WO2013099644A1 - 分光画像処理方法、分光画像処理装置およびプログラム - Google Patents
分光画像処理方法、分光画像処理装置およびプログラム Download PDFInfo
- Publication number
- WO2013099644A1 WO2013099644A1 PCT/JP2012/082417 JP2012082417W WO2013099644A1 WO 2013099644 A1 WO2013099644 A1 WO 2013099644A1 JP 2012082417 W JP2012082417 W JP 2012082417W WO 2013099644 A1 WO2013099644 A1 WO 2013099644A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- illumination
- spectral
- surface reflectance
- spectral distribution
- spectrum
- Prior art date
Links
- 230000003595 spectral effect Effects 0.000 title claims abstract description 209
- 238000003672 processing method Methods 0.000 title claims abstract description 22
- 238000005286 illumination Methods 0.000 claims abstract description 178
- 239000013598 vector Substances 0.000 claims abstract description 100
- 238000001228 spectrum Methods 0.000 claims description 75
- 238000000034 method Methods 0.000 claims description 25
- 238000004364 calculation method Methods 0.000 claims description 24
- 230000005855 radiation Effects 0.000 claims description 15
- 238000005259 measurement Methods 0.000 claims description 8
- 238000010586 diagram Methods 0.000 description 14
- 238000005457 optimization Methods 0.000 description 9
- 238000005516 engineering process Methods 0.000 description 3
- 239000003086 colorant Substances 0.000 description 2
- 238000004590 computer program Methods 0.000 description 2
- 239000000470 constituent Substances 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000000701 chemical imaging Methods 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N1/00—Scanning, transmission or reproduction of documents or the like, e.g. facsimile transmission; Details thereof
- H04N1/46—Colour picture communication systems
- H04N1/56—Processing of colour picture signals
- H04N1/60—Colour correction or control
- H04N1/6083—Colour correction or control controlled by factors external to the apparatus
- H04N1/6086—Colour correction or control controlled by factors external to the apparatus by scene illuminant, i.e. conditions at the time of picture capture, e.g. flash, optical filter used, evening, cloud, daylight, artificial lighting, white point measurement, colour temperature
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N1/00—Scanning, transmission or reproduction of documents or the like, e.g. facsimile transmission; Details thereof
- H04N1/46—Colour picture communication systems
- H04N1/56—Processing of colour picture signals
- H04N1/60—Colour correction or control
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N1/00—Scanning, transmission or reproduction of documents or the like, e.g. facsimile transmission; Details thereof
- H04N1/46—Colour picture communication systems
- H04N1/56—Processing of colour picture signals
- H04N1/60—Colour correction or control
- H04N1/603—Colour correction or control controlled by characteristics of the picture signal generator or the picture reproducer
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N1/00—Scanning, transmission or reproduction of documents or the like, e.g. facsimile transmission; Details thereof
- H04N1/46—Colour picture communication systems
- H04N1/56—Processing of colour picture signals
- H04N1/60—Colour correction or control
- H04N1/6083—Colour correction or control controlled by factors external to the apparatus
- H04N1/6088—Colour correction or control controlled by factors external to the apparatus by viewing conditions, i.e. conditions at picture output
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N17/00—Diagnosis, testing or measuring for television systems or their details
- H04N17/002—Diagnosis, testing or measuring for television systems or their details for television cameras
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N17/00—Diagnosis, testing or measuring for television systems or their details
- H04N17/02—Diagnosis, testing or measuring for television systems or their details for colour television signals
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N9/00—Details of colour television systems
- H04N9/64—Circuits for processing colour signals
- H04N9/67—Circuits for processing colour signals for matrixing
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N9/00—Details of colour television systems
- H04N9/64—Circuits for processing colour signals
- H04N9/73—Colour balance circuits, e.g. white balance circuits or colour temperature control
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/10—Cameras or camera modules comprising electronic image sensors; Control thereof for generating image signals from different wavelengths
Definitions
- the present invention relates to a spectral image processing method, a spectral image processing apparatus, and a program.
- the color information of the image photographed using the photographing apparatus is generated based on the light incident on the photographing apparatus after the ambient illumination light at the time of photographing is reflected by the surface of the object. For this reason, even if an object having the same surface reflectance is photographed, the color information of the photographed image also changes if the ambient illumination light is different.
- the environmental illumination light is white light
- the average value of the reflectance is gray
- the spectral distribution of illumination and the surface reflectance of the object are linear sums of the principal component vector and the average vector that are held in advance, and the energy required to satisfy the above assumption is defined, and this energy is optimized.
- the spectral distribution of illumination and the surface reflectance of the object are estimated.
- FIG. 11 shows a configuration diagram of the illumination / reflectance estimation method according to the related technology.
- the configuration diagram of FIG. 11 is a block diagram generated based on Patent Document 1 and Patent Document 2.
- the related technology includes a color information acquisition unit 1, an illumination / reflectance principal component vector storage memory 2, and a spectral estimation unit 3.
- the color information acquisition unit 1 acquires color information from the scene and outputs it.
- the illumination / reflectance principal component vector storage memory 2 holds the principal component vector and the average vector in order to indicate the restriction range of the spectral distribution of illumination and the surface reflectance of the object.
- Spectral estimation means 3 calculates the energy necessary to satisfy a preset object color or illumination color assumption while making the spectral distribution of illumination and the surface reflectance of the object a linear sum of the principal component vector and the average vector.
- the color information is used for calculation, and the calculated energy is optimized. Furthermore, the spectral distribution of illumination and the surface reflectance of the object when the calculated energy is optimum are output as estimated values.
- the related art restores the spectral distribution of the illumination and the surface reflectance of the object by setting assumptions regarding the object color or illumination color in the scene. Therefore, it is possible to accurately estimate the spectral distribution of illumination and the surface reflectance of objects for a limited number of scenes that satisfy the set assumptions, but for scenes that do not match the assumptions, spectral characteristics of illumination and objects There was a problem that the estimation accuracy of was deteriorated.
- the object of the present invention was invented in view of the above problems, and the object of the present invention is to accurately restore the spectral characteristics of an object and illumination from an observation spectrum without using assumptions regarding the object color and illumination color in the scene.
- the present invention provides a spectral image processing method, a spectral image processing apparatus, and a program.
- the present invention selects observation spectrum information from a multispectral image of a scene, and generates a constrained range of illumination spectral distribution and a constrained range of surface reflectance of an object using the basis vector of illumination and the basis vector of surface reflectance. And using the observation spectrum information, the constraint range of the spectral distribution of the illumination, and the constraint range of the surface reflectance of the object, the weighting coefficient of the basis vector of the illumination and the weight of the basis vector of the surface reflectance.
- this spectral image processing method an observation equation having a coefficient as a parameter is generated, and a spectral distribution of illumination of the scene and a surface reflectance of the object are calculated from the observation equation.
- the present invention uses a spectral information selection means for selecting observation spectral information from a multispectral image of a scene, a basis vector of illumination, and a basis vector of surface reflectance.
- a weighting coefficient of the basis vector of the illumination by using the constraint range generating means for generating the constraint range of the illumination spectrum, the constraint range of the spectral distribution of the illumination, and the constraint range of the surface reflectance of the object
- an observation equation generating means for generating an observation equation using the weight coefficient of the basis vector of the surface reflectance as a parameter
- a calculating means for calculating the spectral distribution of the illumination of the scene and the surface reflectance of the object from the observation equation
- a spectral image processing apparatus A spectral image processing apparatus.
- the present invention allows a computer to select observation spectrum information from a multispectral image of a scene, and to use a basis vector of illumination and a basis vector of surface reflectance to limit the range of illumination spectral distribution and the surface reflectance of an object.
- a weighting coefficient of the basis vector of the illumination and the surface using the processing for generating the constraint range of the observation spectrum information, the constraint range of the spectral distribution of the illumination, and the constraint range of the surface reflectance of the object A program that executes processing for generating an observation equation using a weighting coefficient of a basis vector of reflectance as a parameter, and processing for calculating a spectral distribution of illumination of the scene and a surface reflectance of an object from the observation equation .
- the present invention it is possible to stably and accurately estimate the spectral distribution of illumination and the surface reflectance of an object, using the obtained spectral spectrum information as a clue, regardless of the observed scene.
- FIG. 1 is a schematic configuration diagram of a first embodiment according to the present invention.
- FIG. 2 is a configuration diagram of the first embodiment according to the present invention.
- FIG. 3 is a flowchart of the first embodiment according to the present invention.
- FIG. 4 is a block diagram of a second embodiment according to the present invention.
- FIG. 5 is a flowchart of the second embodiment according to the present invention.
- FIG. 6 is a block diagram of a third embodiment according to the present invention.
- FIG. 7 is a flowchart of the third embodiment according to the present invention.
- FIG. 8 is an example of the basis vector of the illumination spectral distribution.
- FIG. 9 is an example of the basis vector of the surface reflectance created from the DC components of the 24 colors of the Macbeth color chart and the principal component vectors.
- FIG. 10 is a diagram showing an image of the restriction range of the illumination spectral distribution.
- FIG. 11 is a configuration diagram of an illumination / reflectance estimation method according to a related technique
- the present invention relates to a spectral image processing technique for estimating spectral characteristics of illumination and objects in a scene using color information observed from a scene, and in particular, using a spectral spectrum observed with a camera having a high wavelength resolution.
- the present invention relates to a spectral image processing technique for estimating the spectral distribution of illumination and the surface reflectance of an object without making assumptions about the object color or illumination color in the scene.
- FIG. 1 is a block diagram showing an outline of a first embodiment according to the present invention.
- the first embodiment shown in FIG. 1 includes a multispectral information acquisition unit 11, an illumination / reflectance constraint range holding memory 12, and a spectral assumption unit 13 without color assumption.
- the multispectral information acquisition means 11 extracts and outputs a sufficient number of spectrum information for optimization based on the multispectral information acquired from the scene.
- the illumination / reflectance constraint range holding memory 12 holds the constraint range of the spectral distribution of illumination and the surface reflectance of the object.
- the spectral estimation means 13 without color assumption the spectral distribution of illumination held in the illumination / reflectance constraint range holding memory 12 based on a sufficient number of spectrum information calculated by the multispectral information acquisition means 11.
- the observation equation is generated so as to satisfy the constraint range of the surface reflectance of the object, and the spectral distribution of the illumination and the surface reflectance of the object are calculated by solving the observation equation.
- the spectral estimation means 13 without color assumption outputs the calculated spectral distribution of illumination and the surface reflectance of the object.
- the first embodiment according to the present invention is different in the multispectral information acquisition means 11, the illumination / reflectance constraint range holding memory 12, and the spectral estimation means 13 without color assumption, as compared with the configuration of the related art described above. . Details of the configurations of the multispectral information acquisition means 11, the illumination / reflectance constraint range holding memory 12, and the spectral estimation means 13 without color assumption will be described below.
- the multispectral information acquisition unit 11 acquires the multispectral information from the scene in the same manner as the color information acquisition unit 1 in FIG. However, the color information acquisition unit 1 does not consider the number of color information to be output, whereas the multispectral information acquisition unit 11 considers the processing in the spectral estimation unit 13 without color assumption in the subsequent stage and is necessary for optimization. The difference is that a large number of spectrum information is output.
- N is the number of bands of the spectral spectrum observed, M number of spectrum is to be observed, P all the total number of parameters to estimate, the number of parameters needed to recreate the spectral distribution of P I is one lighting , P R is When indicating the number of parameters needed to reproduce a surface reflectance of one object, for example, outputs a sufficient number linearly independent spectral information satisfying equation (1).
- FIG. 2 is a block diagram showing a first embodiment according to the present invention.
- the internal configuration of the multispectral information acquisition unit 11 will be described.
- the multispectral information acquisition unit 11 illustrated in FIG. 2 includes a multispectral image capturing unit 111 and a spectrum selection unit 112.
- the multispectral image capturing means 111 outputs a multispectral image in which multispectral information of the scene is recorded.
- the spectrum selection unit 112 selects and outputs a plurality of pieces of spectral information having a linearly independent relationship that satisfies the condition of the expression (1), for example, from the data of the multispectral image obtained by the multispectral image capturing unit 111. .
- the spectral estimation means 3 in FIG. 11 estimates the spectral distribution of illumination in the scene and the surface reflectance of the object based on a preset assumption regarding the object color or illumination color. Specifically, the energy necessary to satisfy the assumptions regarding the set object color or illumination color is defined and optimized to estimate the spectral distribution of illumination in the scene and the surface reflectance of the object. For example, “average object color is gray” (gray world hypothesis), “the brightest object in the image is white”, “the skin color is included in the image”, “lighting is white” The energy required to satisfy such assumptions is calculated, and the spectral distribution of illumination and the surface reflectance of the object when the calculated energy is optimal are output as estimated values.
- the spectral estimation means 13 without color assumption in the first embodiment calculates the spectral distribution of illumination and the surface reflectance of the object without using assumptions regarding the object color or illumination color.
- the spectral distribution of illumination and the surface reflectance of the object are estimated with high accuracy using the restriction range of the spectral distribution of illumination and the surface reflectance of the object.
- an observation equation is generated from a sufficient number of linearly independent spectral information satisfying Equation (1), a spectral distribution of illumination, and a model formula of the surface reflectance of the object, and this observation equation is solved.
- the spectral distribution of illumination and the surface reflectance of the object are output as estimated values.
- the internal structure of the spectral estimation means 13 without color assumption will be described.
- the spectral estimation unit 13 without color assumption shown in FIG. 2 includes an observation equation generation unit 131 and an equation calculation unit 132.
- a model of the illumination spectral distribution I and the surface reflectance R of the illumination which is an N-dimensional vector, is obtained from the basis vector I pI basis of the illumination spectral distribution and the basis vector R of the surface reflectance of the object.
- the equation calculation means 132 solves the observation equation shown in Equation (3).
- the illumination / reflectance constraint range storage memory 12 of the first embodiment holds basis vectors of illumination spectral distribution and object surface reflectance, and uses these basis vectors to limit the range constrained by various constraint methods. Outputs distributed illumination spectral distribution and object surface reflectance. For example, in the case of a spectral distribution of illumination, as shown in FIG. 8, it is considered that the directly achieved component I 1 basis and the scattered component I 2 basis , which are components of the solar radiation spectrum calculated from the solar radiation model, are used as the basis vectors. It is done. In this solar radiation model, the distribution range of the spectral distribution of illumination can be restricted by easily obtainable illumination conditions such as date and place.
- the mean vector and principal component vector of the relative spectral distribution of natural daylight in CIE daylight shown in Non-Patent Document 1 are used as the basis vector of the spectral distribution of illumination. It is possible.
- the distribution range of the illumination spectral distribution can be restricted by limiting the number of principal component vectors or calculating the principal component vector coefficients based on the correlated color temperature using the method disclosed in Non-Patent Document 1. .
- the surface reflectance of an object it is conceivable to use the average vector and principal component vector of the reflectance database of the object measured as the subject as basis vectors, and by limiting the number of principal component vectors,
- the range of reflectance distribution can be restricted. For example, it is conceivable to use the average vector and the principal component vector generated from the Macbeth color chart shown in FIG. 9 as the basis vectors.
- the illumination / reflectance constraint range storage memory 12 shown in FIG. 2 includes a solar radiation spectrum component calculation means 121, a surface reflectance basis vector storage memory 122, and a finite-dimensional linear sum generation means 123.
- the finite-dimensional linear sum generation means 123 generates variable weight coefficients a pI and b pR for the solar radiation spectrum component I pI basis and the object surface reflectance basis vector R pR basis held as the basis vectors of the illumination spectral distribution. As shown in the equation (2), by taking a linear sum, the spectral distribution I of illumination and the surface reflectance R of the object are generated as an N-dimensional vector.
- FIG. 10 is a diagram showing an image of the restriction range of the illumination spectral distribution.
- N 3
- P I 2
- the spectral distribution of illumination represented as a linear sum of basis vectors is a plane.
- the range that the spectral distribution of illumination can take is restricted.
- the spectral distribution I of the illumination and the surface reflectance R of the object R are generated by the method of Expression (2), so that the spectral distribution I of the illumination and the surface reflectance R of the object can be taken in an N-dimensional space.
- the range is also limited. In this way, the constraint range of the illumination spectral distribution I and the surface reflectance R of the object is calculated.
- FIG. 3 is a flowchart showing the operation of the spectral image processing method according to the first embodiment for carrying out the present invention.
- the multispectral image capturing means 111 acquires multispectral information from the scene and outputs a multispectral image (step S101).
- the spectrum selection unit 112 selects a spectrum from the multispectral image and outputs a sufficient number of linearly independent spectrum information (step S102).
- the solar radiation spectrum component calculating means 121 automatically acquires illumination conditions that can be automatically acquired (step S103).
- the solar radiation spectrum component calculation means 121 calculates and outputs the solar radiation spectrum component (step S104).
- the limited-dimensional linear sum generation means 123 is based on the solar spectral component and the surface reflectance basis vector stored in the surface reflectance basis vector storage memory 122, and restricts the spectral distribution of illumination and the surface reflectance of the object. Each range is generated (step S105).
- the observation equation generating means 131 generates an observation equation that can obtain an optimal solution using a sufficient number of linearly independent spectral information, illumination spectral distribution, and object surface reflectance constraints (step S106).
- the equation calculation means 132 solves the observation equation that obtains the optimum solution, and outputs the surface reflectance of the object and the spectral distribution of illumination (step S107).
- a sufficient number of linearly independent spectral information is acquired from the scene by the multispectral information acquisition means 11, and the illumination spectral distribution and object distribution held in the illumination / reflectance constraint range storage memory 12 are acquired.
- the constraint range of the surface reflectance it is possible to estimate the spectral distribution of the illumination and the surface reflectance of the object using the spectral estimation means 13 without color assumption without using the assumption of the object color or illumination color. Yes. Therefore, even if the scene does not satisfy a specific assumption, it is possible to accurately estimate the spectral distribution of illumination and the surface reflectance of the object with high accuracy.
- the spectral image processing method calculates a solar spectrum component using the solar spectrum component calculation means 121, thereby enabling a narrow restricted range in illumination to be output.
- the observation equation generating unit 131 can generate an observation equation that can be easily calculated by reducing the number of parameters to be estimated.
- the equation calculating unit 132 performs estimation with high accuracy. It is possible to output the measured object surface reflectance and illumination spectral distribution.
- FIG. 4 is a block diagram showing a second embodiment for carrying out the present invention.
- the spectral estimation means 23 without color assumption is the spectral estimation means without color assumption in the first embodiment.
- the other components are the same as in the first embodiment. Constituent elements similar to those in the first embodiment are denoted by the same reference numerals as those in FIG. 2, and detailed description thereof is omitted.
- the spectral estimation unit 23 without color assumption includes an illumination / reflectance estimation value calculation unit 231, an observation spectrum estimation value calculation unit 232, an error calculation unit 233, and an error optimization unit 234.
- the illumination / reflectance estimated value calculation means 231 is a spectral distribution of illumination under the constraint range of the illumination spectral distribution I ( ⁇ n ) and the surface reflectance R ( ⁇ n ) of the object as shown in Equation (2). Then, by giving an initial parameter to the weight coefficient of the surface reflectance of the object, the spectral characteristics of illumination and the estimated value of the surface reflectance of the object are calculated.
- the estimated observation spectrum estimation value calculation means 232 substitutes the estimated values of the illumination spectral characteristics and the object surface reflectance calculated by the illumination / reflectance estimation value calculation means 231 into the right side of the equation (3) to estimate the observation spectrum. Calculate the value.
- the error calculation means 233 considers the error of the observation equation (3), and the observation spectrum estimation value (right side of the expression (3)) calculated by the estimated observation spectrum estimation value calculation means 232 and a sufficient number for optimization.
- An error is calculated using the observed value E m ( ⁇ n ) obtained from a plurality of multispectrums given as color information. Assuming that the spectral distribution of illumination, the surface reflectance of the object, and the measurement error of the observation spectrum are ⁇ I ( ⁇ n ), ⁇ R ( ⁇ n ), and ⁇ L ( ⁇ n ), the observation equation (3) is It can be expressed as 4).
- a method of approximating optimization of these measurement errors for example, there is a method of defining a least square error energy as shown in Expression (5) and minimizing this energy.
- the error optimizing means 234 optimizes the error calculated by the error calculating means 233 using a nonlinear optimization method such as Levenberg-Marquardt method or a coarse / fine search method. Specifically, the parameter given to the illumination / reflectance estimated value calculation means 231 is repeatedly updated so that the error is minimized, and finally the surface reflectance of the object and the spectral distribution of illumination calculated from the optimum parameters Is output as the optimal solution.
- a nonlinear optimization method such as Levenberg-Marquardt method or a coarse / fine search method.
- the spectral estimation means 13 without color assumption outputs the spectral distribution of the illumination calculated by the error optimization means 234 and the estimated value of the surface reflectance of the object.
- FIG. 5 is a flowchart showing an example of the operation of the spectral image processing method in the second embodiment for carrying out the present invention.
- symbol same as FIG. 3 is attached
- subjected and detailed description is abbreviate
- the second embodiment differs from the first embodiment in the following points.
- the illumination / reflectance estimated value calculation means 231 calculates an estimated value of the illumination spectral distribution and the object surface reflectance based on the initial parameter, the restriction range of the illumination spectral distribution and the object surface reflectance (step S206). ).
- the observed spectrum estimated value calculating means 232 calculates an estimated value of the observed spectrum based on the spectral distribution of illumination and the estimated value of the surface reflectance of the object (step S207).
- the error calculation means 233 calculates an error between the observed spectrum estimation value and a sufficient number of observed linearly independent spectrum information (step S208).
- the error optimizing means 234 uses the error to calculate a value set for each error optimizing means, and if the calculated value is smaller than a predetermined value, the operation proceeds to step S108, and if not smaller, the operation proceeds to step S210. Move (step S209).
- the error optimizing unit 234 updates the parameter given to the illumination / reflectance estimated value calculating unit 231 so as to minimize the error, and returns to the operation of S206 (step S210).
- FIG. 6 is a block diagram showing a third embodiment for carrying out the present invention.
- the multispectral information acquisition means 21 is connected to the multispectral information acquisition means 11 of the first embodiment.
- the other components are the same as those in the first embodiment.
- Constituent elements similar to those in the first embodiment are denoted by the same reference numerals as those in FIG. 2, and detailed description thereof is omitted.
- the multispectral information acquisition means 21 includes multispectral multiple times measurement means 211.
- the multispectral multi-times measuring means 211 is shown in the formula (1) within a range in which the illumination colors can be regarded as the same so as to cope with a case where a sufficient number of linearly independent spectral information is not included in the observed scene. Different scenes are photographed until a sufficient number of linearly independent spectral information satisfying the above condition is reached, and a plurality of pieces of spectral information having a sufficient number of linearly independent relationships sufficient for optimization in the subsequent stage are output.
- the multiple times measurement unit 211 repeats the scene shooting until a sufficiently linearly independent observation spectrum is obtained (moving image shooting). Can also be used).
- FIG. 7 is a flowchart showing an example of operation of the spectral image processing method according to the third embodiment for carrying out the present invention.
- symbol same as FIG. 3 is attached
- subjected and detailed description is abbreviate
- the third embodiment is different from the first embodiment in the following points.
- the multispectral multiple times measurement means 211 measures multispectral information from the scene (step S301).
- the multispectral multiple-number measuring unit 211 moves to step S106 if the multispectral information measured from the scene is a sufficient number of linearly independent multispectral information, and otherwise moves to step S301 (step S302). .
- the spectral image processing method includes the multispectral multiple measurement means 211, and photographs different scenes until a sufficient number of linearly independent spectral information satisfying the condition shown in Expression (1) is reached.
- a sufficient number of linearly independent spectral information is not included in the observed scene, an effect of obtaining a sufficient number of linearly independent spectral information can be obtained.
- each unit can be configured by hardware, but can also be realized by a computer program.
- functions and operations similar to those of the above-described embodiments are realized by a processor that operates according to a program stored in the program memory.
- (Appendix 1) Select observation spectrum information from the multispectral image of the scene, Using the illumination basis vector and the surface reflectance basis vector, generate a constraint range for the spectral distribution of the illumination and a constraint range for the surface reflectance of the object, Using the observation spectrum information, the constraint range of the spectral distribution of the illumination, and the constraint range of the surface reflectance of the object, a weighting factor of the basis vector of the illumination and a weighting factor of the basis vector of the surface reflectance, Generate an observation equation with A spectral image processing method for calculating a spectral distribution of illumination of the scene and a surface reflectance of an object from the observation equation.
- Supplementary Note 1 calculates the spectral distribution of illumination and the surface reflectance of an object that constitute the estimated value of the observed spectrum when the error falls within a predetermined range as the spectral distribution of illumination and the surface reflectance of the object in the scene.
- Spectral information selection means for selecting observation spectral information from a multispectral image of a scene; Using a basis vector of illumination and a basis vector of surface reflectance, a constraint range generating means for generating a constraint range of a spectral distribution of illumination and a constraint range of surface reflectance of an object; Using the observation spectrum information, the constraint range of the spectral distribution of the illumination, and the constraint range of the surface reflectance of the object, a weighting factor of the basis vector of the illumination and a weighting factor of the basis vector of the surface reflectance, An observation equation generating means for generating an observation equation having as a parameter; A spectral image processing apparatus comprising: calculation means for calculating a spectral distribution of illumination of the scene and a surface reflectance of an object from the observation equation.
- the calculation means includes: Under the constraint range of the spectral distribution of the illumination and the constraint range of the surface reflectance of the object, a means for calculating an estimated value of the observation spectrum by giving an initial parameter to the weighting coefficient of the observation equation; Means for updating the weighting factor based on an error between the observed spectrum information and the estimated value of the observed spectrum; Means for calculating the spectral distribution of illumination and the surface reflectance of the object that constitute the estimated value of the observed spectrum when the error falls within a predetermined range as the spectral distribution of illumination and the surface reflectance of the object in the scene;
- the spectral image processing apparatus according to appendix 6.
- the spectral information selecting means may observe the linearly independent spectral information equal to or greater than the sum of the number of weighting coefficients of the basis vectors of the illumination and the number of weighting coefficients of the basis vectors of the surface reflectance.
- the spectral image processing apparatus according to appendix 8 or appendix 9, which measures a plurality of times until possible.
- Appendix 11 A process of selecting observed spectral information from a multispectral image of the scene; Using the basis vector of illumination and the basis vector of surface reflectance to generate a constraint range of the spectral distribution of illumination and a constraint range of the surface reflectance of the object, Using the observation spectrum information, the constraint range of the spectral distribution of the illumination, and the constraint range of the surface reflectance of the object, a weighting factor of the basis vector of the illumination and a weighting factor of the basis vector of the surface reflectance, A process for generating an observation equation with a parameter as A program for executing a process of calculating a spectral distribution of illumination of the scene and a surface reflectance of an object from the observation equation.
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Spectrometry And Color Measurement (AREA)
- Image Processing (AREA)
- Facsimile Image Signal Circuits (AREA)
- Color Image Communication Systems (AREA)
Abstract
Description
図1は、本発明による第1の実施の形態の概要を示すブロック図である。図1に示す第1の実施の形態は、マルチスペクトル情報取得手段11と、照明・反射率制約範囲保持メモリ12と、色仮定なし分光推定手段13を備える。
図4は、本発明を実施するための第2の実施の形態を示すブロック図である。図4にも示されるように、本発明を実施するための第2の実施の形態における分光画像処理方法では、色仮定なし分光推定手段23が第1の実施の形態における色仮定なし分光推定手段13と異なり、その他の構成要素については第1の実施の形態と同様である。第1の実施の形態と同様の構成要素については、図2と同一の符号を付し、詳細な説明を省略する。
図6は、本発明を実施するための第3の実施の形態を示すブロック図である。図6にも示されるように、本発明を実施するための第3の実施の形態における分光画像処理方法では、マルチスペクトル情報取得手段21が第1の実施の形態のマルチスペクトル情報取得手段11と異なり、その他の構成要素については第1の実施の形態と同様である。第1の実施の形態と同様の構成要素については、図2と同一の符号を付し、詳細な説明を省略する。
照明の基底ベクトルと表面反射率の基底ベクトルを用いて、照明の分光分布の制約範囲と物体の表面反射率の制約範囲を生成し、
前記観測スペクトル情報と、前記照明の分光分布の制約範囲と、前記物体の表面反射率の制約範囲とを用いて、前記照明の基底ベクトルの重み係数と前記表面反射率の基底ベクトルの重み係数とをパラメタとする観測方程式を生成し、
前記観測方程式から前記シーンの照明の分光分布と物体の表面反射率とを算出する
分光画像処理方法。
前記観測スペクトル情報と前記観測スペクトルの推定値の誤差に基づき、前記重み係数を更新し、
前記誤差が所定内となったときの前記観測スペクトルの推定値を構成する照明の分光分布と物体の表面反射率とを、前記シーンにおける照明の分光分布と物体の表面反射率として算出する
付記1に記載の分光画像処理方法。
付記1又は付記2に記載の分光画像処理方法。
特徴とする付記1から付記3のいずれかに記載の分光画像処理方法。
付記3又は付記4に記載の分光画像処理方法。
照明の基底ベクトルと表面反射率の基底ベクトルを用いて、照明の分光分布の制約範囲と物体の表面反射率の制約範囲を生成する制約範囲生成手段と、
前記観測スペクトル情報と、前記照明の分光分布の制約範囲と、前記物体の表面反射率の制約範囲とを用いて、前記照明の基底ベクトルの重み係数と前記表面反射率の基底ベクトルの重み係数とをパラメタとする観測方程式を生成する観測方程式生成手段と、
前記観測方程式から前記シーンの照明の分光分布と物体の表面反射率とを算出する算出手段と
を有する分光画像処理装置。
前記照明の分光分布の制約範囲と前記物体の表面反射率の制約範囲のもと、前記観測方程式の重み係数に初期パラメタを与えて、前記観測スペクトルの推定値を算出する手段と、
前記観測スペクトル情報と前記観測スペクトルの推定値の誤差に基づき、前記重み係数を更新する手段と、
前記誤差が所定内となったときの前記観測スペクトルの推定値を構成する照明の分光分布と物体の表面反射率とを前記シーンにおける照明の分光分布と物体の表面反射率として算出する手段と
を有する付記6に記載の分光画像処理装置。
付記6又は付記7に記載の分光画像処理装置。
特徴とする付記6から付記8のいずれかに記載の分光画像処理装置。
付記8又は付記9に記載の分光画像処理装置。
シーンのマルチスペクトル画像から観測スペクトル情報を選択する処理と、
照明の基底ベクトルと表面反射率の基底ベクトルを用いて、照明の分光分布の制約範囲と物体の表面反射率の制約範囲を生成する処理と、
前記観測スペクトル情報と、前記照明の分光分布の制約範囲と、前記物体の表面反射率の制約範囲とを用いて、前記照明の基底ベクトルの重み係数と前記表面反射率の基底ベクトルの重み係数とをパラメタとする観測方程式を生成する処理と、
前記観測方程式から前記シーンの照明の分光分布と物体の表面反射率とを算出する処理と
を実行させるプログラム。
2 分光推定手段
3 照明・反射率主成分ベクトル保存メモリ
11 マルチスペクトル情報取得手段
12 色仮定なし分光推定手段
13 照明・反射率制約範囲保存メモリ
21 マルチスペクトル情報取得手段
23 色仮定なし分光推定手段
111 マルチスペクトル画像撮影手段
112 スペクトル選択手段
121 日射スペクトル成分計算手段
122 表面反射率基底ベクトル保存メモリ
123 有限次元線型和生成手段
131 観測方程式生成手段
132 方程式計算手段
211 マルチスペクトル複数回計測手段
231 照明・反射率推定値算出手段
232 観測スペクトル推定値算出手段
233 誤差算出手段
234 誤差最適化手段
Claims (10)
- シーンのマルチスペクトル画像から観測スペクトル情報を選択し、
照明の基底ベクトルと表面反射率の基底ベクトルを用いて、照明の分光分布の制約範囲と物体の表面反射率の制約範囲を生成し、
前記観測スペクトル情報と、前記照明の分光分布の制約範囲と、前記物体の表面反射率の制約範囲とを用いて、前記照明の基底ベクトルの重み係数と前記表面反射率の基底ベクトルの重み係数とをパラメタとする観測方程式を生成し、
前記観測方程式から前記シーンの照明の分光分布と物体の表面反射率とを算出する
分光画像処理方法。 - 前記照明の分光分布の制約範囲と前記物体の表面反射率の制約範囲のもと、前記観測方程式の重み係数に初期パラメタを与えて、前記観測スペクトルの推定値を算出し、
前記観測スペクトル情報と前記観測スペクトルの推定値の誤差に基づき、前記重み係数を更新し、
前記誤差が所定内となったときの前記観測スペクトルの推定値を構成する照明の分光分布と物体の表面反射率とを、前記シーンにおける照明の分光分布と物体の表面反射率として算出する
請求項1に記載の分光画像処理方法。 - 前記観測スペクトル情報として、前記マルチスペクトル画像から線形独立なスペクトル情報を選択する
請求項1又は請求項2に記載の分光画像処理方法。 - 前記照明の基底ベクトルとして、日射スペクトルの直達成分ベクトルと散乱成分ベクトルとを用いる
特徴とする請求項1から請求項3のいずれかに記載の分光画像処理方法。 - 前記線形独立なスペクトル情報が前記照明の基底ベクトルの重み係数の個数と前記表面反射率の基底ベクトルの重み係数の個数の総和と同数かそれより多く観測できるまで複数回計測する
請求項3又は請求項4に記載の分光画像処理方法。 - シーンのマルチスペクトル画像から観測スペクトル情報を選択するスペクトル情報選択手段と、
照明の基底ベクトルと表面反射率の基底ベクトルを用いて、照明の分光分布の制約範囲と物体の表面反射率の制約範囲を生成する制約範囲生成手段と、
前記観測スペクトル情報と、前記照明の分光分布の制約範囲と、前記物体の表面反射率の制約範囲とを用いて、前記照明の基底ベクトルの重み係数と前記表面反射率の基底ベクトルの重み係数とをパラメタとする観測方程式を生成する観測方程式生成手段と、
前記観測方程式から前記シーンの照明の分光分布と物体の表面反射率とを算出する算出手段と
を有する分光画像処理装置。 - 前記算出手段は、
前記照明の分光分布の制約範囲と前記物体の表面反射率の制約範囲のもと、前記観測方程式の重み係数に初期パラメタを与えて、前記観測スペクトルの推定値を算出する手段と、
前記観測スペクトル情報と前記観測スペクトルの推定値の誤差に基づき、前記重み係数を更新する手段と、
前記誤差が所定内となったときの前記観測スペクトルの推定値を構成する照明の分光分布と物体の表面反射率とを前記シーンにおける照明の分光分布と物体の表面反射率として算出する手段と
を有する請求項6に記載の分光画像処理装置。 - 前記スペクトル情報選択手段は、前記マルチスペクトル画像から線形独立なスペクトル情報を選択する
請求項6又は請求項7に記載の分光画像処理装置。 - 前記スペクトル情報選択手段は、前記照明の基底ベクトルとして、日射スペクトルの直達成分ベクトルと散乱成分ベクトルとを用いる
特徴とする請求項6から請求項8のいずれかに記載の分光画像処理装置。 - コンピュータに、
シーンのマルチスペクトル画像から観測スペクトル情報を選択する処理と、
照明の基底ベクトルと表面反射率の基底ベクトルを用いて、照明の分光分布の制約範囲と物体の表面反射率の制約範囲を生成する処理と、
前記観測スペクトル情報と、前記照明の分光分布の制約範囲と、前記物体の表面反射率の制約範囲とを用いて、前記照明の基底ベクトルの重み係数と前記表面反射率の基底ベクトルの重み係数とをパラメタとする観測方程式を生成する処理と、
前記観測方程式から前記シーンの照明の分光分布と物体の表面反射率とを算出する処理と
を実行させるプログラム。
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US14/369,298 US9491331B2 (en) | 2011-12-28 | 2012-12-13 | Image processing method, image processing device, and program for restoring spectral characteristics of an object and an illumination, respectively |
JP2013551611A JP6015964B2 (ja) | 2011-12-28 | 2012-12-13 | 分光画像処理方法、分光画像処理装置およびプログラム |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2011288886 | 2011-12-28 | ||
JP2011-288886 | 2011-12-28 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2013099644A1 true WO2013099644A1 (ja) | 2013-07-04 |
Family
ID=48697136
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2012/082417 WO2013099644A1 (ja) | 2011-12-28 | 2012-12-13 | 分光画像処理方法、分光画像処理装置およびプログラム |
Country Status (3)
Country | Link |
---|---|
US (1) | US9491331B2 (ja) |
JP (1) | JP6015964B2 (ja) |
WO (1) | WO2013099644A1 (ja) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2015036850A (ja) * | 2013-08-12 | 2015-02-23 | 日本電気株式会社 | 画像処理装置、画像処理方法および画像処理プログラム |
JP2015102517A (ja) * | 2013-11-27 | 2015-06-04 | 株式会社リコー | 撮像装置、撮像システムおよび当該撮像装置による撮像方法 |
WO2018016555A1 (ja) * | 2016-07-22 | 2018-01-25 | 日本電気株式会社 | 画像処理装置、画像処理方法及び記録媒体 |
JP2023048996A (ja) * | 2021-09-28 | 2023-04-07 | 三星電子株式会社 | ホワイトバランスを遂行する映像獲得装置、及びそれを含む電子装置、ならびに映像獲得装置の制御方法 |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2016103566A1 (ja) * | 2014-12-25 | 2016-06-30 | 日本電気株式会社 | 画像処理方法及び画像処理装置 |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2007007788A1 (ja) * | 2005-07-13 | 2007-01-18 | Nec Corporation | 色補正方法および色補正装置 |
WO2007007786A1 (ja) * | 2005-07-13 | 2007-01-18 | Nec Corporation | 色補正方法および色補正装置 |
JP2009175926A (ja) * | 2008-01-23 | 2009-08-06 | Toppan Printing Co Ltd | 分光反射率近似装置、分光反射率推定装置及びその方法並びにそのプログラム |
WO2009104787A1 (ja) * | 2008-02-22 | 2009-08-27 | 日本電気株式会社 | カラー画像処理方法、装置およびプログラム |
Family Cites Families (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4648051A (en) * | 1984-10-15 | 1987-03-03 | The Board Of Trustees Of The Leland Stanford Junior University | Color imaging process |
JP3149297B2 (ja) | 1993-08-25 | 2001-03-26 | 東洋インキ製造株式会社 | カラーバランス調整装置および調整方法 |
US6304294B1 (en) * | 1999-10-15 | 2001-10-16 | Sony Corporation | Apparatus for and method of estimating the illumination of an image |
JP3853224B2 (ja) | 2001-11-08 | 2006-12-06 | 富士写真フイルム株式会社 | ホワイトバランス補正方法 |
CN102597744B (zh) * | 2009-09-03 | 2016-09-07 | 澳大利亚国家Ict有限公司 | 照明谱恢复 |
-
2012
- 2012-12-13 JP JP2013551611A patent/JP6015964B2/ja active Active
- 2012-12-13 WO PCT/JP2012/082417 patent/WO2013099644A1/ja active Application Filing
- 2012-12-13 US US14/369,298 patent/US9491331B2/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2007007788A1 (ja) * | 2005-07-13 | 2007-01-18 | Nec Corporation | 色補正方法および色補正装置 |
WO2007007786A1 (ja) * | 2005-07-13 | 2007-01-18 | Nec Corporation | 色補正方法および色補正装置 |
JP2009175926A (ja) * | 2008-01-23 | 2009-08-06 | Toppan Printing Co Ltd | 分光反射率近似装置、分光反射率推定装置及びその方法並びにそのプログラム |
WO2009104787A1 (ja) * | 2008-02-22 | 2009-08-27 | 日本電気株式会社 | カラー画像処理方法、装置およびプログラム |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2015036850A (ja) * | 2013-08-12 | 2015-02-23 | 日本電気株式会社 | 画像処理装置、画像処理方法および画像処理プログラム |
JP2015102517A (ja) * | 2013-11-27 | 2015-06-04 | 株式会社リコー | 撮像装置、撮像システムおよび当該撮像装置による撮像方法 |
WO2018016555A1 (ja) * | 2016-07-22 | 2018-01-25 | 日本電気株式会社 | 画像処理装置、画像処理方法及び記録媒体 |
US11519781B2 (en) | 2016-07-22 | 2022-12-06 | Nec Corporation | Image processing device, image processing method, and recording medium |
JP2023048996A (ja) * | 2021-09-28 | 2023-04-07 | 三星電子株式会社 | ホワイトバランスを遂行する映像獲得装置、及びそれを含む電子装置、ならびに映像獲得装置の制御方法 |
JP7504170B2 (ja) | 2021-09-28 | 2024-06-21 | 三星電子株式会社 | ホワイトバランスを遂行する映像獲得装置、及びそれを含む電子装置、ならびに映像獲得装置の制御方法 |
Also Published As
Publication number | Publication date |
---|---|
US9491331B2 (en) | 2016-11-08 |
JP6015964B2 (ja) | 2016-10-26 |
US20150207959A1 (en) | 2015-07-23 |
JPWO2013099644A1 (ja) | 2015-05-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP6015964B2 (ja) | 分光画像処理方法、分光画像処理装置およびプログラム | |
KR20050074590A (ko) | 광원 추정 장치, 광원 추정 방법, 촬상 장치 및 화상 처리방법 | |
JP5988093B2 (ja) | 画像処理装置、物体識別装置、及びプログラム | |
JP6943251B2 (ja) | 画像処理装置、画像処理方法及びコンピュータ読み取り可能記録媒体 | |
JPWO2016114148A1 (ja) | 画像処理装置、画像処理方法およびプログラム | |
JP6201507B2 (ja) | 画像処理装置、画像処理方法および画像処理プログラム | |
JP6376128B2 (ja) | 照明推定装置、照明推定方法および照明推定プログラム | |
CN109389646B (zh) | 一种利用多光谱图像进行彩色相机辐射定标的方法 | |
JP6443230B2 (ja) | 処理プログラム、処理方法、及び処理装置 | |
JP2013124990A (ja) | 画像処理装置および画像処理方法 | |
JP4174707B2 (ja) | 分光測定システム、色再現システム | |
JP6023653B2 (ja) | 照明光色推定装置、照明光色推定方法及びプログラム | |
JP4615430B2 (ja) | 画像生成装置、画像生成方法および画像生成プログラム | |
Lam et al. | Spectral Imaging Using Basis Lights. | |
CN108430302B (zh) | 图像处理装置、图像处理方法和存储介质 | |
JP6478492B2 (ja) | 画像処理装置およびその方法 | |
WO2016208142A1 (ja) | 画像処理システム、画像処理方法、およびコンピュータ読取可能な記録媒体 | |
JP2011022044A (ja) | 光源推定装置および光源推定方法 | |
KR101630856B1 (ko) | 다분광 포토메트릭 스테레오 시스템 그리고 이의 동작 방법 | |
JP2011095110A (ja) | 光源推定装置および光源推定方法 | |
JP6230504B2 (ja) | 画像処理装置、方法、及びプログラム | |
JP2007221624A (ja) | 照度取得装置、照度取得方法および照度取得プログラム | |
JP2020091256A (ja) | 処理装置、撮像装置、処理方法、およびプログラム | |
JP7309092B1 (ja) | 情報処理装置、判定方法、及び判定プログラム | |
JP2018036177A (ja) | 分光データ処理装置、撮像装置、分光データ処理方法および分光データ処理プログラム |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 12861164 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2013551611 Country of ref document: JP Kind code of ref document: A |
|
WWE | Wipo information: entry into national phase |
Ref document number: 14369298 Country of ref document: US |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 12861164 Country of ref document: EP Kind code of ref document: A1 |