and dark (dark current) correction, linear correction, and variation among sensors are corrected by the signal correction unit 20. . Thereafter, the difference calculation unit 21 calculates the difference from the m-channel spectral characteristics f _{1 to} f _{16 as} follows.

g ₁ = f ₂ −f _1,
g ₂ = f ₃ −f _2,
・
・
g _m-1 = f _m -f _m-1.

As a result of the difference calculation in this way, the m−1 channel signal (V _i : i = 1,..., M−1) has spectral distribution characteristics g _{1 to} g ₁₅ similar to those of the bandpass filter set as shown in FIG. Signal. In the k band detection unit 2, the number m of sensors does not necessarily match the number k of signals to be output. In this configuration, k <m, and in particular, k = m−1.

  By configuring as described above, k-band detection can be performed using a long-pass filter that is easy to manufacture, such as a dye filter (also referred to as colored glass), without using a band-pass filter that is difficult to manufacture with high accuracy. Yes.

  Next, a method for determining the illumination light source and its mixture ratio from the k-band signal detected as described above will be described. That is, the details of the function of the illumination determination unit 5 described above will be described.

  The illumination determination unit 5 is for selecting spectrum data stored in the illumination spectrum database 10 at the subsequent stage. On the other hand, the spectrum data stored in the illumination spectrum database 10 has a characteristic spectrum for each type as shown in FIGS. The specific light source component detection unit 7 determines the illumination light source by paying attention to the spectrum that is a feature of the illumination light source. Here, a fluorescent lamp, a halogen lamp, and LED illumination will be described as typical illumination light sources.

FIG. 4 shows the spectral distribution of a typical fluorescent lamp. As can be seen from the figure, the fluorescent lamp has a bright line spectrum and is characterized by several intense wavelengths. In both of the three-wavelength type and the high color rendering type, a bright line spectrum exists in the vicinity of 440 nm, 544 nm, and 620 nm. Therefore, in the detection signal from the k-band detection unit 2, it is possible to determine whether or not the three spectrums are fluorescent lamps by determining the ratio of these three spectra to the other spectra. For example, as shown in FIG. 13, the center wavelengths of the bandpass obtained from the k-band detector 2 are g ₂ : 436 nm,
g ₁₀ : 544 nm,
g ₁₁ : 612 nm
If so, in a fluorescent lamp, these three signal values are extremely larger than the other values. Therefore, when a fluorescent lamp is used as the illumination light source, E ₁ = (V ₂ + V ₁₀ + V ₁₁ ) / (V ₁ + V ₃ + V ₄ + V ₅ + V ₆ + V ₇ + V ₈ + V ₉ + V ₁₂ + V ₁₃ + V ₁₄ + V ₁₅ )
The value of increases. Thus, by using the E _1, illumination source can determine whether the fluorescent lamp.

Further, as shown in FIG. 5, the halogen lamp has a feature that a long wavelength component is larger than a low wavelength. For this reason, when a halogen lamp is used as the illumination light source, as shown in FIG. 14, the detection signals V ₁₅ , V ₁₄ , V ₁₃ from longer wavelengths are relatively high. Thus, for example, E ₂ = (V ₁₃ + V ₁₄ + V ₁₅ ) / (V ₁ + V ₂ + V ₃ )
The value of becomes larger, and this value can be used to determine whether or not the illumination light source is a halogen lamp.

Furthermore, in the case of LED illumination, as shown in FIG. 6, only a specific wavelength component is larger than the others. For this reason, when LED illumination is used as the illumination light source, as shown in FIG. 15, a specific detection signal is relatively high. So, for example,
E ₃ = max {V _i } / max {V _j }
The value of becomes larger, and this value can be used to determine whether or not the illumination light source is LED illumination.

The illumination mixture ratio determination unit 8 determines the mixture ratio of each light source based on these E ₁ , E ₂ , and E ₃ . The sizes of E ₁ , E ₂ , and E ₃ represent the intensity of the fluorescent lamp, halogen, and LED,
α ₁ = c ₁ E ₁ / (c ₁ E ₁ + c ₂ E ₂ + c ₃ E ₃ ),
α ₂ = c ₂ E ₂ / (c ₁ E ₁ + c ₂ E ₂ + c ₃ E ₃ ),
α ₃ = c ₃ E ₃ / (c ₁ E ₁ + c ₂ E ₂ + c ₃ E ₃ )
Is calculated as α ₁ + α ₂ + α ₃ = 1. Therefore, α ₁ , α ₂ , and α ₃ are set as a mixing ratio of each light source. Note that c ₁ c ₂ c ₃ is a coefficient for adjusting the scales of E ₁ , E ₂ , and E ₃ .

  According to the above configuration, an illumination light detection device that is small and has a small number of components can be configured. In addition, since interpolation is performed according to the type of illumination light source and its ratio, it is possible to obtain a highly accurate spectral distribution comparable to multi-band detection even with detection with a small number of bands.

  FIG. 16 is a block diagram showing a second embodiment of the present invention. As shown in the figure, the signal conversion unit 1 of this embodiment includes a first signal conversion unit 27, a second signal conversion unit 28, and a third signal conversion unit 29, and these first signal conversion units 27. The second signal conversion unit 28 and the third signal conversion unit 29 have substantially the same configuration as the signal conversion unit 1 in the first embodiment. The first signal converter 27 receives the k-band signal from the k-band detector and converts it to the k′-band signal, and the second signal converter 28 receives the k′-band signal from the first signal converter. The third signal conversion unit 29 receives the k ″ band signal from the second signal conversion unit and converts it into an n band signal. Here, k <k ′ <k ″ <n, and the information is sequentially converted into more accurate information.

  The first signal conversion unit 27, the second signal conversion unit 28, and the third signal conversion unit 29 have substantially the same configuration as the signal conversion unit 1 in the first embodiment, but the internal illumination light source spectrum database is different. In the present embodiment, the illumination light source spectrum database 10 is structured into a large classification, a medium classification, and a small classification as shown in FIG. 17, and the first signal conversion section 27, the second signal conversion section 28, and the third classification are respectively arranged. The signal conversion unit 29 is provided. Then, the first signal conversion unit 27 determines large classifications such as natural light and artificial lighting, and the second signal conversion unit 28 uses sunlight, clear sky light, cloudy light, or the like, or a fluorescent lamp, a halogen lamp, LED lighting, or the like. The third signal conversion unit 29 determines a small classification such as the type of fluorescent lamp.

  If the determination of the illumination light source is made only at the large classification level (natural light or artificial illumination), the effect of improving accuracy cannot be greatly expected. On the other hand, there is a risk of making a large misjudgment in the determination of the small classification level (light source fine classification). In this embodiment, the illumination spectrum is estimated in multiple stages, so that the determination can be performed with high accuracy while avoiding the risk of erroneous determination.

  FIG. 18 is a block diagram showing a third embodiment of the present invention. The present embodiment can be considered as a modification of the second embodiment. As described above, the first signal converter 27, the second signal converter 28, and the third signal converter 29 of the second embodiment have substantially the same configuration as the signal converter 1 of the first embodiment. Therefore, in this embodiment, by providing a feedback circuit from the output stage to the input stage of the signal converter 1, the first signal converter 27, the second signal converter 28, and the third signal converter in the second embodiment. 29 is shared.

The signal converter 1 shown in FIG. 18 includes a counter (not shown) inside, and repeats signal conversion while incrementing a k-band signal to which a counter bit is added (this is expressed as a _ki band signal). In this case, estimation of the illumination light source in the interior of the signal converter 1, the major classification according to the counter bits k _i band signal, middle classification, it is possible to change the classification accuracy, such as small classification. According to the present embodiment, it is possible to perform highly accurate determination while avoiding the risk of erroneous determination without increasing the number of parts.

  FIG. 19 is a schematic diagram showing an embodiment in which the present invention is applied to a digital camera, in which a k-band detector 2 is added to the configuration of a typical digital single-lens reflex camera. Here, as the k-band detection unit 2, the k-band detection unit 2 described in the first embodiment can be used, and the attachment position can be attached to an appropriate position for detecting illumination light. .

  In addition, as a configuration of a typical digital single-lens reflex camera, a camera lens 30, a quick return mirror 31, an image sensor 32, a finder unit 33, a video processing unit 34, an image storage unit 35, an operation setting button 36, and a liquid crystal display 37 are provided. It is shown in the figure. The light beam incident from the camera lens 30 is reflected by the quick return mirror 31 to the viewfinder 33, or forms an image on the image sensor 32 by removing the quick return mirror 31 from the optical path during imaging. The image signal captured by the image sensor 32 is processed by the video processing unit 34, stored in the image storage unit 35, and displayed on the liquid crystal display 36. Further, the processing by the video processing unit 34 can be set by the operation setting button 37.

  In Embodiment 4 of the present invention, the k-band signal from the k-band detection unit 2 attached as described above is processed by the video processing unit 34 and converted into an n-band signal. In other words, the signal conversion unit 1 and the video processing unit 35 in the first embodiment (or the second and third embodiments) are used as a configuration. With this configuration, an increase in the number of parts can be minimized. Further, since the acquired n-band signal is obtained by the video processing unit 35, it can be used immediately for processing such as white balance.

  Furthermore, by storing the calculated n-band signal in association with the captured image data in the image storage unit 35, the n-band signal can be used for image processing in an external processing device such as a personal computer.

  FIG. 20 is a block diagram showing an embodiment in which the present invention is applied to auto white balance. The fifth embodiment shown in the figure includes a signal conversion unit 1, a k-band detection unit 2, and an auto white balance unit 38. The auto white balance unit 38 further includes a gain setting unit 39, a matrix selection unit 40, and a gain correction unit. 41, a matrix storage unit 42, a matrix correction unit 43, and a color conversion unit 44.

  In this embodiment, the illumination information data (light source type / ratio) obtained by the signal conversion unit 1 is transmitted to the auto white balance unit and used for gain setting and matrix selection. The gain setting unit 39 sets RGB gain according to the illumination information data, and the gain correction unit 41 corrects the gain of the RGB image data. The matrix selection unit 40 selects a color conversion matrix from among a plurality of color conversion matrices stored in the matrix storage unit 42 according to the type of light source included in the illumination information data, and sends these to the matrix correction unit. The matrix correction unit recalculates the selected color conversion matrix in accordance with the ratio of the light source types to obtain a correction matrix. The color converter 44 performs color conversion of the RGB image data using this correction matrix.

  According to the fifth embodiment of the present invention, by setting the auto white balance coefficient according to the illumination information data, it is possible to reproduce with a preferable color even when the subject is irradiated with a plurality of light sources.

  In Embodiment 6 of the present invention, the illumination detection sensor attached to the remote controller 45 of the camera is the k-band detection unit 2. When the camera is operated by a remote controller, a configuration in which an illumination detection sensor is attached to the remote controller is preferable because the illumination can be detected at a position close to the subject. As shown in FIG. 21, in this embodiment, the k-band signal is transmitted to the remote control receiver 46 provided in the camera body (or PC) using the illumination detection sensor at this time as the k-band detection unit 2. As a result, the amount of data can be reduced (that is, at high speed) and the power consumption can be reduced. Further, since a signal processing unit is not required in the remote controller, costs can be reduced.

  The seventh embodiment shown in FIG. 22 is an application of the present invention to a mobile phone 47 with a camera. Mobile devices such as mobile phones are limited in that they must be small and consume less power. Therefore, a configuration has been devised in which processing such as auto white balance is executed by the server 48 on the network. At this time, the camera-equipped mobile phone 47 may include only a k-band detector, transmit a k-band signal together with an image (or video), and convert it to an n-band signal by a signal conversion unit on the server 48. As a result, the amount of transmission data can be reduced, that is, power consumption can be reduced, which is suitable for use in mobile devices such as mobile phones.

DESCRIPTION OF SYMBOLS 1 Signal conversion part 2 k band detection part 3 Illumination characteristic data calculation part 4 Illumination spectrum estimation part 5 Illumination determination part 6 Data calculation part 7 Specific light source component detection part 8 Illumination mixture ratio determination part 9 Target data reading part 10 Illumination light source spectrum database DESCRIPTION OF SYMBOLS 11 Principal component analysis part 12 Natural light database 13 Artificial light database 14 Estimation matrix calculating part 15 Estimation calculating part 16 Optical system 17 Filter 18 Sensor part 19 A / D converter 20 Signal correction part 21 Difference calculating part 22 Signal storage part 23 Integrating sphere 24 16-color mosaic filter 25 Matrix type detection element 26 Substrate part 27 First signal conversion part 28 Second signal conversion part 29 Third signal conversion part 30 Camera lens 31 Quick return mirror 32 Image sensor 33 Finder part 34 Video processing part 35 Image storage Part 36 Setting button 37 LCD 38 auto white balance unit 39 gain setting unit 40 matrix selector 41 gain correction unit 42 matrix memory 43 matrix correction unit 44 color conversion section 45 remote control 46 Remote control receiver 47 camera phone 48 server

Claims (13)

A k-band detector that detects illumination light and outputs a k-band signal;
A signal converter that converts the k-band signal into an n-band signal with k <n,
The signal converter is
An illumination characteristic data calculating unit that calculates the illumination characteristic data by estimating the type of the illumination light source from the k-band signal;
An illumination spectrum estimation unit that converts the k-band signal into the n-band signal according to the illumination characteristic data;
An illumination information detection apparatus characterized by the above. The k-band detector has m optical long-pass filters having different cutoff wavelengths in stages,
An m-channel signal is detected by passing the illumination light through the m optical long-pass filters;
The illumination information detection apparatus according to claim 1, wherein a k-band signal satisfying m> k is obtained based on a difference between the m channel signals. The m different optical long pass filters are configured as mosaic filters,
The illumination information detection apparatus according to claim 2, wherein m-channel detection is performed by a matrix type detection element close to the mosaic filter.   The illumination characteristic data calculation unit estimates a type of the illumination light source using a ratio of an intensity of at least one band signal of the k band signals and an intensity of at least another band signal as a light source evaluation value. The illumination information detection device according to any one of claims 1 to 3, wherein the illumination information detection device is characterized.   When the illumination characteristic data calculation unit obtains a plurality of the light source evaluation values estimated to be used, the light source evaluation value normalized by the plurality of light source evaluation values is used to determine a mixing ratio of the types of illumination light sources. The illumination information detection device according to claim 4, wherein the illumination information detection device is estimated.   4. The illumination information detection apparatus according to claim 2, wherein the detection band of the k-band signal includes at least one of 436 nm, 544 nm, and 612 nm.   The illumination information detection apparatus according to claim 1, wherein a k-band signal is wirelessly transmitted between the k-band detection unit and the signal conversion unit.   The illumination information detection apparatus according to claim 1, wherein the signal conversion unit is provided in a digital camera and is configured by a video processing unit of the digital camera. An illumination characteristic data calculation unit that estimates the type of illumination light source from a k-band signal having illumination light information and calculates illumination characteristic data;
A signal conversion apparatus comprising: an illumination spectrum estimation unit that converts the k-band signal into an n-band signal with k <n according to the illumination characteristic data. A first illumination characteristic data calculating unit that calculates the first illumination characteristic data by estimating the type of the illumination light source from the k-band signal having the illumination light information;
A first illumination spectrum estimation unit that converts the k-band signal into a k′-band signal with k <k ′ in accordance with the first illumination characteristic data;
A second illumination characteristic data calculation unit that estimates the type of the illumination light source from the k ′ band signal and calculates second illumination characteristic data;
A second illumination spectrum estimation unit that converts the k ′ band signal into a k ″ band signal that satisfies k ′ <k ″ in accordance with the second illumination characteristic data;
A third illumination characteristic data calculating unit that estimates the type of the illumination light source from the k ″ band signal and calculates third illumination characteristic data;
A signal conversion apparatus comprising: a third illumination spectrum estimation unit that converts the k ″ band signal into an n band signal satisfying k ″ <n according to the third illumination characteristic data. The first illumination characteristic data calculation unit estimates a large classification of the illumination light source,
The second illumination characteristic data calculation unit estimates a middle classification of the illumination light source,
The signal conversion apparatus according to claim 10, wherein the third illumination characteristic data calculation unit estimates a small classification of the illumination light source. An input stage for inputting a _ki band signal having illumination light information;
An illumination characteristic data calculation unit that estimates the type of illumination light source from the input _ki band signal and calculates illumination characteristic data;
And in response to said illumination characteristic data, the _{k i} band signal _k i _{<illumination} spectrum estimating unit for converting the _{k i + 1} becomes _{k i + 1} band signal,
A feedback circuit from the output stage of the illumination spectrum estimation unit to the input stage;
A signal conversion apparatus characterized in that the conversion from the k _i band signal to the k _{i + 1} band signal is repeated a plurality of times to obtain an n band signal. An illumination characteristic data calculation step for estimating the type of illumination light source from the k-band signal having illumination light information and calculating illumination characteristic data;
An illumination information signal conversion method comprising: an illumination spectrum estimation step of converting the k-band signal into an n-band signal satisfying k <n according to the illumination characteristic data.

Images