A soft light control method and a mobile terminal

Abstract

The application relates to the technical field of photographic light source control, and discloses a soft light control method and a mobile terminal, wherein the soft light control method comprises the following steps: constructing a four-dimensional light field harmonic model, decomposing a target light effect into a time-space harmonic combination; decomposing the target light effect into a time-space mapping sequence to form light field microstructure control parameters; combining a micro-electro-mechanical system programmable reflection array and a liquid crystal spatial light modulator, cooperating with a multi-channel LED array to realize sub-millimeter light distribution control and microsecond light field dynamic reconstruction; establishing an accurate synchronization mechanism of a shooting device shutter and light effect change; providing a light effect animation programming interface based on key frames, supporting the definition of a dynamic change sequence of light trajectory, intensity and form; and through the fusion of microstructure light field control and time-space harmonic analysis, four-dimensional dynamic light sculpture is realized, and the technical problem that a traditional soft light control method can only adjust overall brightness and color temperature is solved.

Description

Technical Field

[0001] This invention relates to the field of photographic light source control technology, and more specifically, to a soft light control method and a mobile terminal. Background Technology

[0002] In high-end fashion photography, artistic image creation, and medical imaging, photographers and artists pursue unique lighting effects, requiring the creation of complex four-dimensional lighting effects that evolve over time during the shooting process. These professional scenarios demand highly refined and dynamic control of light to create unique artistic expressions or meet specific technical requirements.

[0003] Currently, traditional softlight control technologies mainly employ mechanical adjustment, filter replacement, and electronic dimming, which can only adjust overall brightness and color temperature, failing to achieve precise control of the light field's microstructure. These technologies have significant limitations in spatial resolution and temporal response speed, making it difficult to meet the demands of high-end creative photography. Furthermore, existing dynamic lighting control systems typically use motor-controlled mechanical dimming systems, which struggle to achieve high-frequency, complex, and synchronous spatiotemporal light field changes, thus limiting the dynamic expression of lighting effects.

[0004] Furthermore, existing shooting equipment lacks a precise synchronization mechanism with changes in lighting effects, often leading to inconsistencies in timing during continuous shooting, which affects the repeatability and accuracy of the shooting results. This becomes a key technical obstacle limiting creative expression in professional shooting scenarios that require precise control of light trajectories, intensity, and shapes.

[0005] Therefore, there is an urgent need to provide a method for controlling softlights that can achieve precise and controllable control in both time and space dimensions, in order to meet the high requirements of professional photography and artistic creation for dynamic lighting effects. Summary of the Invention

[0006] This invention provides a soft light control method and mobile terminal, solving the technical problems of precise control of light field microstructure and dynamic four-dimensional light effect expression in related technologies.

[0007] This invention provides a method for controlling a soft light, comprising:

[0008] A four-dimensional light field harmonic model is constructed based on spatiotemporal Fourier analysis, decomposing the target light effect into a spatiotemporal harmonic combination to achieve a precise mathematical expression of the light field.

[0009] By utilizing the output of the four-dimensional light field harmonic model, the target light effect is decomposed into a spatiotemporal mapping sequence through the light field layer synthesis algorithm, forming physically realizable light field microstructure control parameters.

[0010] Based on the control parameters of the light field microstructure, a combination of a microelectromechanical system programmable reflective array and a liquid crystal spatial light modulator is used, along with a multi-channel LED array, to achieve sub-millimeter-level light distribution control and microsecond-level dynamic reconstruction of the light field;

[0011] Based on the results of dynamic reconstruction of the light field, a precise synchronization mechanism between the shutter speed of the shooting device and changes in light effect is established through a predictive control algorithm accelerated by optical computing.

[0012] Based on a precise synchronization mechanism, it provides a keyframe-based programming interface for light effect animation, supporting the definition of dynamic change sequences of light trajectories, intensity, and shape, and realizing these changes as physical lighting effects.

[0013] Furthermore, the steps for constructing a four-dimensional optical field harmonic model based on spatiotemporal Fourier analysis include:

[0014] Collect the target light effect characteristics to form a light field sampling dataset;

[0015] Perform a four-dimensional Fourier transform on the collected light field data to convert the spatiotemporal domain data into a frequency domain representation;

[0016] Based on the frequency domain energy distribution, the main harmonic components whose energy proportion exceeds a preset threshold are selected to construct a simplified four-dimensional optical field harmonic model.

[0017] The parameters of the harmonic model are optimized using the least squares method to minimize the error between the model output and the target light effect.

[0018] Furthermore, the step of decomposing the target ray effect into a spatiotemporal mapping sequence using the light field layering synthesis algorithm includes:

[0019] The four-dimensional light field model is projected onto a two-dimensional plane and the time dimension. The complex spatiotemporal changing light field is decomposed into a combination of spatial basis functions and time modulation functions through the singular value decomposition method.

[0020] Sparsity optimization is performed on the spatial basis function set obtained by decomposition to reduce redundant basis functions and improve computational efficiency;

[0021] The time modulation function is smoothed to eliminate high-frequency noise and ensure the continuity and smoothness of light effect changes;

[0022] The optimized spatial basis function and temporal modulation function are mapped to the control parameters of the physical optical modulator.

[0023] Furthermore, the step of combining a microelectromechanical system programmable reflective array with a liquid crystal spatial light modulator includes:

[0024] A light source array consisting of multiple independently controlled LED units is configured, each of which can independently adjust its brightness, color temperature, and on / off state to form an initial light source layer;

[0025] A reflective array consisting of thousands of miniature adjustable mirrors is constructed based on MEMS technology. Each mirror can independently adjust its angle within microseconds to control the direction of light reflection.

[0026] A liquid crystal spatial light modulator is introduced as the second-stage modulation unit. By adjusting the transmittance of each pixel unit, the light field intensity can be precisely controlled.

[0027] By integrating the three core components mentioned above into a multi-level optical system, the optical field can be accurately reconstructed through optical path configuration.

[0028] Furthermore, the step of establishing a precise synchronization mechanism between the shutter speed of the shooting device and changes in light effect includes:

[0029] Build a hardware interface for communication with various professional shooting equipment to obtain shutter signals and timecode information;

[0030] Based on the status signals of the shooting equipment and the preset light effect animation sequence, calculate the optimal light effect timing control parameters;

[0031] A predictive control algorithm based on optical computing acceleration is adopted to predict the precise time of the next frame capture based on historical data, thereby triggering changes in light effects in advance.

[0032] It enables real-time monitoring of changes in light effect and actual synchronization with camera exposure, dynamically adjusts timing parameters, and achieves closed-loop control.

[0033] Furthermore, the predictive control algorithm is implemented using a Kalman filter, which combines the physical motion model of the shooting device and real-time observation data to predict the next trigger time of the camera.

[0034] Furthermore, the step of providing a keyframe-based lighting effect animation programming interface includes:

[0035] Build a user-friendly graphical interface that allows users to intuitively set keyframes for lighting effect sequences;

[0036] Based on user-defined keyframes, an advanced interpolation algorithm is applied to generate a complete sequence of light effect animations;

[0037] A library of preset lighting effect templates has been built, which includes commonly used lighting and animation effects that users can directly call or modify based on the templates.

[0038] It enables real-time preview of light effect animations, allowing users to instantly view and adjust the animation effects.

[0039] Furthermore, the expression for the four-dimensional optical field harmonic model is:

[0040]

[0041] in This represents the reconstructed light field distribution function. The amplitude coefficient represents the intensity of each harmonic component; These represent the spatial and temporal frequencies in the x, y, and z directions, respectively, with i1, j1, k1, and l1 representing the indices of the spatial and temporal frequencies in the x, y, and z directions, respectively. is the phase offset, representing the initial phase of each harmonic component; sin is the sine function; ∑ represents the summation symbol.

[0042] Furthermore, the expression for decomposing the complex spatiotemporally varying light field into a combination of spatial basis functions and temporal modulation functions is as follows:

[0043]

[0044] Where L proj (x, y, t) is the light field function projected onto the two-dimensional plane, representing the light field distribution on the plane (x, y) that varies with time t; These are spatial basis functions, representing different spatial distribution patterns; This is a time modulation function, representing the variation of each spatial pattern over time; The weight coefficients represent the importance of the i2th basis function; n basis ∑ represents the total number of basis functions, indicating the number of principal components retained after decomposition; i2 represents the index of different basis functions; ∑ represents the summation symbol.

[0045] The present invention provides a mobile terminal, including a memory and one or more processors, wherein the memory stores executable code, and the one or more processors execute the executable code to implement the above-described soft light control method.

[0046] The beneficial effects of this invention are: by integrating microstructure light field control and spatiotemporal harmonic analysis to achieve four-dimensional dynamic light sculpture, it solves the technical problem that traditional soft light control methods can only adjust the overall brightness and color temperature;

[0047] Breaking through the limitation of traditional soft lights that can only be adjusted as a whole, it achieves precise programming control of the light microstructure, with light distribution accuracy reaching the sub-millimeter level and uniformity error reduced;

[0048] The control of soft light is extended from static three-dimensional to dynamic four-dimensional, creating a brand-new expression of light effect in the time dimension. Through high-speed digital micromirror devices and acousto-optic modulators, microsecond-level dynamic reconstruction of light field is achieved, far exceeding the human eye's visual persistence threshold.

[0049] It achieves precise synchronization between the shutter speed of the shooting device and changes in light effects, reduces synchronization errors, ensures that the preset precise light effects are captured in continuous shooting, and improves the accuracy and repeatability of creative photography;

[0050] It provides creative photographers and artists with an intuitive four-dimensional lighting programming tool, encapsulating complex light field technology within a user-friendly interface, lowering the creative threshold, and improving light energy utilization efficiency through precise control of light direction and intensity, achieving the same lighting effect with lower energy consumption than traditional softlights. Attached Figure Description

[0051] Figure 1 This is a flowchart of a soft light control method according to the present invention;

[0052] Figure 2 This is a flowchart of step 1 in this invention;

[0053] Figure 3 This is a flowchart of step 2 in this invention;

[0054] Figure 4 This is a flowchart of step 3 in this invention;

[0055] Figure 5 This is a flowchart of step 4 in this invention;

[0056] Figure 6 This is a flowchart of step 5 in this invention. Detailed Implementation

[0057] The subject matter described herein will now be discussed with reference to exemplary embodiments. It should be understood that these embodiments are discussed only to enable those skilled in the art to better understand and implement the subject matter described herein, and changes may be made to the function and arrangement of the elements discussed without departing from the scope of this specification. Various processes or components may be omitted, substituted, or added as needed in the examples. Furthermore, some features described in the examples may be combined in other examples.

[0058] At least one embodiment of the present invention discloses a soft light control method, such as... Figures 1 to 6 As shown, it includes the following steps:

[0059] Step 1: Construct a four-dimensional light field harmonic model based on spatiotemporal Fourier analysis, decompose the target light effect into a spatiotemporal harmonic combination, and realize the accurate mathematical expression of the light field.

[0060] This step utilizes spatiotemporal Fourier analysis to construct a four-dimensional light field harmonic model, decomposing the target light effect into a spatiotemporal harmonic combination to achieve a precise mathematical expression of the light field. The specific execution is as follows:

[0061] Step 1.1: Collect target light effect characteristics;

[0062] Spatiotemporal light intensity distribution data of the target light effect scene are collected by a high dynamic range imaging device to form a light field sampling dataset.

[0063] Step 1.2: Perform a four-dimensional Fourier transform;

[0064] A four-dimensional Fourier transform is performed on the acquired light field data to convert the spatiotemporal domain data into a frequency domain representation. The formula for the four-dimensional Fourier transform is:

[0065]

[0066] Where F represents the optical field in the frequency domain. Let ω represent the light field distribution function in the spatiotemporal domain, where (x, y, z) are spatial coordinates, t is the time variable, and ω is the time variable. x ω y ω z ω t Let represent the spatial frequency in the X, y, and z directions and the frequency in time t, respectively; e is the base of the natural logarithm; d represents the differential symbol; ∫ represents the integral symbol; and j represents the imaginary unit.

[0067] Optionally, to improve computational efficiency, some implementations may employ the separation of variables method for Fourier transform calculations. This involves first performing a three-dimensional Fourier transform on the spatial dimension, and then a one-dimensional Fourier transform on the temporal dimension, thereby reducing computational complexity. In other implementations, transformation calculations can be selectively performed on the spatial region and time interval of interest, based on the accuracy requirements of the application, further optimizing the utilization of computational resources.

[0068] Step 1.3: Extract the main harmonic components;

[0069] Based on the frequency domain energy distribution, the main harmonic components whose energy proportion exceeds a preset threshold are selected to construct a simplified four-dimensional optical field harmonic model:

[0070]

[0071] in This represents the reconstructed light field distribution function. The amplitude coefficient represents the intensity of each harmonic component; These represent the spatial and temporal frequencies in the x, y, and z directions, respectively, with i1, j1, k1, and l1 representing the indices of the spatial and temporal frequencies in the x, y, and z directions, respectively. is the phase offset, representing the initial phase of each harmonic component; sin is the sine function; ∑ represents the summation symbol.

[0072] The system uses an optimization algorithm based on Fast Fourier Transform (FFT) to calculate the frequency domain representation and selects the main harmonic components through energy contribution analysis. For example, when shooting the flowing light and shadow effects of silk fabric in a fashion photography scene, the system identifies the main frequency components by analyzing the target light effect, extracts key frequencies in three spatial dimensions and one time dimension, and generates a time-varying light effect model that conforms to the characteristics of silk material.

[0073] Step 1.4: Perform model optimization;

[0074] The parameters of the harmonic model are optimized using the least squares method to minimize the error between the model output and the target luminous efficacy.

[0075]

[0076] Where L target Target light effect, representing the desired light field distribution; L model The light effect predicted by the model represents the light field distribution generated by the harmonic model; Represents the coordinates of the spatial sampling point. The coordinates of the time sampling points are represented by p1, q1, r1, and s1, which represent the sampling point indices in the spatial coordinates x, y, z directions and the time dimension t, respectively. Indicates the amplitude coefficient and phase shift Optimize to minimize the objective function; (·) 2 ∑ represents the square norm; ∑ represents the summation symbol.

[0077] This step extends the traditional static three-dimensional light field analysis to a four-dimensional analysis that includes the time dimension. It achieves a precise mathematical expression of complex dynamic light effects through spatiotemporal harmonic decomposition, laying a theoretical foundation for subsequent control of light microstructures.

[0078] Step 2: Using the output of the four-dimensional light field harmonic model, the target light effect is decomposed into a spatiotemporal mapping sequence through the light field layer synthesis algorithm to form physically realizable light field microstructure control parameters;

[0079] This step utilizes a layered optical field synthesis algorithm to convert the four-dimensional optical field harmonic model obtained in step 1 into a physically realizable spatiotemporal mapping sequence, enabling precise control over the optical field microstructure. The specific execution is as follows:

[0080] Step 2.1, Light field decomposition;

[0081] The four-dimensional light field model is projected onto a two-dimensional plane and the time dimension. The complex spatiotemporally varying light field is decomposed into a combination of spatial basis functions and temporal modulation functions using the Singular Value Decomposition (SVD) method.

[0082]

[0083] Where L proj (x,y,t) is the light field function projected onto the two-dimensional plane, representing the light field distribution on the plane (x,y) that varies with time t; These are spatial basis functions, representing different spatial distribution patterns; This is a time modulation function, representing the variation of each spatial pattern over time; The weight coefficients represent the importance of the i2th basis function; n basis ∑ represents the total number of basis functions, indicating the number of principal components retained after decomposition; i2 represents the index of different basis functions; ∑ represents the summation symbol.

[0084] In some implementations, a progressive SVD algorithm can be used. This involves first decomposing the light field snapshots at key time points, and then obtaining the complete time modulation function through time interpolation. This method is particularly suitable for scenarios where light effects change relatively regularly. For example, in medical imaging applications, when it is necessary to simulate the scattered light characteristics of internal human tissues, the progressive SVD algorithm can accurately capture the changes in light scattering patterns at different tissue depths, helping doctors obtain clearer visualizations of tissue structures.

[0085] Step 2.2, spatial basis function optimization;

[0086] The set of spatial basis functions obtained by decomposition Sparsity optimization is performed to reduce redundant basis functions and improve computational efficiency.

[0087]

[0088] Where ||·||1 represents the L1 norm; ||·|| F Denotes the Frobenius norm; ∈ thresh The tolerance threshold represents the maximum allowable value of the reconstruction error. Represents the basis functions Optimize to minimize the objective function; st stands for "subject to", guiding the constraints; n basis L represents the total number of basis functions and the number of principal components retained after decomposition. proj (x, y, t) is the light field function projected onto the two-dimensional plane; The weight coefficients represent the importance of the i2th basis function; ∑ represents the time modulation function, indicating the variation of each spatial pattern over time; ∑ represents the summation symbol.

[0089] The spatial basis function optimization process is implemented through an iterative soft thresholding algorithm (ISTA), gradually converging to the optimal solution for each basis function. For example, in jewelry product photography, it is necessary to highlight the fire effect of a diamond by controlling the microstructure of light. The system first obtains an initial basis function set through SVD decomposition, and then sparse optimization algorithm reduces the initial 128 basis functions to 24 key basis functions, preserving the accurate expression of the reflection characteristics of different diamond facets.

[0090] Step 2.3, smoothing of the time function;

[0091] For time modulation function Smoothing is performed to eliminate high-frequency noise and ensure the continuity and smoothness of light effect changes:

[0092]

[0093] Among them G σ Let be the Gaussian smoothing kernel function, representing a Gaussian function with standard deviation σ as a parameter; σ smooth The smoothing parameter controls the degree of smoothing; the larger the value, the stronger the smoothing effect. τ is the integration variable, representing the sampling points on the time axis. dτ is the smoothed time modulation function; dτ is the differential element in the time dimension. is the time modulation function; ∫ represents the integral symbol.

[0094] Step 2.4, construct the physical mapping relationship;

[0095] The optimized spatial basis functions and temporal modulation functions are mapped to the control parameters of the physical optical modulator, establishing a mapping relationship between the mathematical model and the physical implementation:

[0096]

[0097] Where Θ(x, y, t) is the control parameter matrix of the optical modulator, representing the control signal values ​​at position (x, y) and time t; f map The mapping function converts the theoretical light field distribution into specific physical control parameters, taking into account the nonlinear response characteristics and physical limitations of the device. This is the time modulation function after smoothing. These are spatial basis functions, representing different spatial distribution patterns; The weight coefficients represent the importance of the i2th basis function; n basis∑ represents the total number of basis functions, indicating the number of principal components retained after decomposition; ∑ represents the summation symbol.

[0098] This step decomposes the complex four-dimensional light field model into a combination of spatial and temporal dimensions using a hierarchical synthesis algorithm, significantly reducing computational complexity while maintaining high-precision control over the light field microstructure. Furthermore, sparsity optimization and smoothing processes improve the method's robustness and real-time performance, providing a feasible solution for subsequent physical implementation.

[0099] Step 3: Based on the control parameters of the light field microstructure, a combination of a microelectromechanical system programmable reflective array and a liquid crystal spatial light modulator is used, along with a multi-channel LED array, to achieve sub-millimeter-level light distribution control and microsecond-level dynamic reconstruction of the light field;

[0100] This step combines a micro-electro-mechanical system (MEMS) programmable reflective array with a liquid crystal spatial light modulator (SLM), along with a multi-channel LED array, to transform the light field microstructure control parameters generated in the first two steps into actual physical light control, achieving sub-millimeter-level light distribution control and microsecond-level dynamic reconstruction of the light field. The specific execution is as follows:

[0101] Step 3.1, Construction of a multi-channel LED light source array;

[0102] A light source array consisting of multiple independently controlled LED units is configured, each unit having its brightness, color temperature, and on / off status independently adjustable, forming the initial light source layer:

[0103]

[0104] Where S(x, y, t, λ) represents the spectral radiance at position (x, y), time t, and wavelength λ; N LED Indicates the number of LED units in the light source array; Let i be the time modulation function for the i3rd LED unit; Let be the spatial spectral distribution function of the i3th LED unit at position (x, y) and wavelength λ, which describes the distribution characteristics of the LED light source in the spatial and wavelength dimensions; i3 represents the index of different LED units; ∑ represents the summation symbol.

[0105] Optionally, in some implementations, the LED array may employ a six-color or more color channel configuration to provide wider color gamut coverage and more precise spectral control. For example, in high-end product photography, adding amber and cyan LED units can achieve accurate reproduction of the unique luster of metallic materials. Furthermore, the physical arrangement of the LED units can be optimized according to application requirements; for example, a ring array configuration can be used in portrait photography, while a hemispherical array configuration can be used in product photography to obtain a more uniform light envelope effect.

[0106] Step 3.2, Microelectromechanical reflector array control;

[0107] A reflective array composed of thousands of miniature adjustable mirrors is constructed based on MEMS technology. Each mirror can independently adjust its angle within microseconds to control the direction of light reflection.

[0108]

[0109] Where R(x, y, t) represents the reflection distribution function of the reflective array, describing the reflection characteristics at position (x, y) and time t; N MEMS The total number of microreflectors represents the number of micromirror units in the MEMS array; Let δ be the reflection state of the j2th microreflector at time t; δ is the two-dimensional Dirac function, representing the position of the microreflector, when (x, y) equals The value is 1 when the time is right, and 0 otherwise. Here are the position coordinates of the j2-th microreflector; j2 represents the index of a different microreflector; N LED This indicates the number of LED units in the light source array; ∑ represents the summation symbol.

[0110] Step 3.3, configuration of the liquid crystal spatial light modulator;

[0111] A spatial light modulator (SLM) is introduced as the second-stage modulation unit. By adjusting the transmittance of each pixel unit, fine control of the light field intensity is achieved.

[0112] M(x,y,t)=f SLM (Θ SLM (x,y,t));

[0113] Where M(x, y, t) is the spatial modulation function of the SLM, representing the optical modulation effect at position (x, y) and time t; Θ SLM (x, y, t) are the SLM control parameters obtained from step 2.4, representing the control signals applied to the SLM; f SLMThis is the response function of the SLM, which converts the control signal into the actual light modulation effect.

[0114] In some high-precision applications, phase-modulated SLMs can be chosen instead of amplitude-modulated SLMs to achieve finer light field control by modulating the phase of the light wave rather than its intensity. For example, in photomicrography, phase-modulated SLMs can generate more complex light field distributions to compensate for aberrations caused by the sample or create specific illumination patterns, thereby improving image quality. Another alternative is to use digital micromirror devices (DMDs) instead of liquid crystal SLMs. Although the resolution may be lower, the response speed can be increased to the microsecond level, making it suitable for applications requiring ultra-fast changes in light efficiency.

[0115] Step 3.4, multi-level optical system integration;

[0116] By integrating the three core components (LED array, MEMS reflective array, and SLM) into a multi-level optical system, precise reconstruction of the light field can be achieved through optical path configuration.

[0117]

[0118] Where L out (x, y, z, t) represents the final output four-dimensional light field, indicating the light field distribution at spatial location (x, y, z) and time t; K is the point spread function of the optical system, describing the propagation characteristics of light from the source point (x′, y′) to the target point (x, y, z), considering the influence of wavelength λ; (x′, y′) is the integration variable, representing the coordinates on the light source plane; dx′, dy′, and dλ represent the differential elements of x′, y′, and wavelength dimension λ, respectively; (x, y, z) represents the spatial coordinates, and (x′, y′) represents the coordinates on the light source plane; ∫ represents the integration sign.

[0119] As can be seen, this step employs a multi-level dimming architecture, combining the spectral control capabilities of the LED array, the high-speed directional control capabilities of the MEMS reflective array, and the fine intensity modulation capabilities of the SLM, to achieve comprehensive control of the light field in both spatial and temporal dimensions. In particular, the microsecond-level response speed of MEMS technology breaks through the speed limitations of traditional optical modulators, providing a hardware foundation for high-frequency dynamic lighting effects. Simultaneously, the redundancy configuration of the multi-level dimming architecture enhances the system's fault tolerance and adaptability, enabling it to cope with various complex lighting effect requirements.

[0120] Step 4: Based on the results of dynamic reconstruction of the light field, a precise synchronization mechanism between the shutter speed of the shooting device and changes in light effect is established through a predictive control algorithm accelerated by optical computation.

[0121] This step utilizes a computationally accelerated predictive control algorithm to establish a precise synchronization mechanism between the camera's shutter speed and changes in lighting conditions, ensuring accurate matching between changes in lighting conditions and the imaging timing during continuous shooting. The specific execution is as follows:

[0122] Step 4.1, constructing the signal interface for the shooting device;

[0123] Build a hardware interface for communication with various professional shooting equipment to obtain shutter signals and timecode information:

[0124] S camera (t)=t trigger ,t exposure ,△t latency ,f fps ;

[0125] Where S camera (t) represents the camera status signal, describing the camera's operating state at time t; t trigger The trigger time indicates the moment when the camera shutter is triggered; t exposure Δt represents the exposure time, indicating the duration of exposure for a single shot. latency Signal delay, representing the time delay from the issuance of a trigger signal to the actual response from the camera; f fps Frame rate, representing the number of frames captured by the camera per second.

[0126] Step 4.2, Light Effect-Exposure Timing Mapping;

[0127] Based on the status signals of the shooting equipment and the preset light effect animation sequence, the optimal light effect timing control parameters are calculated:

[0128] T light (i4)=T camera (i4)-△t comp +f offset (i4);

[0129] Where T light (i4) represents the trigger time of the light effect in the i4th frame, indicating the moment when the light effect in the i4th frame should be triggered; T camera (i4) represents the exposure time of the i4th frame of the camera, indicating the moment when the camera begins exposure for the i4th frame; Δt comp This is the system delay compensation time, used to compensate for delays introduced by various components of the system; f offset (i4) is a time offset function set according to creative needs, allowing users to adjust the relative timing of lighting effects and exposure according to artistic requirements; i4 represents the frame number.

[0130] Step 4.3, Implementation of the predictive control algorithm;

[0131] Employing a predictive control algorithm based on optical computing acceleration, the system predicts the precise time for the next frame to be captured based on historical data, thus triggering changes in lighting effects in advance.

[0132] T camera (i4+1)=f predict (T camera (i4-n2), ...,T camera (i4));

[0133]

[0134] Where T camera (i4+1) is the predicted time to capture the next frame; f predict T is a prediction function that predicts future states based on historical data. camera (i4-n2) and T camera (i4) represents the capture time of the i4-n2th and i4th historical frames, respectively, and serves as the input data for prediction; Δt represents the light effect trigger time, indicating the moment when the next frame's light effect should be triggered. system Δt represents the time delay from the issuance of the light effect control signal to the actual change in light effect. safety This is a safety margin to account for prediction errors and system fluctuations; n2 represents the number of historical frames used for prediction.

[0135] Prediction function f predict Based on the Kalman filter, the Kalman filter method combines the physical motion model of the shooting device with real-time observation data, effectively handling randomness and system noise during the shooting process. For example, in applications involving high-speed continuous shooting of moving objects, the system uses the Kalman filter algorithm to predict the camera's next trigger time, taking into account factors such as camera exposure time, frame rate changes, and camera shake. Experimental results show that the predictive control algorithm achieves an average time prediction error within ±0.2 milliseconds in a 60fps high-speed continuous shooting scenario, representing an approximately 3-fold performance improvement compared to a simple linear prediction model. This significantly reduces image blurring or incomplete light capture caused by asynchrony between lighting effects and shooting.

[0136] Optionally, in certain specific scenarios, a deep learning-based temporal prediction model can be used to replace the Kalman filter. By learning from a large amount of historical shooting data, it can better adapt to complex or nonlinear shooting behavior patterns. For example, in applications that track and film wildlife, deep learning models can learn the regularity and randomness of animal movements, providing more accurate predictions and thus achieving more precise control of lighting synchronization.

[0137] Step 4.4, closed-loop feedback calibration mechanism;

[0138] This enables real-time monitoring of changes in light effect and actual synchronization with camera exposure, dynamically adjusting timing parameters to achieve closed-loop control.

[0139]

[0140] Where Δt error (i4) and Δt error (i4-1) represent the timing errors of the i4th frame and the i4-1th frame, respectively; and These are the actual trigger times for the camera and lighting effects, respectively; Δt comp (i4+1) and Δt comp (i4) represent the system delay compensation time for the next frame and the current frame, respectively; K p and K d These represent the proportional coefficient and the derivative coefficient, respectively, used to control the response characteristics of the system.

[0141] This step introduces a high-precision synchronization mechanism that combines predictive control with closed-loop feedback. It predicts the behavior of the shooting equipment and triggers changes in lighting effects in advance, while continuously optimizing synchronization accuracy through closed-loop feedback. In particular, the use of optical computing acceleration technology significantly reduces signal processing latency, enabling the system to handle high-speed continuous shooting scenarios and adapt to different types of professional shooting equipment, demonstrating broad compatibility and practicality.

[0142] Step 5: Based on the precise synchronization mechanism, a keyframe-based lighting animation programming interface is provided to support the definition of dynamic change sequences of light trajectories, intensity, and shape, and to realize these changes as physical lighting effects;

[0143] This step provides a keyframe-based programming interface for lighting effects animation, allowing photographers to define dynamic sequences of light trajectories, intensity, and shapes, enabling intuitive programming control of the four-dimensional light field. The specific execution is as follows:

[0144] Step 5.1, Define the keyframe for the lighting effect;

[0145] Build a user-friendly graphical interface that allows users to intuitively set keyframes for lighting effect sequences:

[0146]

[0147] Where K frame Let n represent the set of keyframes, and n3 represent the total number of keyframes. For the i5th keyframe, include:

[0148]

[0149] in Here, represents the time corresponding to the i5th keyframe; Spatial distribution describes the distribution pattern of light effects in three-dimensional space; Intensity indicates the brightness level of the luminous effect; For color, it represents the color characteristics of the light effect; These are special effects parameters, including special lighting effects such as scattering and focusing; i5 represents the index of the keyframe; (x, y, z) represents the spatial coordinates.

[0150] Step 5.2, keyframe interpolation algorithm;

[0151] Based on user-defined keyframes, an advanced interpolation algorithm is applied to generate a complete sequence of light effect animations:

[0152]

[0153] Where L field (x, y, z, t) represents the light field at time t and spatial location (x, y, z); f interp As an adaptive interpolation function, it automatically selects an appropriate interpolation method (such as linear interpolation, Bézier curve interpolation, spline interpolation, etc.) based on the luminous efficacy characteristics; and These represent the i5th and i5+1th keyframes, respectively; t represents the current time point. and These represent the time points of the i5th and i5+1th keyframes, respectively.

[0154] In its implementation, the system provides an adaptive interpolation algorithm library that automatically selects the most suitable interpolation method based on the characteristics of the changing light effects. For example, for soft light effects requiring smooth transitions, the system uses third-order Hermitian spline interpolation; for light effects requiring clear acceleration and deceleration effects, the system uses Bézier curve interpolation based on easing functions; and for light effects requiring precise mathematical changes, the system uses an interpolation method based on a physical model. In a creative photography example, the photographer needed to simulate the changing effect of sunlight filtering through leaves in a forest at sunset. By analyzing the characteristics of keyframes, the system automatically selected an interpolation algorithm based on a physical light propagation model, generating a dynamic light effect sequence that includes natural light spot movement and intensity changes. This resulted in portrait photos that presented a realistic forest sunset atmosphere, while the photographer only needed to set the initial and final keyframes for the light effects.

[0155] Step 5.3, Building the Light Effect Template Library;

[0156] A library of preset lighting effect templates has been built, containing commonly used lighting animation effects, which users can directly call or modify based on the templates:

[0157]

[0158] Where L template (x, y, z, t) represents the template light field at time t and spatial location (x, y, z); For template type, j3 represents the template index; P user Allow users to define parameters to adjust the specific appearance of the template; f template This is a template generation function that converts the template type and user parameters into a specific light field distribution.

[0159] Step 5.4, Real-time preview and adjustment mechanism;

[0160] Enables real-time preview of lighting effects animations, allowing users to instantly view and adjust the animation effects:

[0161] L preview (x,y,z,t)=f simplify (L field (x,y,z,t),β simplify );

[0162] Where L preview (x, y, z, t) represents the simplified light field at time t and spatial location (x, y, z); f simplify To simplify the calculation function and reduce computational complexity; L field (x,y,z,t) represents the light field at time t and spatial location (x,y,z); β simplify This is a simplification parameter used to reduce computational complexity during preview; a larger value indicates a higher degree of simplification.

[0163] Optionally, on a high-performance computing platform, the system can employ parallel computing techniques to accelerate light field simulation. For example, it can utilize the parallel processing capabilities of the Graphics Processing Unit (GPU) to perform large-scale ray tracing calculations, thereby achieving real-time preview without sacrificing accuracy. Another optional implementation approach is to adopt a progressive rendering strategy, first quickly generating a low-precision preview, and then gradually increasing the precision as the user observes, providing a high-quality preview effect while ensuring interactive responsiveness. For example, in the design of cross-media art installations, artists can first quickly preview the general effect of the lighting, confirm the creative direction, and then the system can gradually present more refined lighting details, making the design process more efficient.

[0164] Furthermore, this step provides a set of intuitive and easy-to-use four-dimensional lighting effects programming tools for creative professionals, encapsulating complex light field control technology within a user-friendly interface. Through mechanisms such as keyframe animation, preset templates, and real-time preview, the creative threshold is greatly lowered, enabling non-technical professionals to easily create complex dynamic lighting effects. At the same time, the system retains advanced customization capabilities to meet the fine-grained control needs of professional users.

[0165] A mobile terminal includes a memory and one or more processors, wherein the memory stores executable code, and when the one or more processors execute the executable code, they are used to implement the above-described soft light control method.

[0166] Here, the present invention provides an implementation example:

[0167] Dynamic lighting control in jewelry photography

[0168] In a product photography assignment for a jewelry brand, the photographer needed to showcase the changing fire and luster of diamond jewelry under different lighting conditions. Traditional shooting methods typically require multiple adjustments to light positions and parameters, followed by taking multiple photos, a cumbersome process that struggles to capture dynamic lighting effects. The soft light control method provided in this implementation simplifies and enhances the shooting workflow.

[0169] In the light field spatiotemporal analysis stage: the photographer first selects the target light effect type as "Dynamic Display of Diamond Fire". The system automatically determines the key light field parameters describing the diamond's fire through a preset spectral analysis model, including 8 main spatial frequency components and 3 temporal frequency components. These components together describe the complex dynamic patterns of light reflection, refraction, and dispersion on different facets of the diamond.

[0170] In the light field microstructure generation stage: based on the geometric characteristics and optical properties of diamond, the system generates 18 spatial basis functions and 6 temporal modulation functions. By combining these functions, a complete dynamic light field description is formed. After singular value decomposition, the system simplifies the 128 initial light field components into 24 key components, significantly reducing computational complexity while maintaining visual effects.

[0171] Physical Implementation Mechanism Stage: The system is configured with a hemispherical light source array consisting of 64 RGB+WLED units, combined with a 1920×1080 resolution liquid crystal SLM and a MEMS array with 7500 micro-reflective units. This configuration allows the system to precisely control the angle, intensity, and color of incident light on each facet of the diamond, thereby producing a realistic dynamic fire effect. Actual testing shows that the system can control the light distribution with sub-millimeter precision, with uniformity error controlled within ±1.5%.

[0172] In the high-precision synchronization control stage, the photographer continuously captures the diamond's changes under different lighting conditions at a frame rate of 24fps. The system uses a Kalman filter prediction algorithm to ensure that the changes in lighting effects in each frame are precisely synchronized with the camera's exposure, with an average synchronization error of only 0.15 milliseconds, far lower than the time difference perceptible to the human eye.

[0173] Programmable lighting animation stage: Photographers only need to set three key lighting states: the initial state simulates the appearance of a diamond under natural light, the intermediate state showcases the colorful fire effect, and the final state highlights the white luster of the diamond. The system automatically generates a smooth 60-second transition animation through Hermit spline interpolation, making the changes in the diamond's lighting effect both natural and dramatic.

[0174] By comparing traditional shooting methods with the shooting method of this embodiment, the following quantifiable technical effects were obtained:

[0175] Improved shooting efficiency: Traditional methods require 4-6 hours of repeated lighting adjustments and multiple shots, while this method only requires 40 minutes to complete the entire shooting process, increasing efficiency by approximately 85%.

[0176] Light efficiency accuracy: The light distribution accuracy achieved by this method reaches 0.8 mm, and the uniformity error is controlled within ±1.5%, which is an improvement compared to the 3-5 mm accuracy and ±7% uniformity error of the traditional method.

[0177] Dynamic performance capability: This method captures the continuous dynamic effects of diamonds under changing light, showcasing complex optical changes that traditional static photography methods cannot present, improving product display effect score by 65%.

[0178] Creative control: Photographers complete lighting design through an intuitive graphical interface without needing to deeply understand complex optical principles, reducing creation time by 75% and achieving final results that better meet expectations.

[0179] The above application examples demonstrate that the soft light control method provided in this embodiment has technical advantages and application value in actual professional photography scenarios, and can effectively solve the problems of accuracy, efficiency and expressiveness faced by traditional methods.

[0180] Detailed data recording and analysis were conducted for the two most critical technical effects in this embodiment. The test results of the optical field microstructure control accuracy are shown in Table 1:

[0181] Table 1: Test results of optical field microstructure control accuracy (average of 5 repeated tests)

[0182] Test parameters This method Traditional methods Improved proportions Spatial accuracy of light distribution 0.8±0.1 mm 4.2 ± 0.8 mm Increased by 81% Light field uniformity error ±1.5％ ±7.3％ 79% increase Minimum resolvable spot size 0.5 mm 2.8 mm Increased by 82% Brightness gradient transition band width 0.6 mm 5.5 mm Increased by 89% Spectral color accuracy (ΔE) 1.8 4.7 62% increase

[0183] The test results of light effect-shutter synchronization accuracy are shown in Table 2:

[0184] Table 2: Results of Light Effect-Shutter Synchronization Accuracy Test (Statistics from 100 Consecutive Shootings)

[0185] Test parameters This method Traditional methods Improved proportions Average synchronization error 0.15 milliseconds 8.6 milliseconds 98% increase Maximum synchronization error 0.42 milliseconds 15.3 milliseconds 97% increase Synchronization stability (standard deviation) 0.08 milliseconds 4.2 milliseconds 98% increase Prediction accuracy (24fps) 98.5％ 65.3％ Increased by 51% Light effect conversion response time 5 microseconds 8.2 milliseconds Improved by 99.9%

[0186] The data above clearly demonstrates that this implementation method achieves an average improvement of approximately 79% in the precision of light field microstructure control compared to traditional methods, and an average improvement of approximately 89% in the precision of light effect-shutter synchronization. These technological breakthroughs enable this method to achieve unprecedented creative light effect control capabilities in the field of high-end photography, providing photographers with entirely new tools for artistic expression.

[0187] The embodiments of the present invention have been described above. However, the embodiments are not limited to the specific implementation methods described above. The specific implementation methods described above are merely illustrative and not restrictive. Those skilled in the art can make more equivalent embodiments under the guidance of the present embodiments, and all of them are within the protection scope of the present embodiments.

Claims

1. A method for controlling a soft light, characterized in that, include: A four-dimensional light field harmonic model is constructed based on spatiotemporal Fourier analysis, decomposing the target light effect into a spatiotemporal harmonic combination to achieve a precise mathematical expression of the light field. The steps for constructing a four-dimensional light field harmonic model based on spatiotemporal Fourier analysis include: Collect the target light effect characteristics to form a light field sampling dataset; Perform a four-dimensional Fourier transform on the collected light field data to convert the spatiotemporal domain data into a frequency domain representation; Based on the frequency domain energy distribution, the main harmonic components whose energy proportion exceeds a preset threshold are selected to construct a simplified four-dimensional optical field harmonic model. The parameters of the harmonic model are optimized using the least squares method to minimize the error between the model output and the target light effect. Using the output of a four-dimensional light field harmonic model, the target light effect is decomposed into a spatiotemporal mapping sequence through a light field layered synthesis algorithm, forming physically realizable light field microstructure control parameters; the step of decomposing the target light effect into a spatiotemporal mapping sequence through the light field layered synthesis algorithm includes: The four-dimensional light field model is projected onto a two-dimensional plane and the time dimension. The complex spatiotemporal changing light field is decomposed into a combination of spatial basis functions and time modulation functions through the singular value decomposition method. Sparsity optimization is performed on the spatial basis function set obtained by decomposition to reduce redundant basis functions and improve computational efficiency; The time modulation function is smoothed to eliminate high-frequency noise and ensure the continuity and smoothness of light effect changes; The optimized spatial basis function and temporal modulation function are mapped to the control parameters of the physical optical modulator; Based on the control parameters of the light field microstructure, a combination of a microelectromechanical system programmable reflective array and a liquid crystal spatial light modulator is used, along with a multi-channel LED array, to achieve sub-millimeter-level light distribution control and microsecond-level dynamic reconstruction of the light field; Based on the results of dynamic reconstruction of the light field, a precise synchronization mechanism between the shutter speed of the shooting device and changes in light effect is established through a predictive control algorithm accelerated by optical computing. The predictive control algorithm is implemented through a Kalman filter and, combined with the physical motion model of the shooting device and real-time observation data, predicts the next trigger time of the camera. Based on a precise synchronization mechanism, it provides a keyframe-based programming interface for light effect animation, supporting the definition of dynamic change sequences of light trajectories, intensity, and shape, and realizing these changes as physical lighting effects.

2. The soft light control method according to claim 1, characterized in that, The step of combining a microelectromechanical system programmable reflective array with a liquid crystal spatial light modulator includes: A light source array consisting of multiple independently controlled LED units is configured, each of which can independently adjust its brightness, color temperature, and on / off state to form an initial light source layer; A reflective array consisting of thousands of miniature adjustable mirrors is constructed based on MEMS technology. Each mirror can independently adjust its angle within microseconds to control the direction of light reflection. A liquid crystal spatial light modulator is introduced as the second-stage modulation unit. By adjusting the transmittance of each pixel unit, the light field intensity can be precisely controlled. The initial light source layer, reflection array, and second-level modulation unit are integrated into a multi-level optical system, and the optical field is accurately reconstructed through optical path configuration.

3. The soft light control method according to claim 1, characterized in that, The steps for establishing a precise synchronization mechanism between the shutter speed of the shooting device and changes in light effects include: Build a hardware interface for communication with various professional shooting equipment to obtain shutter signals and timecode information; Based on the status signals of the shooting equipment and the preset light effect animation sequence, calculate the optimal light effect timing control parameters; A predictive control algorithm based on optical computing acceleration is adopted to predict the precise time of the next frame capture based on historical data, thereby triggering changes in light effects in advance. It enables real-time monitoring of changes in light effect and actual synchronization with camera exposure, dynamically adjusts timing parameters, and achieves closed-loop control.

4. The soft light control method according to claim 1, characterized in that, The steps for providing a keyframe-based lighting effect animation programming interface include: Build a user-friendly graphical interface that allows users to intuitively set keyframes for lighting effect sequences; Based on user-defined keyframes, an advanced interpolation algorithm is applied to generate a complete sequence of light effect animations; A library of preset lighting effect templates has been built, which includes commonly used lighting and animation effects that users can directly call or modify based on the templates. It enables real-time preview of light effect animations, allowing users to instantly view and adjust the animation effects.

5. The soft light control method according to claim 1, characterized in that, The expression for the four-dimensional optical field harmonic model is: ； in This represents the reconstructed light field distribution function. The amplitude coefficient represents the intensity of each harmonic component; , , , They are respectively direction, direction, Spatial and temporal frequencies of direction , , , They represent direction, direction, Directional spatial frequency and Index of time frequency; The phase shift represents the initial phase of each harmonic component. It is a sine function; The summation symbol is used.

6. The soft light control method according to claim 1, characterized in that, The expression for decomposing the complex spatiotemporally varying light field into a combination of spatial basis functions and temporal modulation functions is as follows: ； in Let be the light field function projected onto a two-dimensional plane, representing the light field on the plane. Over time The changing light field distribution; These are spatial basis functions, representing different spatial distribution patterns; This is a time modulation function, representing the variation of each spatial pattern over time; Indicates the first Weighting coefficients for the importance of each basis function; The total number of basis functions represents the number of principal components retained after decomposition; Indicates the index of different basis functions; The summation symbol is used.

7. A mobile terminal, characterized in that, The device includes a memory and one or more processors, wherein the memory stores executable code, and the one or more processors execute the executable code to implement a soft light control method according to any one of claims 1-6.

Images