Signal processing device, signal processing method, program, and lighting device

Abstract

According to the signal processing device of the present technology, the signal processing device performs a calculation process satisfying (condition 1) as a calculation process for reproducing a phase distribution of a target light intensity distribution on a projection surface by spatial light phase modulation of incident light: uses a feedback loop in which updating of the phase distribution is repeated by including a nonlinear ray optics model that is a ray optics model including a nonlinear term and an inverse operation model of the model that is obtained by linearizing the nonlinear ray optics model, finds an error distribution between the target light intensity distribution and a light intensity distribution calculated based on a provisional value of the phase distribution based on the nonlinear ray optics model, and obtains a value obtained by multiplying the error distribution by a feedback gain as a light intensity correction value, and adds a phase correction value to the provisional value by an output at the time when the light intensity correction value is input to the inverse operation model as a phase correction value.

Description

Technical Field

[0001] This technology relates to a signal processing apparatus and method, a program for performing computational processing of a phase distribution for reproducing a target light intensity distribution on a projection plane by spatial light phase modulation of incident light, and to an illumination apparatus for reproducing a target light intensity distribution on a projection plane by spatial light phase modulation of incident light. Background Technology

[0002] Techniques for reproducing a desired image (light intensity distribution) by performing spatial light modulation on incident light using a liquid crystal panel and a spatial light modulator (SLM) such as a DMD (digital micromirror device) are known. For example, a technique for reproducing a desired image by performing spatial light intensity modulation on incident light is well known.

[0003] Meanwhile, techniques for projecting a desired reconstructed image by performing spatial light phase modulation on the incident light are also known (see, for example, Patent Document 1 below). When spatial light intensity modulation is performed, the incident light is partially attenuated or blocked in the SLM when reconstructing the desired light intensity distribution. However, when spatial light phase modulation is performed, the desired light intensity distribution can be reconstructed without incident light attenuation or blocking in the SLM, thus improving light utilization efficiency.

[0004] Existing technical documents

[0005] Patent documents

[0006] Patent Document 1

[0007] International Patent Application No. 2017-520022 Summary of the Invention

[0008] [Technical Issues]

[0009] In the case of using spatial light phase modulation, free-form methods, represented by the method disclosed in Patent Document 1, are known as methods for determining the phase distribution used to reproduce the target image (target light intensity distribution).

[0010] However, the conventional free-form method disclosed in Patent Document 1 rewrites the problem into an easily solvable form by approximating the formula of the ray-optics model (based on the ray-optics light propagation model), which originally included nonlinear terms, when calculating the phase distribution used to reproduce the target light intensity distribution. This leads to a trend of reduced reproducibility of the reproduced image relative to the target light intensity distribution.

[0011] In view of the above, this technique is proposed. The purpose of this technique is to improve the reproducibility of the reproduced image with respect to the target light intensity distribution.

[0012] [Solution to the problem]

[0013] The signal processing apparatus according to this technology performs a computational process satisfying "Condition 1" as a phase distribution computational process for reproducing the target light intensity distribution on a projection plane by spatial light phase modulation of the incident light. "Condition 1" specifies that the computational process includes a nonlinear ray-optical model comprising a ray-optical model including nonlinear terms and an inverse computational model relating to the model obtained by linearizing the nonlinear ray-optical model. It determines an error distribution of the error between the target light intensity distribution and the light intensity distribution calculated by the nonlinear ray-optical model based on a temporary value of the phase distribution. A light intensity correction value is obtained by multiplying the error distribution by a feedback gain. The light intensity correction value is input to the inverse computational model to obtain an output. The obtained output is considered as a phase correction value, and a feedback loop is used to repeatedly update the phase distribution by adding the phase correction value to the temporary value.

[0014] Using models such as the aforementioned ray-optics model that includes nonlinear terms makes it possible to accurately determine the phase distribution used to reproduce the target light intensity distribution.

[0015] The signal processing apparatus described above according to this technology can alternatively be configured to perform the calculation processing of the phase distribution in a manner that satisfies "Condition 1" and "Condition 2" above. "Condition 2" specifies the incorporation of a term concerning the light intensity distribution of the incident light into the nonlinear ray-optical model.

[0016] This alternative configuration ensures that the phase distribution calculation is performed by using the feedback loop specified by "Condition 1", which enables the determination of the phase distribution to eliminate the incident light intensity distribution and reproduce the target light intensity distribution.

[0017] The signal processing apparatus described above according to this technology can alternatively be configured to perform the calculation processing of the phase distribution in a manner that satisfies "Condition 1" and "Condition 3" above. "Condition 3" specifies that the term of the light intensity distribution of the incident light is incorporated into both the nonlinear ray-optical model and the inverse calculation model.

[0018] Since the incident light intensity distribution term is incorporated not only into the ray optical model but also into the inverse calculation model, convergence stabilization can be achieved in the phase distribution calculation performed by using the feedback loop specified in "Condition 1".

[0019] The signal processing apparatus described above according to this technology can alternatively be configured to control the feedback gain based on the absolute value of the error distribution.

[0020] When using the feedback loop specified in "Condition 1", the light intensity correction value input to the inverse calculation model needs to be small enough to ensure the reliability of the phase correction value calculated by the inverse calculation model. As mentioned above, by controlling the feedback gain according to the absolute value of the error distribution, it is possible to prevent the light intensity correction value input to the inverse calculation model from being too large.

[0021] The signal processing apparatus described above according to the present technology may alternatively be configured such that, if the maximum value of the absolute value of the light intensity correction value obtained by multiplying the error distribution by a constant-based feedback gain exceeds a predetermined value, the feedback gain is controlled to reduce the maximum value of the absolute value of the light intensity correction value to a value not greater than the predetermined value, and if the maximum value of the absolute value of the light intensity correction value obtained by multiplying the error distribution by a constant-based feedback gain does not exceed the predetermined value, the constant is used as the feedback gain.

[0022] This alternative configuration allows temporary values ​​of the phase distribution to be repeatedly corrected by multiplying the error distribution by a small light intensity correction value obtained through progressively adjusted feedback gain, even under conditions of large errors. Then, when the error distribution is equal to or less than a predetermined value through repeated correction, increased convergence is provided by changing the feedback gain to a constant.

[0023] According to the signal processing method of this technology, a signal processing device performs a computational process that satisfies "Condition 1" as a computational process for reproducing the phase distribution of the target light intensity distribution on the projection plane by spatial light phase modulation of the incident light. "Condition 1" specifies that the computational process includes a nonlinear ray-optical model that includes a nonlinear term and an inverse computational model about the model obtained by linearizing the nonlinear ray-optical model. The error distribution between the target light intensity distribution and the light intensity distribution calculated by the nonlinear ray-optical model based on the temporary value of the phase distribution is determined. A light intensity correction value is obtained by multiplying the error distribution by the feedback gain. The light intensity correction value is input to the inverse computational model to obtain an output. The obtained output is regarded as the phase correction value. A feedback loop is used to repeatedly update the phase distribution by adding the phase correction value to the temporary value.

[0024] The above-described signal processing method also provides operations similar to those performed by the above-described signal processing apparatus according to the present technology.

[0025] Furthermore, the program according to this technology is a computer-readable program, and the program is adapted to cause the computer device to perform a computational process satisfying "Condition 1" as a computational process for reproducing the phase distribution of the target light intensity distribution on the projection plane by spatial light phase modulation of the incident light. "Condition 1" specifies that the computational process includes a nonlinear ray-optical model including a nonlinear term and an inverse computational model of the model obtained by linearizing the nonlinear ray-optical model, determining the error distribution between the target light intensity distribution and the light intensity distribution calculated by the nonlinear ray-optical model based on the temporary value of the phase distribution, obtaining a light intensity correction value by multiplying the error distribution by the feedback gain, inputting the light intensity correction value into the inverse computational model to obtain an output, treating the obtained output as a phase correction value, and using a feedback loop to repeatedly update the phase distribution by adding the phase correction value to the temporary value.

[0026] The above procedure implements the signal processing apparatus according to the previously described art.

[0027] Furthermore, the lighting device according to this technology includes a light source unit, a phase modulation unit, and a signal processing unit. The light source unit has a light-emitting element. The phase modulation unit performs spatial light phase modulation on the incident light from the light source unit. The signal processing unit performs calculation processing that satisfies "Condition 1" as a calculation processing for reproducing the phase distribution of the target light intensity distribution on the projection plane by spatial light phase modulation. "Condition 1" specifies that the calculation processing includes a nonlinear ray-optical model including a nonlinear term and an inverse calculation model of the model obtained by linearizing the nonlinear ray-optical model, determining the error distribution between the target light intensity distribution and the light intensity distribution calculated by the nonlinear ray-optical model based on the temporary value of the phase distribution, obtaining a light intensity correction value by multiplying the error distribution by the feedback gain, inputting the light intensity correction value into the inverse calculation model to obtain an output, treating the obtained output as a phase correction value, and using a feedback loop to repeatedly update the phase distribution by adding the phase correction value to the temporary value.

[0028] The lighting device described above also provides operations similar to those performed by the signal processing device previously described according to the present technology.

[0029] The lighting device described above according to this technology can be alternatively configured such that the light source section has multiple light-emitting elements.

[0030] This alternative configuration eliminates the need to use a single high-output light-emitting element in the light source section to meet predetermined light intensity requirements.

[0031] The above-described illumination device according to the present technology can be alternatively configured such that the signal processing unit performs calculation processing to calculate the phase distribution that satisfies the above-described "condition 1" and "condition 2", wherein "condition 1" and "condition 2" define the terms for incorporating the light intensity distribution of the incident light into the nonlinear ray optical model, including an intensity distribution detection unit for detecting the light intensity distribution of the incident light, and using the light intensity distribution detected by the intensity distribution detection unit as the light intensity distribution to be incorporated into the nonlinear ray-optical model.

[0032] This alternative configuration ensures that, when the incident light intensity distribution changes over time, the nonlinear ray-optics model can reflect the incident light intensity distribution that has changed over time. Attached Figure Description

[0033] Figure 1 This is a diagram illustrating an example configuration of a lighting device according to a first embodiment of the present technology.

[0034] Figure 2 This is an explanatory diagram showing an example of the configuration of a light source unit included in a lighting device according to an embodiment of the present technology.

[0035] Figure 3 This is an explanatory diagram illustrating the principle of image reproduction through spatial phase modulation.

[0036] Figure 4 This is an explanatory diagram illustrating the problems of conventional freeform forming methods.

[0037] Figure 5 It is a diagram showing the relationship between the projection distance and the coordinate system of the phase modulation plane and the projection plane.

[0038] Figure 6 A diagram illustrating the optimized loop provided by the new algorithm used as the basis in the implementation.

[0039] Figure 7 This is a schematic diagram illustrating the dynamic adjustment of the feedback gain in the implementation embodiment.

[0040] Figure 8 This is a diagram illustrating an optimized loop provided by a phase distribution calculation algorithm according to an embodiment.

[0041] Figure 9 This is a diagram illustrating an example configuration of a lighting device according to a second embodiment of the present technology.

[0042] Figure 10 This is a diagram illustrating an example of a configuration of a projector device to which the lighting device according to the first embodiment is applied.

[0043] Figure 11This is a diagram illustrating an example configuration of a projector device using a lighting apparatus according to the second embodiment. Detailed Implementation

[0044] Embodiments of this technology will now be described in the following order with reference to the accompanying drawings.

[0045] <1. First Implementation Method>

[0046] [1-1. Configuration of Lighting Fixture]

[0047] [1-2. Phase distribution calculation method according to the implementation method]

[0048] <2. Second Implementation Method>

[0049] <3. Third Implementation Method (Applicable to Projector Devices)>

[0050] <4. Variations>

[0051] <5. Summary of Implementation Methods>

[0052] <6. This technology>

[0053] <1. First Implementation Method>

[0054] [1-1. Configuration of Lighting Fixture]

[0055] Figure 1 This is a diagram showing an example of the configuration of a lighting device 1 according to a first embodiment of the present technology.

[0056] like Figure 1 As shown, the lighting device 1 includes a light source 2, a phase modulation SLM (spatial light modulator) 3, a driving unit 4, and a control unit 5.

[0057] The lighting device 1 is configured to reproduce a desired image (light intensity distribution) on the projection plane Sp by allowing the phase modulation SLM3 to perform spatial light phase modulation on the incident light from the light source 2. The lighting device 1 described above can be applied, for example, to a vehicle's headlight. When the lighting device 1 is applied to a headlight, the lighting device 1 can be configured to cause the phase modulation SLM 3 to perform spatial light phase modulation to change the illumination range of the high beam or low beam.

[0058] The light source unit 2 serves as a light source that directs light onto the phase-modulated SLM 3. In this embodiment, the light source unit 2, for example... Figure 2 The diagram shows a plurality of light-emitting elements 2a. More specifically, the light source 2 includes a light source having a two-dimensional array of a plurality of light-emitting elements 2a, and light emitted from the plurality of light-emitting elements 2a is incident on a phase modulation SLM 3.

[0059] In this embodiment, a laser light-emitting element is used as the light-emitting element 2a. It should be noted that the light-emitting element 2a is not limited to a laser light-emitting element. For example, an LED (light-emitting diode), a discharge lamp, or other light-emitting elements can be used alternatively as the light-emitting element 2a.

[0060] Phase-modulated SLM 3 includes, for example, a transmissive liquid crystal panel, and performs spatial light phase modulation on the incident light.

[0061] It should be noted that the phase modulation SLM 3 can be alternatively configured as a reflective spatial light phase modulator rather than a transmissive spatial light phase modulator. For example, a reflective liquid crystal panel or a DMD (digital micromirror device) can be used as a reflective spatial light phase modulator.

[0062] The driving unit 4 includes a driving circuit for driving the phase modulation SLM 3. The driving unit 4 is configured to drive the pixels in the phase modulation SLM 3 individually.

[0063] The control unit 5 is configured, for example, as a microcomputer including, for example, a CPU (Central Processing Unit), ROM (Read-Only Memory), and RAM (Random Access Memory). The control unit 5 receives input of a target image and calculates the phase distribution of the phase modulation SLM3 for reproducing the target image on the projection plane Sp. The control unit 5 controls the drive unit 4 to drive the phase modulation SLM3 according to the calculated phase distribution.

[0064] like Figure 1 As shown, the control unit 5 includes a target intensity distribution calculation unit 5a and a phase distribution calculation unit 5b, serving as functional units for calculating the phase distribution of phase modulation SLM3 based on the target image. Based on the target image, the target intensity distribution calculation unit 5a calculates the light intensity distribution to be reproduced on the projection plane Sp (this target light intensity distribution may be referred to as the "target intensity distribution" below). The phase distribution calculation unit 5b calculates the phase distribution of the phase modulation SLM3 used to reproduce the target intensity distribution calculated by the target intensity distribution calculation unit 5a on the projection plane Sp using a free-form method. Here, the free-form method is a general term for methods based on ray-optical calculations used to reproduce the phase distribution of the target light intensity distribution on the projection plane Sp by performing spatial light phase modulation.

[0065] In this example, the processing performed by the target intensity distribution calculation unit 5a and the phase distribution calculation unit 5b allows the CPU to perform software processing based on a program stored in a memory such as ROM.

[0066] Furthermore, the processing of the phase distribution calculation unit 5b will be explained in detail later.

[0067] [1-2. Phase distribution calculation method according to the embodiment]

[0068] First, refer to the following Figure 3 The principle of image reproduction by spatial phase modulation according to embodiments of this technology is described.

[0069] Figure 3 The diagram schematically illustrates the relationship between the light rays incident on the phase modulation surface Sm of the phase modulation SLM 3, the wavefront of the phase distribution in the phase modulation SLM 3, the phase-modulated light rays, and the light intensity distribution formed on the projection surface Sp by the phase-modulated light rays.

[0070] First, as a prerequisite, a smooth curve is plotted to indicate the wavefront of the phase distribution in the phase-modulated SLM 3, since a freeform shaping method is employed. Because the phase-modulated SLM 3 performs spatial light phase modulation, the incident light is refracted to travel in the normal direction of the wavefront of the phase distribution. Due to this refraction, portions with high light density and portions with low light density are formed on the projection plane Sp. This results in a light intensity distribution on the projection plane Sp.

[0071] Based on the above principle, the desired image can be reproduced on the projection plane Sp by setting the phase distribution pattern in the phase modulation SLM 3.

[0072] Here, the free-form method described in the previously mentioned patent document 1 assumes that the light intensity distribution of the light incident on the phase modulation surface Sm is uniform, as in... Figure 4 As shown in A, there is no uniform distribution of light intensity variation in the in-plane direction.

[0073] Therefore, in, for example, Figure 4 As shown in B, because the light incident on the phase modulation plane Sm is partially blocked by the shielding object Oa, the incident light intensity distribution is uneven. In this case, the incident light intensity distribution may be superimposed on the reconstructed image, resulting in the inability to achieve proper image reconstruction.

[0074] Therefore, in order to prevent the incident light intensity distribution from being superimposed on the reconstructed image, this embodiment is configured by revisiting the phase distribution calculation method based on past free-form methods.

[0075] First, refer to equations 1 to 35 and Figures 5 to 7 This describes the phase distribution calculation method on which this embodiment is based.

[0076] The following explanation is provided in reference 1.

[0077] Reference 1: High Brightness HDR Projection Using Dynamic FreeformLensing GERWIN DAMBERG and JAMES GREGSO (DOI: http: / / dx.doi.org / 10.1145 / 2857051)

[0078] Furthermore, the prerequisites described below are as follows.

[0079] Let the imaginary unit be represented as j.

[0080] *The set of (M, N) matrices of scalars belonging to the set of complex numbers C is denoted as C M×N .

[0081] *Matrix element indices start from 0.

[0082] Having A m，n The matrix A, whose elements are in the m-th row and n-th column, is represented as A = {A}. m，n} m，n .

[0083] *By kernel K∈C M×N Performed matrix A∈C M×N The convolution operation is defined as indicated in Equation 1 below.

[0084] [Mathematical Expression 1]

[0085]

[0086] Where mod(·,M) represents the remainder of M.

[0087] First, as a fundamental principle, the intensity distribution reconstructed from a given phase distribution can be calculated using an optical propagation model. However, to reconstruct the intensity distribution, an inverse problem needs to be solved, whereby the phase distribution that realizes this intensity distribution needs to be elucidated. Typically, rigorously solving this inverse problem is very difficult. Therefore, the phase distribution is approximated. Methods for estimating the phase distribution are broadly classified into two types: computer-synthesized holograms (CGH) based on wave optics and free-form methods based on ray-optics. The CGH method considers optical interference phenomena in phase estimation, thus exhibiting excellent rendering capabilities when using coherent light as the incident light source. However, the CGH method requires discretizing the computational region with fine sampling intervals, thus involving high computational costs. Meanwhile, the free-form method is affected by interferences not considered when calculating under coherent light sources, making it difficult to render high-frequency components with fine precision compared to the CGH method. However, algorithms capable of high-speed computation have been proposed based on the free-form method. Previous free-form methods did not perform optimization calculations that converged to an exact solution of the phase distribution (i.e., "any phase distribution with a ray density distribution that best approximates the target intensity distribution"). In contrast, previous free-form methods transformed the problem into a solvable form using formulas, for example, based on an approximate light propagation model of ray optics (ray-optics model).

[0088] The paper designated as Reference 1 proposes an algorithm that uses a proximity method based on the free-form method to estimate the phase distribution (this algorithm is referred to as the "proximity algorithm" below). The mathematical proof and implementation of the proximity algorithm are described below, where ambiguity is eliminated wherever possible.

[0089] First, such as Figure 5 As shown, the coordinate system and projection distance f of the phase modulation plane Sm and the projection plane Sp are defined. Here, it is assumed that the projection distance f is the distance between the phase modulation plane Sm and the projection plane Sp.

[0090] The relationship between the phase distribution P(x, y) in the phase modulation surface Sm and the light intensity distribution I in the projection surface Sp at the propagation destination is based on a ray-optics formulation. In the following description, the light incident on the phase modulation surface Sm is assumed to be a plane wave, in order to consider equidistant grid points x = (x, y) perpendicularly incident on the phase modulation surface Sm. T The ray group is constructed on the projection plane Sp. Furthermore, the average intensity of all light intensity distributions (incident light, the image reproduced by phase modulation, and the target image) is normalized to 1 (i.e., the intensity value at each point corresponds to a rise factor relative to the average intensity). Using the phase distribution P(x, y) in the phase modulation surface Sm, the ray group penetrates the grid points u = (ux, uy) on the projection plane Sp. T It can be represented as shown in Equation 2 below.

[0091] [Mathematical Expression 2]

[0092]

[0093] Let us consider a square microregion surrounded by a grid of points uniformly distributed on the phase modulation plane Sm and adjacent grid points. The microregion on the projection plane Sp corresponding to the microregion on the phase modulation plane Sm is shaped like a parallelogram. The area magnification m(ux, uy) in this case is calculated as follows.

[0094] [Mathematical Expression 3]

[0095]

[0096] When the electric field intensity I(ux, uy) at the grid point (ux, uy) on the projection plane Sp is calculated as the light density distribution 1 / m(ux, uy), the following equation 4 is obtained.

[0097] [Mathematical Expression 4]

[0098]

[0099] Here, I(ux, uy) represents the electric field intensity at the grid point (ux, uy) on the projection plane Sp corresponding to the grid point (x, y) on the phase modulation plane Sm. Therefore, it should be noted that even when the coordinates of the phase modulation plane Sm are sampled at equally spaced grid points (x, y) for numerical calculation purposes, I(ux, uy) only represents the sampled electric field intensity at non-equally spaced grid points (ux, uy) on the projection plane Sp.

[0100] In the proximity algorithm, the intensity distribution indicated in Equation 4 is linearly approximated around P=0, as shown in Equation 5 below.

[0101] [Mathematical Expression 5]

[0102]

[0103] Therefore, the phase distribution P^ (“P^” is the symbol obtained by placing the “~” mark on “I”) that should be determined for the target intensity I^~ (“I^~” is the symbol obtained by placing the “^” mark on “P”) is obtained as shown in Equation 6 below.

[0104] [Mathematical Expression 6]

[0105]

[0106] Next, the problem is discretized. The coordinates of the phase modulation plane Sm are sampled by using grid points that are equally spaced with a spacing d and have a size of M×N.

[0107] [Mathematical Expression 7]

[0108]

[0109] [Mathematical Expression 8]

[0110]

[0111] The phase distribution is also sampled at the aforementioned equally spaced grid points and represented in matrix form, as shown in equation P = {P} m，n} m，n ={P(x m，n y m，n )} m，n As indicated in the text. The differentials of this phase distribution P are defined as follows.

[0112] [Mathematical Expression 9]

[0113]

[0114] [Mathematical Expression 10]

[0115]

[0116] [Mathematical Expression 11]

[0117]

[0118] Under the above discretization rules, the optimization problem of phase distribution can be analyzed within the framework of linear algebra (Equation 4).

[0119] From here on, the (M, N) matrix itself will be treated as an abstract vector. That is, for A, B∈C M×N And α∈C, when A+B and αA are defined as the sum of the matrix and scalar product respectively, C M×N The result of the above calculations becomes a linear space.

[0120] Furthermore, the differential operators in equations 9 to 11 become C M×N Upper closed linear operator. Note that the matrix inner product is defined as follows.

[0121] [Mathematical Expression 12]

[0122]

[0123] Therefore, the norm of the matrix is ​​shown below.

[0124] [Mathematical Expression 13]

[0125]

[0126] Within the above framework, Equation 4 is rewritten as Equation 14 below.

[0127] [Mathematical Expression 14]

[0128]

[0129] Here, I p ={I^~(ux m，n uy m，n )} m，n However, we establish the following equation 15.

[0130] [Mathematical Expression 15]

[0131]

[0132] As a result, it should be noted that, in fact, I p It is a matrix with P-dependency.

[0133] Due to the above I p Given the nonlinearity of P, it is difficult to solve the optimization problem of the phase distribution (Equation 4). If I p If it is a constant matrix, then equation 14 indicates the linear equation " The least squares problem can be solved as shown in equation " The least squares solution is obtained precisely as indicated in the instructions. Here, It is relative to calculation The Moore-Penrose generalized inverse element. Specifically, in the case of this problem, the convolution operation... The relationship between the eigenvalues ​​and eigenvectors of the Discrete Fourier Transform can be used to configure the Moore-Penrose generalized inverse element for operations that can perform high-speed computations.

[0134] To solve the nonlinear optimization problem (Equation 14), the proximity algorithm performs iterative optimization by using a proximity method. More specifically, it is assumed that the temporary value P(i) relative to the phase distribution is at the grid point (ux) calculated by Equation 15. (i) uy (i) The value of the target intensity distribution sampled at point () is given by equation "I (i) p ={I^~(ux (i) m,n ,uy (i) m,n )} m,n This indicates that a further reduction in norm can be determined near P(i). The phase modulation distribution P, where I p The P-dependency can be sufficiently ignored, as it is achieved by using a constant matrix I. (i) P replaces I in equation 14 p P obtained from the linear least squares problem (i) A near-solution. Therefore, P should be obtained through this near-solution P( i+1 The update is performed. The iterative optimization steps based on the above strategy are shown in Equation 16 below.

[0135] [Mathematical Expression 16]

[0136]

[0137] Here, γ here,] new. Based on (i) || 2 It indicates P and P (i) The proximity algorithm introduces a regularization term for P. To increase the stability of the optimization, the proximity algorithm introduces a curvature with respect to P. Regularization is applied, and an update formula is defined as indicated below.

[0138] [Mathematical Expression 17]

[0139]

[0140] By using the matrix set {F} defined by the following formula (k，l) M，N ∈C M×N We can use (k = 0, 1, ..., M-1, I = 0, 1, ..., N-1) to solve equation 17.

[0141] [Mathematical Expression 18]

[0142]

[0143] Here, we establish the following formula.

[0144] [Mathematical Expression 19]

[0145]

[0146] The matrix set {F (k，l) M，N It has the following important characteristics.

[0147] (1) <F (k’，l’) M，N F (k，l) M，N >=δ k’，k δ l’，l , {F (k，l) M，N} represents CM×N The standard orthogonal basis.

[0148] (2) When a matrix A∈CM×N has {F (k，l) M，N The expansion coefficient {F} provided (k，l) M，N The result obtained when {A} is set as an (M, N) matrix is ​​called the Discrete Fourier Transform and is written as DFT[A]:={F} (k，l) M，N A} k，l Furthermore, conversely, in contrast to the discrete Fourier transform B of matrix A, the original matrix A is called the inverse discrete Fourier transform of B and is written as IDFT[B]: = Σ k，l B k，l F (k，l) M，N The Discrete Fourier Transform and the Inverse Discrete Fourier Transform can be computed quickly using an algorithm called the Fast Fourier Transform.

[0149] (3){F (k，l) M，N} is through any kernel K∈C M×N The eigenvectors of the convolution operation. And each eigenvector F (k，l) M，N The corresponding eigenvalues ​​are calculated as shown in equation λ. (k，l) M，N =(√MN)DFT[K] k，l (Note that (√MN) represents the square root of MN) as expressed in the text.

[0150] Equation 20 below can be derived from a paper on proximity algorithms (Reference 1).

[0151] [Mathematical Expression 20]

[0152]

[0153] Therefore, Equation 17 is derived from the normal basis {F} (k，l) M，N} Rewrite to represent the components as shown in Equation 21 below.

[0154] [Mathematical Expression 21]

[0155]

[0156] When using the term "lemma" as described later in this text, P (i+1) The components of the discrete Fourier transform are shown in the following equation.

[0157] [Mathematical Expression 22]

[0158]

[0159] P (i+1) It can be calculated using the discrete inverse Fourier transform described above.

[0160] The Fourier Transform FT[·] in the paper on the nearest neighbor algorithm and the Discrete Fourier Transform DFT[·] in this embodiment correspond to DFT[·] = FT[·] / √MN. Furthermore, Eigenvalue DFT[A] k，l It is a real number. Therefore, when equation 22 is rewritten as shown in reference 1, we obtain the following equation 23.

[0161] [Mathematical Expression 23]

[0162]

[0163] In reference 1, the complex conjugation symbol is appended to the molecular FT[A] of equation 23. k，l However, due to FT[A] k，l Since it is a real number as described above, the existence of the complex conjugate symbol is meaningless.

[0164] "lemma"

[0165] For a1, a2, a3, b1, b2, b3 and z∈c, f(z)=|a1z-b1| 2 +|a2z-b2| 2 +|a3z-b3| 2 It is minimized, as shown in Equation 24 below.

[0166] [Mathematical Expression 24]

[0167]

[0168] "prove"

[0169] It is divided into real and imaginary parts, such as a1 = a1 (real) +ja1 (imag) , z = x + jy.

[0170] [Mathematical Expression 25]

[0171]

[0172] When calculating from the above, it should be understood that the following formula is obtained.

[0173] [Mathematical Expression 26]

[0174]

[0175] Furthermore, the Hessian matrix is ​​positive definite for all x and y, as shown below.

[0176] [Mathematical Expression 27]

[0177]

[0178] Therefore, the minimum point of f(z) is represented by the following equation.

[0179] [Mathematical Expression 28]

[0180]

[0181] QED

[0182] It can be said that the approach algorithm described in Reference 1 is a method that prioritizes velocity calculation by sacrificing the reproducibility of the target intensity distribution.

[0183] Meanwhile, this implementation proposes a new free-form algorithm (hereinafter referred to as the "new algorithm") that reproduces the target intensity distribution as faithfully as possible.

[0184] The new algorithm aims to determine the phase distribution that minimizes the error between the actual reconstructed image and the target intensity distribution. More specifically, the phase distribution to be determined is the solution P^ of the nonlinear optimization problem of Equation 29 below.

[0185] [Mathematical Expression 29]

[0186]

[0187] In the i-th step of the optimization loop, the proximity problem determines the updated value P of the phase distribution based on the linearized model around P=0. (i+1) Without considering the temporary value P of the phase distribution at this time. (i) Meanwhile, the new algorithm first calculates the temporary value P based on the accurate ray-optics model shown in Equation 30 below. (i) Error in intensity distribution (error distribution) (i) Here, "precise ray-optic model" refers to a ray-optic model that includes nonlinear terms as shown in Equation 4 above.

[0188] [Mathematical Expression 30]

[0189]

[0190] Assuming that the error quantity (error distribution) is expressed as error (i) The result obtained by multiplying the result by the feedback gain G is used as the light intensity correction value ΔI. (i) =G·error (i)In this case, when around the temporary value P (i) When linearizing the accurate ray-optics model, we obtain Equation 31 below.

[0191] [Mathematical Expression 31]

[0192]

[0193] Therefore, in the light intensity correction value ΔI (i) When the value is sufficiently small, the phase correction value ΔP used to achieve this light intensity correction is given below. (i) As a linear term in Equation 31 "relative to ΔI" (i) Inverse computation.

[0194] [Mathematical Expression 32]

[0195]

[0196] When the previously mentioned discrete representation is applied again here, Equation 33 is derived from Equation 32.

[0197] [Mathematical Expression 33]

[0198]

[0199] Here, we should pay attention When the phase correction value ΔP(i) is at that time, the components of the discrete Fourier transform are shown in Equation 34 below.

[0200] [Mathematical Expression 34]

[0201]

[0202] The phase distribution should be obtained by using the phase correction value ΔP obtained from the above inverse discrete Fourier transform. (i) With P (i) Add them together to update.

[0203] exist Figure 6 The optimized loop provided by the new algorithm described above is shown in the figure.

[0204] exist Figure 6 In this context, according to Equation 4 above, the ray-optical model F1 originates from P. (i) Calculate I (i) .

[0205] Regarding the temporary value P of the phase distribution (i) The target intensity resampling unit F2 calculates non-equidistant grid points (ux) on the projection plane Sp. (i) uy (i)The resampled target intensity distribution value is used as the target intensity I^~(ux) (i) uy (i) ), which is calculated by Equation 15. Here, as is evident from Equations 2 and 4, I in the ray-optical model F1 (i) During the computation process, non-equidistant grid points (ux) on the projection plane Sp are calculated simultaneously. (i) uy (i) Based on the grid points (ux) calculated in the ray-optic model F1 as described above. (i) uy (i) The target intensity resampling unit F2 calculates the target intensity I^~(ux) based on the information. (i) uy (i) ).

[0206] like Figure 6 As shown, I will be calculated in the ray-optical model F1. (i) With target intensity I^~(ux) (i) uy (i) The difference between ) is calculated as the error distribution. (i) And by distributing the error... (i) Multiply by the feedback gain G to obtain the light intensity correction value ΔI (i) The linear term inverse calculation unit F3 calculates the linear term based on the light intensity correction value ΔI. (i) Perform the calculation shown in Equation 34 to calculate the phase correction value ΔP. (i) Then, the calculated phase correction value ΔP (i) With temporary value P (i) Adding them together yields a new temporary value P. (i+1) The resulting phase distribution P (i+1) It is input into the ray-optics model F1 to form a feedback loop.

[0207] Here, to ensure the phase correction value ΔP (i) The reliability requires ΔI (i) Small enough, the phase correction value ΔP (i) This originates from the inverse calculation of the linear term. Therefore, a small value should preferably be chosen as the feedback gain G. However, even when G = 0.1, error... (i) In the initial stage of the optimized loop relative to the high-contrast target image, the target intensity I^~(ux) is approached. (i) uy (i) ) itself. Therefore, ΔI derived by multiplying by the feedback gain G (i) =0.1·error (i) This is outside the effective range of the linear approximation. Therefore, the feedback gain G is not initially fixed at a constant G0. Instead, ΔI... (i)The absolute value of the maximum value and the allowable value ΔI of the minimum value max It is pre-planned. Then, in the initial stage of optimizing the loop, according to ΔI... (i) All components in -ΔI max to +ΔI max Scaling errors within a range (i) Subsequently, when error (i) When the value is reduced to a certain extent, the feedback gain G is dynamically adjusted in the manner of G = G0.

[0208] Figure 7 A schematic diagram illustrating the dynamic adjustment of gain G is shown. This is achieved by using error... (i) The light intensity correction value ΔI obtained by multiplying by the constant G0 (i) The maximum absolute value exceeds the allowable value ΔI max In the case of exceeding the allowable value ΔI, the gain G is adjusted in such a way that... max Part of ( Figure 7 (Shadow in -ΔI) max to +ΔI max Within the range.

[0209] For example, choosing G0 = 0.1 and ΔI max =0.01 (the average intensity of the incident light and target intensity distribution is 1). Then, in each step i, it should be based on error. (i) Set G as shown in Equation 35 below.

[0210] [Mathematical Expression 35]

[0211]

[0212] By using a ray-optics model that includes nonlinear terms, the new algorithm enables the accurate determination of the phase distribution used to reproduce the target light intensity distribution.

[0213] However, it is assumed that the intensity distribution of the light incident on the phase modulation plane Sm is uniform. Therefore, the intensity distribution of the incident light is superimposed on the reconstructed image with a non-uniform phase distribution derived from the new algorithm described above.

[0214] At the same time, when the incident light intensity distribution I (Incident) When known, reflection I in the ray-optical model F1 can be achieved through the nonlinear optimization loop of the new algorithm and its linear inverse computation unit F3. (Incident) To obtain the phase distribution used to correct the incident light intensity distribution and reproduce the target intensity distribution.

[0215] Figure 8An optimized loop is shown, provided by a phase distribution calculation algorithm according to an embodiment for reflecting the incident light intensity distribution as described above.

[0216] like Figure 8 As shown, the optimization loop in the above case is respectively equipped with a ray-optical model F1' and a linear term inverse calculation unit F3', replacing... Figure 6 The light-optical model F1 and the inverse linear term calculation unit F3 are shown in the diagram. Incident light intensity distribution I. (Incident) This is reflected in the ray-optic model F1' and the inverse computation unit of the linear term F3'.

[0217] When weighted I based on individual coordinate components in a numerical formula described by the new algorithm (Incident) At that time, it can reflect the intensity distribution of incident light I (Incident) .

[0218] More specifically, calculate equation 36 as shown below.

[0219] [Mathematical Expression 36]

[0220]

[0221] Correct the linear terms of the model to Therefore, the phase correction value ΔP (i) It should be calculated as the least squares solution (Equation 37) of the linear equation expressed in Equation 38.

[0222] [Mathematical Expression 37]

[0223]

[0224] [Mathematical Expression 38]

[0225]

[0226] However, for simplicity, as shown in Equation 40, the phase correction value ΔP(i) is calculated as the least squares solution of the linear equation (Equation 39).

[0227] [Mathematical Expression 39]

[0228]

[0229] [Mathematical Expression 40]

[0230]

[0231] Phase correction value ΔP (i) The components of the discrete Fourier transform are shown in Equation 41 below.

[0232] [Mathematical Expression.41]#

[0233]

[0234] However, in practice, information about the incident light intensity distribution does not always need to be additionally incorporated into the inverse calculation model, as shown in Equation 41 above. Even when the phase correction amount is determined using the earlier Equation 34, a phase distribution that cancels out the influence of the incident light intensity distribution can be obtained. However, it is necessary to incorporate information about the incident light intensity distribution into the ray-optical model that performs the forward calculation.

[0235] When information about the incident light intensity distribution is additionally incorporated into the inverse calculation model, as shown in Equation 41, stable convergence can generally be achieved. However, for example, in cases where the incident light intensity distribution has very low light intensity values, "I" in Equation 41... (i) / I (incident) "It has extremely high component values. This may lead to unstable convergence. In this case, based on the assumption that information about the intensity distribution of incident light is incorporated into the ray optical model, Equation 34 should preferably be used (i.e., information about the intensity distribution of incident light should not be incorporated into the inverse calculation model).

[0236] Here, in the illumination device 1 according to the first embodiment, the incident light intensity distribution I relative to the phase modulation surface Sm is represented. (incident) The information is stored in a memory such as a ROM in, for example, the control unit 5. Based on the stored information about the incident light intensity distribution I... (incident) The phase distribution calculation unit 5b performs phase distribution calculation processing for reproducing the target light intensity distribution on the projection plane Sp by using the calculation method described with reference to the aforementioned numerical formula.

[0237] In the first embodiment, the incident light intensity distribution I relative to the phase modulation surface Sm (incident) For example, it is pre-measured and stored in the lighting device 1 before being shipped from the factory.

[0238] <2. Second Embodiment>

[0239] The second embodiment of this technology will now be described.

[0240] The second embodiment is configured to process the change in incident light intensity distribution over time.

[0241] Figure 9 This is a diagram showing an example of the configuration of the lighting device 1A according to the second embodiment.

[0242] It should be noted that in the following description, those parts similar to those described so far are indicated by the same reference numerals as their counterparts and will not be described redundantly.

[0243] Lighting device 1A and Figure 1 The difference of the lighting device 1 shown is that the lighting device 1A additionally includes an imaging unit 6 and includes a control unit 5A instead of the control unit 5.

[0244] The imaging unit 6 includes an imaging element such as a CCD (charge-coupled device) sensor or a CMOS (complementary metal-oxide-semiconductor) sensor, and captures an image of the light-emitting surface of the light source unit 2 to obtain a captured image reflecting the light intensity distribution of light incident on the phase modulation surface Sm.

[0245] The difference between control unit 5A and control unit 5 is that control unit 5bA replaces phase distribution calculation unit 5b. Phase distribution calculation unit 5bA generates information about the incident light intensity distribution I based on the image captured by imaging unit 6. (incident) Information, and based on the generated information about the incident light intensity distribution I (incident) The information is used to calculate the phase distribution for reproducing the target light intensity distribution on the projection plane Sp by using a calculation method similar to that used by the phase distribution calculation unit 5b in the first embodiment.

[0246] As described above, the second embodiment incorporates the incident light intensity distribution I obtained from the image captured by the imaging unit 6 into both the ray-optical model F1' and the linear term inverse calculation unit F3'. (incident) .

[0247] Therefore, even in the incident light intensity distribution I (incident) Even when the intensity distribution of incident light changes over time, it can still be represented by the altered distribution. (incident) This is reflected in the ray-optical model F1' and the inverse calculation unit of the linear term F3'.

[0248] Therefore, even in the incident light intensity distribution I (入射) This also prevents the incident light intensity distribution from overlapping on the reconstructed image when it changes over time.

[0249] It should be noted that the second embodiment may also have information regarding the incident light intensity distribution I. (incident) The information is not incorporated into the configuration of the inverse calculation model. With this configuration, the phase distribution calculation unit 5bA only incorporates the incident light intensity distribution I obtained from the image captured by the imaging unit 6 into the ray optical model F1'. (入射) .

[0250] <3. Third Embodiment (Application of Projection Device)>

[0251] A third embodiment of this technology is configured such that the lighting device according to the previously described embodiments is applied to a projector device.

[0252] Figure 10 This is a diagram showing an example of a configuration of the lighting device 1 according to the first embodiment applied to the projector device 10.

[0253] like Figure 10 As shown, the projection device 10 and Figure 1 The lighting device 1 shown is similar, including a light source unit 2, a phase modulation SLM 3 and a driving unit 4, and the projection device 10 also includes an intensity modulation SLM 11, a control unit 12 and a driving unit 13.

[0254] Intensity-modulated SLM 11 includes, for example, a transmissive liquid crystal panel, and performs spatial intensity modulation on the incident light. For example... Figure 10 As shown, the intensity modulation SLM 11 is connected to the output stage of the phase modulation SLM 3. Therefore, light emitted from the light source 2 and spatially phase-modulated by the phase modulation SLM 3 is incident on the intensity modulation SLM 11.

[0255] The projection device 10 projects a reproduced image of the target image onto the projection plane Sp' by projecting light that has undergone spatial light intensity modulation by intensity modulation SLM 11 onto the projection plane Sp'.

[0256] Here, it is evident from the position of the depicted projection plane Sp that the phase distribution in this case is calculated using a free-form method to reproduce the target light intensity distribution on the intensity modulation plane of the intensity modulation SLM 11.

[0257] It should be noted that, for example, reflective spatial light phase modulators such as reflective liquid crystal panels or DMDs can be used as intensity modulation SLM 11.

[0258] The control unit 12 is similar to the control unit 5, including a microcomputer, which includes, for example, a CPU, ROM and RAM. It calculates the phase distribution of the phase modulation SLM 3 based on the target image, calculates the light intensity distribution of the intensity modulation SLM 11, drives the phase modulation SLM 3 with the drive unit 4 based on the calculated phase distribution, and drives the intensity modulation SLM 11 with the drive unit 13 based on the calculated light intensity distribution.

[0259] like Figure 10 As shown, similar to the control unit 5, the control unit 12 includes a target intensity distribution calculation unit 5a and a phase distribution calculation unit 5b. The target intensity distribution calculation unit 5a and the phase distribution calculation unit 5b calculate the phase distribution of the phase modulation SLM 3 used to reproduce the target light intensity distribution on the projection plane Sp (in this case, the intensity modulation plane of the intensity modulation SLM 11) using a calculation method similar to that used in the first embodiment.

[0260] In addition, the control unit 12 has an intensity distribution calculation unit 12a. The intensity distribution calculation unit 12a calculates the light intensity distribution set in the intensity modulation SLM 11 so as to reproduce the light intensity distribution of the target image on the projection plane Sp'.

[0261] More specifically, the intensity distribution calculation unit 12a takes the target image and the light intensity distribution I on the projection plane Sp calculated by the phase distribution calculation unit 5b as input, and calculates the light intensity distribution of the intensity modulation SLM 11 based on the input target image and the light intensity distribution I. Here, in reference... Figure 8 The light intensity distribution I on the projection plane Sp is calculated in the described ray-optic model F1'. The light intensity distribution I used here is the light intensity distribution I calculated in the ray-optic model F1' when the phase distribution P is determined as a solution of the optimal loop.

[0262] In the projection device 10, a light intensity distribution based on the phase distribution P is reproduced on the intensity modulation plane of the intensity modulation SLM 11. Therefore, in order to reproduce the light intensity distribution of the target image on the projection plane Sp', the light intensity distribution in the intensity modulation SLM 11 can be set to eliminate the difference between the light intensity distribution reproduced on the intensity modulation plane based on the phase distribution P and the light intensity distribution of the target image reproduced on the projection plane Sp'.

[0263] Therefore, the intensity distribution calculation unit 12a calculates the light intensity distribution of the target image to be reproduced on the projection plane Sp', and calculates the difference between the calculated light intensity distribution and the light intensity distribution I input from the phase distribution calculation unit 5b. Then, based on the light intensity distribution calculated as the difference, the driving unit 13 drives the intensity modulation SLM 11.

[0264] Incidentally, in the past, projector devices obtained reproduced images by allowing an intensity modulation SLM 11 to perform spatial light intensity modulation on the light from the light source. However, spatial light intensity modulation partially blocks or dims the light incident from the light source. Therefore, light utilization efficiency is low, and contrast enhancement is difficult to achieve.

[0265] Simultaneously, when the lighting device 1 used to reproduce the desired light intensity distribution through spatial light phase modulation is applied to, for example... Figure 10 When using the projector device shown, light utilization efficiency can be improved and the contrast of the reproduced image can be enhanced. Figure 10In the configuration shown, by allowing the phase modulation SLM 3 to perform spatial light phase modulation, the light intensity distribution based on the target image is reproduced on the intensity modulation plane of the intensity modulation SLM 11. This corresponds to an approximate light intensity distribution of the target image formed before spatial light intensity modulation by the intensity modulation SLM 11, and is similar to the control of what is commonly known as region segmentation driving used to provide backlight for a liquid crystal display. However, in this case, the light intensity distribution is formed by phase modulation. This prevents a reduction in the utilization efficiency of light from the light source.

[0266] In the above scenario, the intensity modulation SLM 11 arranges the details of the reproduced image, commonly referred to as the low-frequency image, reproduced by the phase modulation SLM3, and reproduces the light intensity distribution according to the target image on the projection plane Sp'. This allows for enhanced contrast of the reproduced image while suppressing a decrease in resolution.

[0267] It should be noted that the lighting device 1A according to the second embodiment can be applied to the configuration of a projector device.

[0268] Figure 11 An example configuration of a projector device 10A with an applied lighting device 1A is shown.

[0269] Projector device 10A and Figure 10 The difference between the projector device 10 shown is that it additionally includes an imaging unit 6 and includes a control unit 12A instead of the control unit 12. The difference from the control unit 12 is that the control unit 12A has a phase distribution calculation unit 5bA instead of the phase distribution calculation unit 5b.

[0270] Furthermore, the camera unit 6 and the phase distribution calculation unit 5bA have already been described in the second embodiment, so their description will be omitted here.

[0271] <4. Variations>

[0272] It should be noted that the implementation of this technology is not limited to the specific examples described above. Various modifications can be made to the configuration of the foregoing embodiments.

[0273] For example, there might be cases where the target image is a video image rather than a still image. In the case of a video image, it's conceivable that the phase distribution of the phase-modulated SLM 3 and the light intensity distribution of the intensity-modulated SLM 11 can be calculated based on each frame. Meanwhile, if the image content of the frame remains unchanged, the phase and light intensity distributions do not need to be calculated. Conversely, if the image content of the frame changes, the phase and light intensity distributions can be calculated.

[0274] Furthermore, the second embodiment has been described based on the assumption that the imaging unit 6 detects the light intensity distribution of light incident on the phase modulation surface Sm. However, the method for detecting the light intensity distribution of light incident on the phase modulation surface Sm is not limited to the method using the imaging unit 6. Alternatively, for example, a light emission sensor (a sensor for detecting the amount of light emitted from the relevant light emission element 2a) provided for each light-emitting element 2a in the light source unit 2 can be used to detect the light intensity distribution of light incident on the phase modulation surface Sm.

[0275] Furthermore, the incident light intensity distribution I has been incorporated into both the ray optical model and the inverse calculation model. (incident) The situation and reference only the incident light intensity distribution I in the ray optical model (incident) The situation has been described above. However, alternatively, based on the light intensity distribution morphology in the incident light intensity distribution, the incident light intensity distribution I can be... (incident) The choice is to integrate it into both the ray optics model and the inverse computation model, or to offer the option only within the ray optics model.

[0276] <5. Overview of Implementation Examples>

[0277] As described above, the signal processing apparatus (control unit 5, 5A, 12 or 12A) according to the embodiments of this technology performs calculations satisfying "Condition 1" as calculation processing for reproducing the target light intensity distribution on the projection plane by spatial light phase modulation of the incident light. "Condition 1" specifies that the calculation processing includes a nonlinear ray-optical model (ray-optical model F1'), i.e., a ray-optical model including nonlinear terms; and an inverse calculation model (linear inverse calculation unit F3') relating to the model obtained by linearizing the nonlinear ray-optical model; determining an error distribution (error distribution error) of the error between the target light intensity distribution and the light intensity distribution calculated by the nonlinear ray-optical model based on the temporary value of the phase distribution; obtaining a light intensity correction value (ΔI) by multiplying the error distribution by the feedback gain (feedback gain G); inputting the light intensity correction value into the inverse calculation model to obtain an output; treating the obtained output as a phase correction value (phase correction value ΔP); and using a feedback loop to repeatedly update the phase distribution by adding the phase correction value to the temporary value.

[0278] Using models such as the aforementioned ray-optics model that includes nonlinear terms makes it possible to accurately determine the phase distribution used to reproduce the target light intensity distribution.

[0279] Therefore, the reproducibility of the regenerated image can be improved relative to the target light intensity distribution.

[0280] Furthermore, the signal processing apparatus according to embodiments of this technology performs the calculation processing of the phase distribution in a manner that satisfies "Condition 1" and "Condition 2". "Condition 2" specifies the term of incorporating the light intensity distribution of the incident light into the nonlinear ray-optical model.

[0281] This ensures that by performing phase distribution calculations using the feedback loop specified by "Condition 1", the phase distribution can be determined in a way that eliminates the incident light intensity distribution and reproduces the target light intensity distribution.

[0282] Therefore, it can prevent the intensity distribution of incident light from being superimposed on the reconstructed image.

[0283] Furthermore, the signal processing apparatus according to the embodiments of this technology performs the calculation processing of the phase distribution in a manner that satisfies "Condition 1" and "Condition 3". "Condition 3" specifies that the term of the light intensity distribution of the incident light is incorporated into both the nonlinear ray-optical model and the inverse calculation model.

[0284] Since the incident light intensity distribution term is incorporated not only into the ray-optics model but also into the inverse calculation model, convergence stabilization can be achieved in the phase distribution calculation performed using the feedback loop specified in "Condition 1". More specifically, for example, as described above, convergence stabilization can be achieved when the incident light intensity distribution does not have extremely low light intensity values ​​in some parts.

[0285] Furthermore, the signal processing device according to embodiments of this technology controls the feedback gain based on the absolute value of the error distribution (see [link]). Figure 7 And Equation 35).

[0286] When using the feedback loop specified in "Condition 1", the light intensity correction value input to the inverse calculation model needs to be small enough to ensure the reliability of the phase correction value calculated by the inverse calculation model. As mentioned above, by controlling the feedback gain according to the absolute value of the error distribution, it is possible to prevent the light intensity correction value input to the inverse calculation model from being too large.

[0287] Therefore, it is possible to appropriately determine the phase distribution used to reproduce the target light intensity distribution.

[0288] Furthermore, when the maximum value of the absolute value of the light intensity correction obtained by multiplying the error distribution by the feedback gain based on a constant (G0) exceeds a predetermined value (ΔI) max In the case of [missing information], the signal processing apparatus according to the embodiments of the present technology controls the feedback gain to reduce the maximum value of the absolute value of the light intensity correction value to a value not greater than a predetermined value. Meanwhile, if the maximum value of the absolute value of the light intensity correction value obtained by multiplying the error distribution by the constant-based feedback gain does not exceed the predetermined value, the signal processing apparatus uses a constant as the feedback gain.

[0289] This allows for the repeated correction of temporary values ​​of the phase distribution by multiplying the error distribution by a small light intensity correction value obtained through progressively adjusted feedback gain, even under conditions of large errors. Then, when the error distribution is equal to or less than a predetermined value through repeated correction, increased convergence is provided by changing the feedback gain to a constant.

[0290] Therefore, it is possible to appropriately determine the phase distribution used to reproduce the target light intensity distribution.

[0291] Furthermore, the signal processing apparatus employs a signal processing method according to an embodiment of the present technology, which performs a computational processing that satisfies "Condition 1" as a calculation of the phase distribution of the target light intensity distribution reproduced on the projection plane by spatial light phase modulation of the incident light. "Condition 1" specifies that the computational processing includes a nonlinear ray-optical model including a nonlinear term and an inverse computational model of the model obtained by linearizing the nonlinear ray-optical model, determining the error distribution between the target light intensity distribution and the light intensity distribution calculated by the nonlinear ray-optical model based on a temporary value of the phase distribution, obtaining a light intensity correction value by multiplying the error distribution by the feedback gain, inputting the light intensity correction value into the inverse computational model to obtain an output, treating the obtained output as a phase correction value, and using a feedback loop that repeatedly updates the phase distribution by adding the phase correction value to the temporary value.

[0292] The above-described signal processing method also provides similar operation and advantages as the signal processing apparatus described above according to embodiments of the present technology.

[0293] Furthermore, the program according to the embodiments of this technology is a computer-readable program, and the program is adapted to cause the computer device to perform a computational process that satisfies "Condition 1" as a computational process for reproducing the phase distribution of the target light intensity distribution on the projection plane by spatial light phase modulation of the incident light. "Condition 1" specifies that the computational process includes a nonlinear ray-optical model including a nonlinear term and an inverse computational model of the model obtained by linearizing the nonlinear ray-optical model, determining the error distribution between the target light intensity distribution and the light intensity distribution calculated by the nonlinear ray-optical model based on the temporary value of the phase distribution, obtaining a light intensity correction value by multiplying the error distribution by the feedback gain, inputting the light intensity correction value into the inverse computational model to obtain an output, treating the obtained output as a phase correction value, and using a feedback loop to repeatedly update the phase distribution by adding the phase correction value to the temporary value.

[0294] More specifically, for example, the program according to the embodiments of the present technology is a program that causes a computer device such as control unit 5 (or 5A) or control unit 12 (or 12A) to perform the processing of phase distribution calculation unit 5b or 5bA.

[0295] The above procedure enables the implementation of the signal processing apparatus according to the embodiments of the present technology described above.

[0296] Furthermore, the lighting device (lighting device 1 or 1A or projector device 10 or 10A) according to embodiments of this technology includes a light source unit (light source unit 2), a phase modulation unit (phase modulation SLM3), and a signal processing unit (control unit 5, 5A, 12, or 12A). The light source unit has a light-emitting element (light-emitting element 2a). The phase modulation unit performs spatial light phase modulation on the incident light from the light source unit. The signal processing unit performs calculation processing that satisfies "condition 1" as a calculation processing for reproducing the phase distribution of the target light intensity distribution on the projection plane through spatial light phase modulation. Condition 1 specifies that the calculation process includes a nonlinear ray-optical model and an inverse calculation model of the model obtained by linearizing the nonlinear ray-optical model. The error distribution between the target light intensity distribution and the light intensity distribution calculated by the nonlinear ray-optical model based on the temporary value of the phase distribution is determined. The light intensity correction value is obtained by multiplying the error distribution by the feedback gain. The light intensity correction value is input into the inverse calculation model to obtain the output. The obtained output is regarded as the phase correction value. The feedback loop is used to repeatedly update the phase distribution by adding the phase correction value to the temporary value.

[0297] The lighting device described above according to embodiments of the present technology also provides similar operation and advantages as the signal processing device described above according to embodiments of the present technology.

[0298] Furthermore, the lighting device according to the embodiments of this technology is configured such that the light source section has multiple light-emitting elements.

[0299] This eliminates the need to use a single high-output light-emitting element in the light source section to meet the predetermined light intensity requirements.

[0300] Therefore, the cost of the light source can be reduced.

[0301] The illumination device (illumination device 1A or projector device 10A) according to an embodiment of the present technology is configured such that the signal processing unit performs calculation processing to calculate the phase distribution satisfying "condition 1" and "condition 2". "Condition 2" specifies the term for incorporating the light intensity distribution of the incident light into the nonlinear ray-optical model. The illumination device includes an intensity distribution detection unit (imaging unit 6) for detecting the light intensity distribution of the incident light. The signal processing unit (control unit 5A or 12A) uses the light intensity distribution detected by the intensity distribution detection unit as the light intensity distribution to be incorporated into the nonlinear ray optical model.

[0302] This ensures that, given a change in incident light intensity distribution over time, the nonlinear ray-optics model can reflect the incident light intensity distribution that occurs over time.

[0303] Therefore, even when the incident light intensity distribution changes over time, it can prevent the incident light intensity distribution from superimposing on the reconstructed image.

[0304] It should be noted that the advantages described in this document are illustrative rather than limiting. This technology may provide additional advantages beyond those described in this document.

[0305] <6. This technology>

[0306] It should be noted that this technology can also be implemented using the following configuration. (1)

[0308] A signal processing apparatus performs calculation processing of the phase distribution for reproducing a target light intensity distribution on a projection plane by spatial light phase modulation of incident light.

[0309] The calculation process is performed in a manner that satisfies "Condition 1", and...

[0310] Condition 1 specifies that the computational process includes a nonlinear ray-optical model that includes nonlinear terms and an inverse computational model of the model obtained by linearizing the nonlinear ray-optical model. The error distribution is determined to be the error between the target light intensity distribution and the light intensity distribution calculated by the nonlinear ray-optical model based on the temporary value of the phase distribution. The light intensity correction value is obtained by multiplying the error distribution by the feedback gain. The light intensity correction value is input into the inverse computational model to obtain the output. The obtained output is regarded as the phase correction value. The feedback loop is used to repeatedly update the phase distribution by adding the phase correction value to the temporary value. (2)

[0312] According to the signal processing device described in (1),

[0313] The signal processing device performs the calculation of the phase distribution in a manner that satisfies "Condition 1" and "Condition 2" above, and

[0314] Condition 2 specifies the inclusion of a term in the nonlinear ray-optics model that describes the distribution of the light intensity of the incident light. (3)

[0316] According to the signal processing device described in (1),

[0317] The signal processing device performs the calculation of the phase distribution in a manner that satisfies conditions 1 and 3 above, and

[0318] Condition 3 specifies that the term for the intensity distribution of incident light should be incorporated into both the nonlinear ray-optical model and the inverse computation model. (4)

[0320] The signal processing apparatus according to any one of (1) to (3),

[0321] The signal processing device controls the feedback gain based on the absolute value of the error distribution. (5)

[0323] According to the signal processing device described in (4),

[0324] Wherein, if the maximum value of the absolute value of the light intensity correction value obtained by multiplying the error distribution by the constant-based feedback gain exceeds a predetermined value, the feedback gain is controlled to reduce the maximum value of the absolute value of the light intensity correction value to a value not greater than the predetermined value, and

[0325] The constant is used as the feedback gain if the maximum value of the absolute value of the light intensity correction value obtained by multiplying the error distribution by the feedback gain based on the constant does not exceed the predetermined value. (6)

[0327] A signal processing method is employed by a signal processing device configured to perform computational processing for reproducing the phase distribution of a target light intensity distribution on a projection plane by spatial light phase modulation of the incident light.

[0328] The calculation process is performed in a manner that satisfies "Condition 1", and...

[0329] Condition 1 specifies that the computational process includes a nonlinear ray-optical model that includes nonlinear terms and an inverse computational model of the model obtained by linearizing the nonlinear ray-optical model. The error distribution is determined to be the error between the target light intensity distribution and the light intensity distribution calculated by the nonlinear ray-optical model based on the temporary value of the phase distribution. The light intensity correction value is obtained by multiplying the error distribution by the feedback gain. The light intensity correction value is input into the inverse computational model to obtain the output. The obtained output is regarded as the phase correction value. The feedback loop is used to repeatedly update the phase distribution by adding the phase correction value to the temporary value. (7)

[0331] A program, readable by a computer device and adapted to cause the computer device to perform calculations of the phase distribution for reproducing a target light intensity distribution on a projection plane by spatial light phase modulation of incident light.

[0332] The calculation process is performed in a manner that satisfies "Condition 1", and

[0333] Condition 1 specifies that the computational process includes a nonlinear ray-optical model that includes a nonlinear term and an inverse computational model of the model obtained by linearizing the nonlinear ray-optical model. The error distribution is determined to be the error between the target light intensity distribution and the light intensity distribution calculated by the nonlinear ray-optical model based on the temporary value of the phase distribution. The light intensity correction value is obtained by multiplying the error distribution by the feedback gain. The light intensity correction value is input into the inverse computational model to obtain the output. The obtained output is regarded as the phase correction value. A feedback loop is used to repeatedly update the phase distribution by adding the phase correction value to the temporary value. (8)

[0335] A lighting device, comprising:

[0336] The light source section has a light-emitting element;

[0337] The phase modulation unit performs spatial light phase modulation on the incident light from the light source unit; and

[0338] The signal processing unit performs calculations to reproduce the phase distribution of the target light intensity distribution on the projection plane through spatial light phase modulation.

[0339] The calculation process is performed in a manner that satisfies "Condition 1", and

[0340] Condition 1 specifies that the computational process includes a nonlinear ray-optical model that includes a nonlinear term and an inverse computational model of the model obtained by linearizing the nonlinear ray-optical model. The error distribution is determined to be the error between the target light intensity distribution and the light intensity distribution calculated by the nonlinear ray-optical model based on the temporary value of the phase distribution. The light intensity correction value is obtained by multiplying the error distribution by the feedback gain. The light intensity correction value is input into the inverse computational model to obtain the output. The obtained output is regarded as the phase correction value. A feedback loop is used to repeatedly update the phase distribution by adding the phase correction value to the temporary value. (9)

[0342] According to the lighting device described in (8),

[0343] The light source section has multiple light-emitting elements. (10)

[0345] According to the lighting device described in (8) or (9),

[0346] The signal processing unit performs the calculation of the phase distribution in a manner that satisfies the above-mentioned "condition 1" and "condition 2".

[0347] Condition 2 specifies that a term for the intensity distribution of incident light be incorporated into the nonlinear ray-optics model.

[0348] The signal processing unit includes an intensity distribution detection unit for detecting the light intensity distribution of the incident light, and

[0349] The signal processing unit uses the light intensity distribution detected by the intensity distribution detection unit as the light intensity distribution incorporated into the nonlinear ray optical model.

[0350] [Symbol Explanation]

[0351] 1. 1A: Lighting device

[0352] 2: Light source section

[0353] 2a: Light-emitting element

[0354] 3: Phase Modulation SLM

[0355] 4: Drive Unit

[0356] 5. 5A: Control Department

[0357] 5a: Target Intensity Distribution Calculation Department

[0358] 5b, 5bA: Phase distribution calculation unit

[0359] 6: Imaging section

[0360] Sp, Sp': Projection plane

[0361] Sm: Phase modulation plane

[0362] F1, F1': Ray-optical model

[0363] F2: Target Intensity Resampling Unit

[0364] F3, F3': Inverse computation unit of linear terms

[0365] 10, 10A: Projector device

[0366] 11: Intensity Modulation SLM

[0367] 12, 12A: Control Unit

[0368] 12a: Intensity Distribution Calculation Section

[0369] 13: Drive unit.

Claims

1. A signal processing apparatus that performs a calculation process for reproducing a phase distribution of a target light intensity distribution on a projection plane by spatial light phase modulation of incident light, wherein the calculation process is performed in a manner satisfying a "condition 1", and the "condition 1" provides that the calculation process includes a nonlinear ray-optical model that is a ray-optical model including a nonlinear term and an inverse calculation model with respect to a model obtained by linearizing the nonlinear ray-optical model, an error distribution that determines an error between the target light intensity distribution and a light intensity distribution calculated from a provisional value of the phase distribution by the nonlinear ray-optical model, a light intensity correction value obtained by multiplying the error distribution by a feedback gain, the light intensity correction value is input to the inverse calculation model to obtain an output, the obtained output is regarded as a phase correction value, and a feedback loop that repeatedly updates the phase distribution by adding the phase correction value to the provisional value is used.

2. The signal processing apparatus according to claim 1, wherein the signal processing apparatus performs a calculation process that calculates the phase distribution in a manner satisfying the "condition 1" and a "condition 2", and the "condition 2" provides that a term that incorporates a light intensity distribution of the incident light is included in the nonlinear ray-optical model.

3. The signal processing apparatus according to claim 1, wherein the signal processing apparatus performs a calculation process that calculates the phase distribution in a manner satisfying the "condition 1" and a "condition 3", and the "condition 3" provides that a term that incorporates a light intensity distribution of the incident light is included in both the nonlinear ray-optical model and the inverse calculation model.

4. The signal processing apparatus according to claim 1, wherein, the signal processing apparatus controls the feedback gain in accordance with an absolute value of the error distribution.

5. The signal processing apparatus according to claim 4, wherein in a case where a maximum value of an absolute value of the light intensity correction value obtained by multiplying the error distribution by the feedback gain based on a constant exceeds a predetermined value, the feedback gain is controlled to reduce the maximum value of the absolute value of the light intensity correction value to a value that is not greater than the predetermined value, and in a case where the maximum value of the absolute value of the light intensity correction value obtained by multiplying the error distribution by the feedback gain based on the constant does not exceed the predetermined value, the constant is used as the feedback gain.

6. A signal processing method employed by a signal processing apparatus configured to perform a calculation process for reproducing a phase distribution of a target light intensity distribution on a projection plane by spatial light phase modulation of incident light, wherein the calculation process is performed in a manner satisfying a "condition 1", and the "condition 1" provides that the calculation process includes a nonlinear ray-optical model that is a ray-optical model including a nonlinear term and an inverse calculation model with respect to a model obtained by linearizing the nonlinear ray-optical model, an error distribution that determines an error between the target light intensity distribution and a light intensity distribution calculated from a provisional value of the phase distribution by the nonlinear ray-optical model, a light intensity correction value obtained by multiplying the error distribution by a feedback gain, the light intensity correction value is input to the inverse calculation model to obtain an output, the obtained output is regarded as a phase correction value, and a feedback loop that repeatedly updates the phase distribution by adding the phase correction value to the provisional value is used. The "condition 1" provides that the calculation processing includes a nonlinear ray-optical model as a ray-optical model including a nonlinear term and an inverse calculation model with respect to a model obtained by linearizing the nonlinear ray-optical model, an error distribution of an error between the target light intensity distribution and a light intensity distribution calculated by the nonlinear ray-optical model from a provisional value of the phase distribution is determined, a light intensity correction value is obtained by multiplying the error distribution by a feedback gain, the light intensity correction value is input to the inverse calculation model to obtain an output, the obtained output is regarded as a phase correction value, and a feedback loop in which the phase distribution is repeatedly updated by adding the phase correction value to the provisional value is used.

7. A computer-readable storage medium storing a program readable by a computer device and adapted to cause the computer device to execute a calculation processing for reproducing a phase distribution of a target light intensity distribution on a projection plane by spatial light phase modulation of incident light, wherein the calculation processing is executed in a manner satisfying "condition 1", and the "condition 1" provides that the calculation processing includes a nonlinear ray-optical model as a ray-optical model including a nonlinear term and an inverse calculation model with respect to a model obtained by linearizing the nonlinear ray-optical model, an error distribution of an error between the target light intensity distribution and a light intensity distribution calculated by the nonlinear ray-optical model from a provisional value of the phase distribution is determined, a light intensity correction value is obtained by multiplying the error distribution by a feedback gain, the light intensity correction value is input to the inverse calculation model to obtain an output, the obtained output is regarded as a phase correction value, and a feedback loop in which the phase distribution is repeatedly updated by adding the phase correction value to the provisional value is used.

8. An illumination device comprising: a light source section having a light emitting element; a phase modulation section that performs spatial light phase modulation on incident light from the light source section; and a signal processing section that executes a calculation processing for reproducing a phase distribution of a target light intensity distribution on a projection plane by the spatial light phase modulation, wherein the calculation processing is executed in a manner satisfying "condition 1", and the "condition 1" provides that the calculation processing includes a nonlinear ray-optical model as a ray-optical model including a nonlinear term and an inverse calculation model with respect to a model obtained by linearizing the nonlinear ray-optical model, an error distribution of an error between the target light intensity distribution and a light intensity distribution calculated by the nonlinear ray-optical model from a provisional value of the phase distribution is determined, a light intensity correction value is obtained by multiplying the error distribution by a feedback gain, the light intensity correction value is input to the inverse calculation model to obtain an output, the obtained output is regarded as a phase correction value, and a feedback loop in which the phase distribution is repeatedly updated by adding the phase correction value to the provisional value is used.

9. The illumination device according to claim 8, wherein the light source section has a plurality of light emitting elements.

10. The illumination device according to claim 8, wherein, The signal processing section performs a calculation process of calculating the phase distribution in a manner satisfying the "condition 1" and the "condition 2"; The "condition 2" specifies a term incorporating the light intensity distribution of the incident light in the nonlinear ray-optical model, The signal processing section includes an intensity distribution detection section configured to detect the light intensity distribution of the incident light, and The signal processing section uses the light intensity distribution detected by the intensity distribution detection section as the light intensity distribution incorporated in the nonlinear ray-optical model.

Images