Intelligent adaptive projection lamp and projection control method

Abstract

The invention provides an intelligent adaptive projection lamp and a projection control method, and the method comprises the steps: collecting multi-source environment data which comprises at least one of environment illumination data, projection surface characteristic data and user operation data; performing preprocessing, feature extraction and feature fusion on the multi-source environment data to generate an environment feature vector; performing projection mapping on the environment feature vector through a parameter prediction model based on rate distortion optimization to generate projection parameters; performing image deblurring processing and adaptive k-space sampling reconstruction on the projection parameters to generate an optimized projection image; collecting projection effect feedback data for the optimized projection image; and performing online updating optimization on the parameter prediction model according to the projection effect feedback data. According to the invention, intelligent adaptive adjustment of the projection parameters is realized, and the projection quality is improved.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence and image processing technology, and in particular to an intelligent adaptive projection lamp and a projection control method. Background Art

[0002] With the rapid development of artificial intelligence technology, intelligent projection equipment has been widely used in office, education, entertainment and other fields. The projection effect of projection equipment directly affects the user experience, so how to achieve intelligent optimization and adjustment of projection parameters has become the focus of the industry.

[0003] At present, common projection equipment mainly adopts projection correction methods based on fixed parameters or simple adaptive adjustment schemes. The projection correction method based on fixed parameters optimizes the projection effect through preset parameters such as brightness and contrast, but cannot be adaptively adjusted for different environments. Simple adaptive adjustment schemes only consider a single environmental factor (such as ambient light) for parameter adjustment and cannot achieve multi-dimensional collaborative optimization.

[0004] Some improved solutions introduce machine learning technology to optimize projection parameters through traditional machine learning models. For example, the prior art discloses a method for adaptively adjusting projection parameters based on a support vector machine, which achieves predictive optimization of projection parameters by building a support vector machine model. However, due to the limited expressive power of traditional machine learning models, it is difficult to achieve precise control in complex scenes, resulting in less than ideal projection effects.

[0005] Therefore, how to achieve intelligent and precise adjustment of projection parameters and improve projection quality has become a technical problem that needs to be solved urgently. Summary of the invention

[0006] In view of this, the present application provides an intelligent adaptive projection lamp and a projection control method, which can realize intelligent and precise adjustment of projection parameters and improve projection quality.

[0007] The embodiment of the present application provides an intelligent adaptive projection control method, including: Collecting multi-source environmental data, wherein the multi-source environmental data includes at least one of environmental lighting data, projection surface characteristic data, and user operation data; Preprocessing, feature extraction and feature fusion of the multi-source environmental data to generate an environmental feature vector; Projecting the environment feature vector using a parameter prediction model based on rate-distortion optimization to generate projection parameters; Performing image deblurring processing and adaptive k-space sampling reconstruction on the projection parameters to generate an optimized projection image; Collecting projection effect feedback data for the optimized projection image; The parameter prediction model is updated and optimized online according to the projection effect feedback data.

[0008] Optionally, collecting multi-source environmental data includes: Collecting the ambient light data by an ambient light sensor, wherein the ambient light data includes ambient light intensity data; Acquiring the projection surface characteristic data through a projection surface detector, wherein the projection surface characteristic data includes reflectivity and texture feature data of the projection surface; Through the user interaction interface, the user's adjustment operation data on the projection parameters is collected to obtain the user operation data.

[0009] Optionally, the preprocessing, feature extraction and feature fusion of the multi-source environmental data to generate an environmental feature vector includes: Performing noise filtering and outlier detection processing on the ambient light data to obtain cleaned ambient light data; Performing geometric correction and texture feature extraction on the projection surface characteristic data to obtain standardized surface characteristic data; Performing time sequence alignment and behavior pattern analysis on the user operation data to obtain standardized user operation data; A convolutional neural network is used to extract standardized surface feature data to obtain image features; Use recurrent neural networks to analyze normalized user operation data and obtain time series features; The image features, the time series features and the cleaned ambient light data are fused through an attention mechanism, and the fused data is the ambient feature vector.

[0010] Optionally, the parameter prediction model is a multi-layer deep neural network, and the multi-layer deep neural network includes a feature receiving layer, a multi-layer fully connected layer based on a rectified linear unit activation function, and a parameter prediction layer; The step of performing projection mapping on the environmental feature vector to generate projection parameters by using a parameter prediction model based on rate distortion optimization includes: Receiving the environment feature vector through the feature receiving layer; In the fully connected layer, a deep representation corresponding to the environment feature vector is generated through a nonlinear transformation based on a ReLU activation function and a residual connection; In the parameter prediction layer, projection mapping is performed on the deep representation corresponding to the environmental feature vector based on multiple independent prediction units to generate the projection parameters, wherein one prediction unit corresponds to one projection parameter.

[0011] Optionally, a parameter prediction model based on rate-distortion optimization is obtained in the following manner; Construct training data sets and validation sets based on pre-collected historical environmental data and corresponding optimal projection parameters; Constructing a loss function based on rate-distortion theory, wherein the loss function includes a parameter prediction error term and a rate constraint term; Using an improved stochastic gradient descent algorithm, based on the loss function and the training data set, to train the preset neural network model, and in the training process, using an adaptive moment estimation optimizer and a dynamic learning rate adjustment strategy; The model effect is evaluated based on the validation set, and hyperparameters including batch size and learning rate are dynamically adjusted to generate the parameter prediction model.

[0012] Optionally, performing image deblurring processing and adaptive k-space sampling reconstruction on the projection parameters to generate an optimized projection image includes: A neural network encoder based on prior knowledge encodes the projection image corresponding to the projection parameters to obtain an implicit representation of the feature encoding of the projection image; A decoder network based on a multi-scale structure restores the implicit representation of the feature encoding by layer-by-layer feature reconstruction to obtain a reconstructed image corresponding to the projected image; The reconstructed image is fused with the projection image through a residual learning strategy to generate a clear projection image; Based on a learnable sampling pattern generator, an adaptive acquisition operation is performed to determine the optimal sampling trajectory according to the content characteristics of the clear projection image; A reconstruction model based on a deep convolutional network of the U-Net architecture is used to preserve image details of different scales through skip connections. Based on Fourier consistency constraints, the optimal sampling trajectory and image details of different scales are operated consistently to generate the initial optimized projection image; An image registration network is used to perform precise alignment, and geometric correction is performed on the initial optimized projection image to generate the optimized projection image.

[0013] Optionally, in the step of obtaining a reconstructed image corresponding to the projected image, the projection control method further includes: adopting a spatial attention mechanism to enhance the detail performance of important areas, and adding an adaptive weight adjustment strategy to dynamically adjust the deblurring intensity according to the blur degree of the image area.

[0014] Optionally, the collecting projection effect feedback data for the optimized projection image includes: The brightness uniformity, color reproduction and contrast of the projected image are collected through optical sensors to form a time-series data stream; Determining quantitative indicators of the projection image by a preset evaluation algorithm, wherein the quantitative indicators include peak signal-to-noise ratio and structural similarity; Collecting user evaluation data through a user interaction interface, wherein the evaluation data includes satisfaction ratings and feedback information; Record the user's parameter adjustment operation behavior data; The time series data stream, the quantitative index, the user evaluation data and the operation behavior data are cleaned and standardized to obtain standardized projection effect feedback data.

[0015] Optionally, the online updating and optimizing of the parameter prediction model according to the projection effect feedback data includes: Based on the normalized feedback data, an incremental learning strategy is adopted to gradually update the parameters of the parameter prediction model through a small batch gradient descent method. In the parameter updating process, the parameter prediction model is adjusted according to the consistency of the feedback data based on a dynamic learning rate adjustment mechanism, and the empirical knowledge of the historical model is transferred to the updated parameter prediction model based on the knowledge distillation technology. By comparing the projection qualities of the projection images before and after the parameter prediction model is updated, and based on the comparison result, verifying the performance improvement of the parameter prediction model, and deciding whether to accept the update based on the comparison result.

[0016] An embodiment of the present application further provides an intelligent adaptive projection lamp, comprising a processor and a memory, wherein the memory is used to store a computer program, and when the processor executes the computer program, the above-mentioned projection control method is implemented.

[0017] The present application also provides a computer device, the computer device comprising: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the above-mentioned intelligent adaptive projection control method.

[0018] An embodiment of the present application also provides a computer-readable storage medium, which stores computer instructions, and the computer instructions are used to enable a computer to execute the above-mentioned intelligent adaptive projection control method.

[0019] The embodiment of the present application also provides a computer program product, including computer instructions, which implement the steps of the above-mentioned intelligent adaptive projection control method when executed by a processor.

[0020] This application has the following technical effects: By adopting a rate-distortion optimized parameter prediction model, combined with deep feature extraction and fusion of multi-source environmental data, accurate prediction and adaptive adjustment of projection parameters are achieved, significantly improving the projection quality. The implicit neural representation method based on prior knowledge and the deep end-to-end K-space processing solution effectively improve the image deblurring and reconstruction effects. Through online learning and model optimization mechanisms, the system can continue to learn and evolve, and continuously improve the projection effect. In other words, this application can realize intelligent and precise adjustment of projection parameters and improve projection quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the disclosed embodiments in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without paying any creative work.

[0022] Figure 1 A schematic diagram of a flow chart of an intelligent adaptive projection control method provided in an embodiment of the present application; Figure 2 This is a structural diagram of the intelligent adaptive projection lamp in an embodiment of the present application. DETAILED DESCRIPTION

[0023] The embodiments of the present disclosure are described in detail below with reference to the accompanying drawings.

[0024] It should be clear that the following embodiments of the present disclosure are described by specific specific examples, and those skilled in the art can easily understand other advantages and effects of the present disclosure from the contents disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of the present disclosure, rather than all of the embodiments. The present disclosure can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present disclosure. It should be noted that the following embodiments and features in the embodiments can be combined with each other in the absence of conflict. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present disclosure.

[0025] It should be noted that various aspects of the embodiments within the scope of the appended claims are described below. It should be apparent that the aspects described herein may be embodied in a wide variety of forms, and any specific structure and / or function described herein is merely illustrative. Based on the present disclosure, it should be understood by those skilled in the art that an aspect described herein may be implemented independently of any other aspect, and two or more of these aspects may be combined in various ways. For example, any number of aspects described herein may be used to implement the device and / or practice the method. In addition, other structures and / or functionalities other than one or more of the aspects described herein may be used to implement this device and / or practice this method.

[0026] It should also be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present disclosure. The drawings only show components related to the present disclosure rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

[0027] Additionally, in the following description, specific details are provided to facilitate a thorough understanding of the examples. However, it will be understood by those skilled in the art that the aspects described may be practiced without these specific details.

[0028] The specific implementation of the present application is described in detail below in conjunction with the accompanying drawings. It should be noted that the drawings in the specification of the present application are only used to explain and illustrate the present application, and are not intended to limit the scope of protection of the present application.

[0029] The technical solution of the embodiments of the present application is described in detail with reference to the accompanying drawings.

[0030] It should be noted that the technical terms used in the embodiments of the present application should be understood to have the meanings commonly understood by those skilled in the art, unless otherwise clearly specified and limited. The descriptions such as "first" and "second" used in the embodiments of the present application are only used to distinguish different entities or operations, and should not be understood as limiting the specific order of these entities or operations.

[0031] Combine the following Figure 1 , the specific implementation method of the embodiment of the present application is described in detail.

[0032] like Figure 1 As shown, the embodiment of the present application provides an intelligent adaptive projection control method, which includes the steps of multi-source environmental data collection, feature extraction and fusion, parameter prediction, image optimization processing, and online learning and updating.

[0033] It should be noted that this application uses deep learning technology to achieve intelligent optimization of projection parameters, which can effectively improve the projection quality.

[0034] More specifically, the intelligent adaptive projection control method includes: S1: Collect multi-source environmental data, where the multi-source environmental data includes at least one of environmental lighting data, projection surface characteristic data, and user operation data.

[0035] Specifically, in step S1, the step of collecting multi-source environmental data is first performed, and this step achieves comprehensive acquisition of environmental information through multiple data collection modules.

[0036] Exemplarily, the ambient light sensor collects ambient light intensity data and monitors the light changes of the projection environment in real time.

[0037] The projection surface detector obtains the reflectivity and texture feature data of the projection surface, providing an important basis for subsequent parameter optimization. In addition, the user's adjustment operation data is collected through the user interaction interface to understand the user's personalized needs.

[0038] Step S1 specifically includes: S1.1: collecting the ambient light data through an ambient light sensor, where the ambient light data includes ambient light intensity data; In step S1.1, ambient light data is collected by an ambient light sensor.

[0039] It should be noted that the sensor uses a high-precision photosensitive element array, which can capture changes in lighting in the environment in real time.

[0040] Specifically, the ambient light sensor can measure the light intensity of the projection area and its surrounding environment at a sampling frequency of 10 times per second. The data collected by the ambient light sensor includes information such as the intensity value of the ambient light (in lux), the uniformity of the light distribution, and the light fluctuation trend.

[0041] In addition, the ambient light sensor in this application also has an anti-glare processing function, which can effectively filter out the interference of strong light sources and ensure the accuracy of collected data.

[0042] S1.2: Acquire the projection surface characteristic data through a projection surface detector, where the projection surface characteristic data includes the projection surface.

[0043] In step S1.2, the projection surface detector is used to obtain the projection surface characteristic data. First, the projection surface detector uses a high-resolution image sensor to perform multi-angle scanning and acquisition of the projection surface. Secondly, the reflectivity data of the projection surface is measured through the built-in spectral analysis module, including diffuse reflectivity and specular reflectivity.

[0044] It should be noted that when acquiring the projection surface characteristic data, the texture feature data of the projection surface is also collected, and the texture feature data may include information such as surface roughness, color uniformity, and material changes. The collection of these characteristic data provides an important basis for subsequent parameter optimization.

[0045] S1.3: Collecting the user's adjustment operation data on the projection parameters through the user interaction interface to obtain the user operation data.

[0046] The user interface provides intuitive parameter adjustment controls, including a brightness slider, contrast knob, color temperature selector, etc. When the user adjusts these controls, information such as the type of parameter adjusted, the adjustment direction, the adjustment amplitude, and the adjustment time will be recorded.

[0047] In addition, the user interaction interface also has quick operation buttons to facilitate users to quickly switch between preset parameter combinations.

[0048] It should be noted that this solution not only records the user's specific operation data, but also analyzes the user's operating habits and preferences to provide a reference for personalized parameter optimization.

[0049] It should be noted that, in actual operation, the projection parameters may be determined based on at least two of the ambient illumination data, the projection surface characteristic data and the user operation data.

[0050] S2: Preprocessing, feature extraction and feature fusion of the multi-source environmental data to generate an environmental feature vector.

[0051] It should be noted that this step uses multiple deep learning models for feature extraction and fusion.

[0052] First, the ambient light data is filtered to eliminate the influence of outliers. Secondly, the projection surface characteristic data is geometrically corrected to ensure the accuracy of feature extraction.

[0053] Specifically, a convolutional neural network is used to extract image features. The network contains multiple convolutional layers and pooling layers, which can effectively capture the spatial features of surface textures. At the same time, a recurrent neural network is used to analyze the temporal features of user operations, and the user behavior pattern is modeled through LSTM units. Finally, an attention mechanism is introduced to achieve feature fusion, which can adaptively adjust the weights of different features and generate high-quality environmental feature vectors.

[0054] S2 specifically includes: S2.1: Perform noise filtering and outlier detection processing on the ambient light data to obtain cleaned ambient light data.

[0055] In step S2.1, noise filtering and outlier detection processing are performed on the collected ambient light data.

[0056] First, a Gaussian filter is used to preprocess the ambient light data to eliminate the influence of random noise. Second, outlier detection is performed based on statistical methods to calculate the mean and standard deviation of the light data, and data points that deviate from the mean by more than a preset multiple of the standard deviation (for example, three times the standard deviation) are marked as potential outliers.

[0057] In addition, considering the dynamic change characteristics of ambient light, the system also introduces an adaptive threshold mechanism, which can dynamically adjust the abnormal value judgment standard according to the change law of light in different periods of time. Through these processes, high-quality ambient light data after cleaning is finally obtained.

[0058] It should be noted that this solution will not directly delete these outliers, but will use an interpolation method based on temporal context to correct them in order to expand the amount of ambient lighting data.

[0059] S2.2: Performing geometric correction and texture feature extraction on the projection surface characteristic data to obtain standardized surface characteristic data.

[0060] Specifically, the perspective transformation algorithm is first used to correct the geometric distortion of the projection surface to ensure the accuracy of feature extraction. Secondly, a multi-scale Gabor filter group is used to extract the texture features of the projection surface. The filter group contains filter kernels of different directions and scales, which can fully capture the texture information of the surface.

[0061] It should be noted that in the process of texture feature extraction, the non-uniformity of the surface material is also taken into account, and the robustness of feature extraction is improved through adaptive local feature enhancement technology. After these processes, the system obtains standardized surface feature data, laying the foundation for subsequent feature fusion.

[0062] S2.3: Perform time sequence alignment and behavior pattern analysis on the user operation data to obtain standardized user operation data.

[0063] Optionally, in S2.3, a dynamic time warping algorithm is used to align operation sequences of different time scales to ensure the consistency of temporal features. Secondly, the user operation is segmented by a sliding window technique so that each window contains continuous 30 seconds of operation data.

[0064] It should be noted that a hierarchical feature extraction strategy is used in the behavior pattern analysis, including low-level operation frequency statistics, mid-level operation combination pattern recognition, and high-level user preference inference. Through this multi-level analysis and processing, standardized user operation data is finally obtained.

[0065] S2.4: Use a convolutional neural network to extract standardized surface feature data to obtain image features.

[0066] Among them, the convolutional neural network adopts an improved ResNet structure, which contains five convolution blocks, each of which consists of two 3×3 convolution layers and a maximum pooling layer.

[0067] In addition, considering the diversity of projection surface features, an adaptive feature normalization layer is added to the convolutional neural network. The adaptive feature normalization layer can dynamically adjust the normalization parameters according to the feature distribution of different surfaces. Therefore, through this deep network structure, rich image features can be successfully extracted.

[0068] It should be noted that in order to improve the effect of feature extraction, attention mechanism and residual connection can be introduced in each convolution block.

[0069] S2.5: Use a recurrent neural network to analyze the normalized user operation data to obtain time series features.

[0070] First, a bidirectional LSTM network structure is adopted, which contains two layers of LSTM units, each layer contains 128 hidden states. Secondly, in order to improve the memory capacity of the network, the system also adopts a residual temporal connection structure, which enables the network to better capture long-term dependencies. Through the analysis of this recurrent neural network, the system obtains the temporal feature representation of user operation behavior.

[0071] It should be noted that a gated attention mechanism is introduced in the LSTM unit, which can adaptively focus on important temporal features.

[0072] S2.6: The image features, the time series features and the cleaned ambient lighting data are fused through an attention mechanism, and the fused data is the ambient feature vector.

[0073] In step S2.6, the image features, temporal features and cleaned ambient lighting data are fused through the attention mechanism.

[0074] Specifically, a multi-head self-attention mechanism is first used to adaptively weight features of different modalities. Secondly, features of different dimensions are mapped to the same feature space through a feature conversion network. In addition, in order to improve the expressiveness of fused features, a residual feature fusion structure is also used to combine the original feature information with the converted features. Finally, through this multimodal feature fusion strategy, the system generates high-quality environmental feature vectors, which provide reliable input for subsequent parameter prediction.

[0075] It should be noted that a dynamic weight adjustment mechanism is introduced in the feature fusion process, which can automatically adjust the importance weights of different features according to the characteristics of the current scene.

[0076] S3: Projection mapping is performed on the environmental feature vector through a parameter prediction model based on rate-distortion optimization to generate projection parameters.

[0077] In step S3, projection mapping is performed using a parameter prediction model based on rate-distortion optimization.

[0078] Exemplarily, the parameter prediction model adopts a multi-layer deep neural network structure, and the multi-layer deep neural network includes a feature receiving layer, a multi-layer fully connected layer based on a rectified linear unit activation function, and a parameter prediction layer.

[0079] Specifically, first, the feature receiving layer receives the environment feature vector to maintain the integrity of the input data. Secondly, the ReLU activation function is used in the fully connected layer for nonlinear transformation, and residual connections are introduced to alleviate the gradient vanishing problem.

[0080] It should be noted that the parameter prediction layer contains multiple independent prediction units, each of which is responsible for predicting a specific projection parameter (such as brightness, contrast, etc.). In addition, the model training uses a loss function based on rate-distortion theory, which comprehensively considers the accuracy and computational complexity of parameter prediction, and can optimize the use of computing resources while ensuring the prediction effect.

[0081] To facilitate understanding, a specific example is used to illustrate the working process of the parameter prediction model.

[0082] Assume that in a conference room environment, the ambient light intensity is 800 lux and the reflectivity of the projection surface is 0.75. First, these ambient light intensities and projection surface reflectivity are encoded as feature vectors. Then, the feature vectors pass through the first hidden layer with 64 neurons and undergo nonlinear transformation through the ReLU activation function. Next, the information is passed to the second hidden layer containing 128 neurons to further extract deep features. Finally, multiple prediction units in the output layer generate different projection parameters, such as adjusting the brightness to 75%, setting the contrast to 1.8, etc. This multi-level feature extraction and mapping process ensures accurate prediction of projection parameters.

[0083] S3 specifically includes: S3.1: Receive the environment feature vector through the feature receiving layer.

[0084] In step S3.1, an environment feature vector is received through a feature receiving layer.

[0085] It should be noted that the feature receiving layer adopts an innovative adaptive feature normalization structure, which can dynamically adjust the normalization parameters according to the statistical characteristics of the input features.

[0086] Specifically, the input environment feature vector is first dimensionally checked and aligned to ensure that the feature dimension matches the network structure. Secondly, batch normalization is used to standardize the feature distribution to an appropriate range, which is crucial for improving the stability of subsequent processing. In addition, the feature receiving layer also integrates a feature integrity verification mechanism. When a feature is missing or abnormal, the feature completion module is triggered to make a reasonable estimate of the missing features through context information.

[0087] For example, when the ambient lighting features are temporarily missing, they are interpolated and supplemented based on historical data and feature values ​​at adjacent moments.

[0088] S3.2: In the fully connected layer, a deep representation corresponding to the environment feature vector is generated through a nonlinear transformation based on a ReLU activation function and a residual connection.

[0089] It should be noted that the fully connected layer adopts a multi-layer structure design, which includes three hidden layers with 256, 512 and 256 neurons respectively. Through this progressive structural design, the abstract representation of features can be gradually extracted.

[0090] It should be noted that each hidden layer is connected to the ReLU activation function, which can not only introduce nonlinear transformation capabilities, but also effectively alleviate the gradient vanishing problem. Secondly, this solution also innovatively adds residual connections between adjacent layers. These jump connections can ensure lossless transmission of information and provide more optimization paths for the network.

[0091] In addition, in order to further improve the effect of feature extraction, this solution also introduces the Dropout mechanism after each fully connected layer, which effectively prevents model overfitting by randomly inactivating some neurons.

[0092] For example, during training, the Dropout ratio is set to 0.3, which means that 30% of the neurons are randomly inactivated in each forward propagation.

[0093] S3.3: In the parameter prediction layer, based on multiple independent prediction units, projection mapping is performed on the deep representation corresponding to the environmental feature vector to generate the projection parameter, wherein one prediction unit corresponds to one projection parameter.

[0094] Specifically, the parameter prediction layer contains multiple parallel prediction branches, each of which is responsible for the prediction of a specific projection parameter.

[0095] It should be noted that each prediction unit uses a special network structure and is optimized according to the characteristics of the prediction parameters. For example, for the prediction unit of the brightness parameter, a three-layer neural network is used, and the Sigmoid activation function is used in the output layer to ensure that the predicted value falls within a reasonable range; for the prediction unit of the contrast parameter, a two-layer structure with residual connections is used, and the value range mapping is achieved through the Tanh activation function.

[0096] In addition, this solution also introduces a parameter constraint mechanism in the prediction layer, and ensures the coordination between prediction parameters by adding regularization terms. In order to improve the robustness of the prediction, the system also designs a prediction confidence evaluation module. When the prediction confidence is lower than the threshold, the backup prediction strategy will be triggered to provide more reliable output by integrating multiple prediction results.

[0097] For example, in actual applications, if the prediction confidence of a parameter is lower than 0.8, multiple prediction models will be used at the same time, and the final prediction value will be obtained by weighted averaging.

[0098] For ease of understanding, a specific scene in a projection environment is taken as an example.

[0099] Assuming that in a conference room environment, the environmental feature vector contains the following information: {ambient light intensity: 800 lux, projection surface reflectivity: 0.75, projection distance: 3 meters, room temperature: 25°C, user preferred brightness level: 4}, the prediction model adopts a five-layer neural network structure, and each layer is designed for its specific purpose.

[0100] First, the input layer contains 32 neurons for receiving the above-mentioned environmental feature vectors; second, the three hidden layers contain 64, 128, and 64 neurons respectively, and each layer uses the ReLU activation function and introduces residual connections. Taking the first hidden layer as an example, it can convert raw features such as ambient light intensity and surface reflectivity into higher-level representations, such as "ambient brightness suitability index"; finally, the output layer is designed with 4 prediction units, corresponding to the brightness (0-100%), contrast (1.0-2.0), color temperature (2000K-6500K), and sharpness (0-10) of the projection. Through this structure, the model can map the input environmental feature vector to the optimal combination of projection parameters, for example: {brightness: 75%, contrast: 1.8, color temperature: 5500K, sharpness: 8}.

[0101] Taking the above conference room scenario as an example, we first construct the loss function L=α L_pred+β L_rate+γ (L1+L2), where L_pred represents the parameter prediction error, L_rate represents the rate constraint, L1 and L2 are regularization terms, and α, β, and γ are weight coefficients. Specifically, when the model predicts that the projection brightness is 75% and the actual optimal brightness is 72%, the L_pred term calculates the 3% deviation; at the same time, the L_rate term evaluates the computational complexity of the model when giving this prediction. For example, if the model uses too many parameters or layers, the L_rate value will increase. In addition, a dynamic batch strategy is adopted during training, with a starting batch size of 32 and an initial learning rate of 0.001. It should be noted that in order to improve the robustness of the model, different interference conditions are added to the training data: for example, by randomly adjusting the ambient light intensity by ±100lux, or changing the reflectivity of the projection surface by ±0.1 to simulate different usage scenarios. Finally, the model parameters are updated through the back-propagation algorithm, and training is stopped when the mean square error on the validation set is less than 0.05.

[0102] S4: performing image deblurring processing and adaptive k-space sampling reconstruction on the projection parameters to generate an optimized projection image.

[0103] This step uses advanced image processing technology to improve projection quality.

[0104] Specifically, the projected image is first encoded into an implicit representation through a neural network encoder based on prior knowledge, which can effectively capture the essential features of the image. Then, a decoder network with a multi-scale structure is used to reconstruct the image, and a spatial attention mechanism is introduced during the reconstruction process to enhance the details of important areas. In addition, a learnable sampling pattern generator is used to determine the optimal sampling trajectory, and high-quality image reconstruction is achieved through a deep convolutional network based on the U-Net architecture.

[0105] For example, take processing a commercial presentation slide with a resolution of 1920×1080 as an example.

[0106] First, a prior knowledge base containing common fonts and graphic features is constructed. The encoder encodes the input image into a 256-dimensional feature vector, paying special attention to the feature expression of the blurred area. For the text area with a font size of 24pt, a higher weight (for example, a weight value of 0.8) is assigned through the spatial attention mechanism, while a lower weight (for example, a weight value of 0.2) is assigned to the solid background area. The decoder gradually restores the image details through 3 upsampling blocks, each of which contains a residual learning unit. Finally, through the adaptive weight fusion module, the deblurred image is fused with the original image in a ratio of 0.7:0.3, which significantly improves the image clarity.

[0107] S4 specifically includes: S4.1: A neural network encoder based on prior knowledge encodes the projection image corresponding to the projection parameters to obtain an implicit representation of the feature encoding of the projection image.

[0108] The encoder adopts an innovative knowledge-driven architecture design that can effectively utilize pre-accumulated prior knowledge of image features.

[0109] Specifically, we first built a knowledge base containing common image feature templates, which cover standard feature representations of various visual elements such as text, graphics, and charts. Secondly, the encoder extracts features through a multi-layer convolutional network structure, and each layer is equipped with an attention mechanism that can adaptively adjust the focus of feature extraction according to content characteristics. In addition, feature consistency constraints are introduced during the encoding process to ensure that the encoding results are highly correlated with prior knowledge.

[0110] For example, when processing an area containing text, the encoder pays special attention to the stroke features and edge information of the font, and generates more accurate feature encoding by matching them with standard font features in the prior knowledge base.

[0111] S4.2: A decoder network based on a multi-scale structure restores the implicit representation of the feature encoding through layer-by-layer feature reconstruction to obtain a reconstructed image corresponding to the projected image.

[0112] In step S4.2, a decoder network based on a multi-scale structure is used for feature reconstruction.

[0113] First, the decoder adopts a hierarchical reconstruction strategy, which consists of four reconstruction stages, each of which is equipped with feature reconstruction modules of different scales.

[0114] It should be noted that the system introduces a spatial attention mechanism in each reconstruction stage, which can dynamically adjust the reconstruction strength according to the importance of the image content.

[0115] In addition, in order to improve the reconstruction effect, the system has also designed an adaptive weight adjustment strategy, which can automatically adjust the reconstruction parameters according to the blurriness of the image area. For example, for high-frequency detail areas such as text areas, higher reconstruction weights will be given to ensure clear restoration of details.

[0116] S4.3: The reconstructed image is fused with the projection image through a residual learning strategy to generate a clear projection image.

[0117] Specifically, a multi-level residual connection network is first established, which can capture the difference information between the original image and the reconstructed image at different scales.

[0118] It should be noted that the system uses an adaptive weighting mechanism for image fusion, and the fusion weights are dynamically adjusted according to the content features and quality indicators of the image area.

[0119] In addition, the system also introduces a local enhancement processing module, which can perform refined fusion optimization for specific areas. For example, when processing high-contrast edge areas, the system will appropriately increase the weight of the original image to retain more edge details.

[0120] S4.4: Based on the learnable sampling pattern generator, an adaptive acquisition operation is performed to determine the optimal sampling trajectory according to the content characteristics of the clear projection image.

[0121] First, the generator adopts a deep reinforcement learning framework to optimize the sampling strategy through continuous interaction with the environment.

[0122] It should be noted that the system has designed a content-aware reward function that can dynamically adjust the sampling parameters according to the quality characteristics of the sampling results.

[0123] Secondly, the system also introduces a sampling efficiency evaluation mechanism to ensure the efficiency of the sampling process by real-time monitoring of sampling quality and computational overhead. For example, for areas with complex textures, the system will automatically increase the sampling density, while for flat areas, the sampling frequency will be appropriately reduced.

[0124] S4.5: Reconstruction model using a deep convolutional network based on the U-Net architecture, preserving image details at different scales through skip connections.

[0125] Specifically, the network adopts a symmetric encoder-decoder structure, which contains five layers of downsampling and five layers of upsampling modules.

[0126] It should be noted that the system sets up jump connections between each layer, which can effectively retain the multi-scale feature information of the image.

[0127] In addition, the system has added a feature selection mechanism to the network, which can automatically identify and retain key image details. For example, during the reconstruction process, the system will pay special attention to the retention of high-frequency information such as edges and textures to ensure the clarity of the reconstructed image.

[0128] S4.6: Based on the Fourier consistency constraint, the optimal sampling trajectory and image details of different scales are subjected to consistency operations to generate the initial optimized projection image.

[0129] First, the system converts the sampled data into frequency domain space and analyzes the frequency characteristics of the image through Fourier transform.

[0130] It should be noted that the system is designed with an adaptive frequency filter, which can dynamically adjust the retention ratio of frequency components according to the image content characteristics.

[0131] Secondly, the system also introduces phase consistency constraints to ensure that the reconstruction results are highly consistent in both the frequency domain and the spatial domain. For example, the system pays special attention to the smooth transition between low-frequency and mid-frequency components while retaining sufficient high-frequency details to achieve the best visual effect.

[0132] S4.7: Use an image registration network to perform precise alignment, perform geometric correction on the initial optimized projection image, and generate the optimized projection image.

[0133] The network adopts an end-to-end learning architecture and can automatically estimate and correct the geometric transformation parameters of the image.

[0134] It should be noted that the system introduces a multi-scale feature matching strategy in the registration process, which improves the accuracy of registration by performing feature matching at different resolution levels.

[0135] In addition, the system is also designed with a registration quality assessment module, which can monitor the registration effect in real time and trigger optimization adjustments when necessary. For example, when it is detected that the registration error in a local area exceeds the threshold, the system will automatically start the local fine registration procedure to ensure the consistency of the overall registration effect.

[0136] Take, for example, projecting a business presentation slide at 1920×1080 resolution.

[0137] It should be noted that the slide contains multiple elements such as text, charts and logos. Due to the influence of projection distance and angle, the original image appears blurred in some areas.

[0138] First, a prior knowledge base is constructed, which contains clear feature templates of common fonts and graphics, such as the stroke features of Arial fonts and the edge features of common charts. Based on this prior knowledge, an encoder-decoder network structure is designed. The encoder encodes the input image into a 256-dimensional feature vector, paying special attention to the feature expression of the blurred area. For example, for the text area with a font size of 24pt in the slide, a higher weight (such as a weight value of 0.8) is given through the spatial attention mechanism, while a lower weight (such as a weight value of 0.2) is given to the solid background area. Then, the decoder network gradually restores the image details through 3 upsampling blocks, each of which contains residual learning units. Taking the text area as an example, the blurred font edges are restored to clear strokes by matching with the prior font features. Finally, the deblurred image is fused with the original image at a ratio of 0.7:0.3 through the adaptive weight fusion module to obtain an output image with a 30% increase in clarity.

[0139] First, K-space sampling optimization is performed. Specifically, the sampling pattern generator will adaptively generate sampling trajectories based on the image content features. For example, for areas containing important text information (such as titles), a dense sampling strategy is adopted with a sampling rate of 90%; while for non-critical areas such as backgrounds, sparse sampling is adopted with a sampling rate of 30%, thereby achieving efficient allocation of computing resources.

[0140] It should be noted that in the reconstruction stage, an improved U-Net structure is adopted, which includes 4 layers of downsampling and 4 layers of upsampling, and the number of feature channels increases from 64 to 512.

[0141] Taking a chart in the slide as an example, firstly, multi-scale features are extracted by downsampling to capture information from detailed texture to overall structure; then, high-frequency details of the original image are retained through jump connections to ensure the sharpness of the edges of the reconstructed chart. In the registration stage, the spatial transformation parameters learned by the registration network include: translation (Δx=5 pixels, Δy=3 pixels), rotation angle (θ=2°) and scaling factor (s=1.02). Through the precise transformation of these parameters, the projected image is ensured to be completely aligned with the target position. In addition, the quality assessment module monitors the peak signal-to-noise ratio (PSNR) of the output image in real time. When the PSNR is lower than 35dB, the parameter fine-tuning mechanism is triggered.

[0142] S5: Collecting projection effect feedback data for the optimized projection image.

[0143] S5 specifically includes: S5.1: The brightness uniformity, color reproduction and contrast of the projected image are collected through optical sensors to form a time series data stream.

[0144] It should be noted that the sensor uses high-precision CMOS imaging elements and is equipped with a professional-grade optical filter group, which can accurately capture various characteristic parameters of the projected image.

[0145] Specifically, sampling is performed every 100 milliseconds, and key indicators such as brightness uniformity, color reproduction and contrast are recorded simultaneously.

[0146] In terms of brightness uniformity measurement, the sensor scans the projection area using a 9-point sampling method to calculate the brightness difference between the center and edge areas. For color reproduction measurement, a standard color card comparison method is used to evaluate color accuracy by calculating the color difference value △E.

[0147] In addition, the system also monitors contrast changes in real time, and ensures the dynamic range of the image by measuring the brightness ratio of the brightest area to the darkest area. These data are organized in chronological order to form a continuous time series data stream, providing basic support for subsequent quality assessment.

[0148] S5.2: Determine a quantitative index of the projection image by a preset evaluation algorithm, wherein the quantitative index includes a peak signal-to-noise ratio and a structural similarity; This scheme adopts an improved peak signal-to-noise ratio (PSNR) calculation method, which takes into account the weighted factors of human visual characteristics and can more accurately reflect the image quality.

[0149] It should be noted that in the PSNR calculation process, an adaptive region division strategy is also introduced, using different evaluation weights for regions of different content types.

[0150] Secondly, the system evaluates the structural fidelity of the image through the structural similarity (SSIM) indicator, which comprehensively evaluates the image quality from three dimensions: brightness, contrast, and structure. In addition, the system also innovatively designs a temporal consistency evaluation module to timely detect and respond to quality fluctuations by analyzing the quality change trend between consecutive frames.

[0151] For example, when it is detected that the PSNR value drops by more than 3dB in a short period of time, the system will automatically trigger the quality alarm mechanism.

[0152] S5.3: Collect user evaluation data through the user interaction interface, where the evaluation data includes satisfaction ratings and feedback information.

[0153] Specifically, the interface features an intuitive rating system, including a 1-5 point satisfaction scale and a detailed feedback input area.

[0154] It should be noted that the system has designed an intelligent feedback collection mechanism that not only records the user's direct ratings, but also captures the user's operational behavior characteristics. For example, when a user frequently adjusts a parameter, the system will automatically pop up a related feedback inquiry window to understand the user's adjustment intentions and expected effects.

[0155] In addition, the system has established a semantic analysis module for user feedback, which can extract key information from users' text descriptions and convert them into quantifiable evaluation indicators.

[0156] S5.4: Record the user's parameter adjustment operation behavior data.

[0157] This solution designs a fine-grained operation recording mechanism that can capture every parameter adjustment action of the user, including information such as the adjusted parameter type, adjustment direction, adjustment amplitude, and adjustment time.

[0158] It should be noted that in the actual processing process, detailed behaviors such as pauses and rollbacks during the user's adjustment process will also be recorded. This information can reflect the user's decision-making process and preference characteristics.

[0159] In addition, an operation mode recognition module is introduced to identify the user's commonly used parameter combinations and adjustment habits by analyzing the user's adjustment sequence. For example, the system can discover the user's preferred parameter setting mode under different lighting environments.

[0160] S5.5: Perform data cleaning and standardization processing on the time series data stream, the quantitative indicators, the user evaluation data and the operation behavior data to obtain standardized projection effect feedback data.

[0161] First, an outlier detection algorithm is used to perform a preliminary screening of the data to remove data points that are obviously abnormal. Secondly, the system converts different types of feedback data into a unified evaluation framework through multi-dimensional standardization processing. In addition, the system has established a data quality assessment mechanism to ensure that the standardized feedback data has a high degree of credibility by calculating indicators such as data integrity, consistency, and timeliness. For example, for user rating data, the system will combine factors such as the time interval between ratings and the context of the operation to determine the effectiveness of the rating, and adjust its weight in the overall evaluation accordingly.

[0162] It should be noted that the system uses a hybrid method based on statistics and machine learning for data cleaning, which takes into account both the statistical characteristics of the data and the time series correlation of the data.

[0163] S6: Performing online updating and optimization on the parameter prediction model according to the projection effect feedback data.

[0164] In steps S5 and S6, the system establishes a complete feedback optimization mechanism, which continuously improves model performance through multi-dimensional data collection and online learning updates.

[0165] For example, the system collects quality data of the projected image every 100ms, including parameters such as brightness uniformity and color reproduction. At the same time, satisfaction scores and feedback information are collected through the user interface. Based on these feedback data, the system uses an incremental learning strategy to gradually optimize the prediction model, and ensures the stability of the update through a dynamic learning rate adjustment mechanism. In addition, the knowledge distillation technology is introduced to transfer the empirical knowledge of the historical model to the new model to further improve the model performance.

[0166] S6 specifically includes: S6.1: Based on the normalized feedback data, an incremental learning strategy is adopted to gradually update the parameters of the parameter prediction model through a small batch gradient descent method. During the parameter updating process, based on a dynamic learning rate adjustment mechanism, the parameter prediction model is adjusted according to the consistency of the feedback data. Based on the knowledge distillation technology, the empirical knowledge of the historical model is transferred to the updated parameter prediction model.

[0167] First, the system uses a small batch gradient descent method to update the model parameters.

[0168] Specifically, 32 samples are selected each time to form a small batch, and the gradient of the loss function is calculated to guide the adjustment direction of the parameters.

[0169] It should be noted that the system uses a dynamic learning rate adjustment mechanism during the update process, which can adaptively adjust the learning rate according to the consistency of the feedback data. For example, when the feedback data of multiple consecutive batches show high consistency, the system will appropriately increase the learning rate to speed up the convergence; conversely, when the feedback data shows large fluctuations, the system will reduce the learning rate to ensure the stability of the update.

[0170] In addition, the system innovatively introduces knowledge distillation technology, and realizes the effective inheritance of historical knowledge by designing a dual architecture of teacher model and student model. Specifically, the existing model is used as the teacher model, and the newly trained model is used as the student model. The degree of knowledge transfer is controlled by the temperature parameter T to ensure that the updated model can maintain the original generalization ability and adapt to new scenario requirements.

[0171] During the incremental learning process, the system also designs a feature importance evaluation mechanism. Specifically, by calculating the contribution of different features to the prediction results, the system can identify key features and focus on optimizing the model parameters related to these features during the update process. It should be noted that the system uses a feature importance analysis method based on SHAP values, which can accurately quantify the influence of each feature. For example, when the system finds that the ambient lighting feature has a significant impact on the prediction results, it will increase the weight of the feature in the model update accordingly. In addition, the system has also established a verification mechanism for parameter updates, which evaluates the effectiveness of the update through cross-validation to ensure the steady improvement of model performance.

[0172] S6.2: By comparing the projection qualities of the projection images before and after the parameter prediction model is updated, and based on the comparison result, verifying the performance improvement of the parameter prediction model, and deciding whether to accept the update based on the comparison result.

[0173] In step S6.2, the system verifies the effect of the model update through a comprehensive performance evaluation mechanism. First, the system establishes a multi-dimensional evaluation index system, including key indicators such as prediction accuracy, response time, and resource consumption. It should be noted that the evaluation process adopts a sliding window mechanism to accurately reflect the changing trend of model performance by comparing the performance indicators in the continuous time window. Specifically, the system will calculate the rate of change of prediction error before and after the update, the degree of improvement of calculation delay, and the optimization of resource utilization. In addition, the system has also designed an adaptive performance threshold mechanism to dynamically adjust the acceptance criteria according to the requirements of different application scenarios. For example, in scenarios with high real-time requirements, the system will pay more attention to the improvement of response time.

[0174] Based on performance verification, the system implements an intelligent update decision-making mechanism. Specifically, the system calculates the updated benefit score by comprehensively analyzing various performance indicators. It should be noted that the benefit score is weighted summed, and the weight coefficient will be dynamically adjusted according to the characteristics of the application scenario. For example, in scenarios with high image quality requirements, the prediction accuracy indicator will be given a higher weight. When the benefit score exceeds the preset threshold, the system will accept the update; otherwise, it will trigger the rollback mechanism and restore the model to the state before the update. In addition, the system also maintains the historical record of the model version and supports version switching when necessary to ensure that the system always runs on the optimal model version.

[0175] The present application also provides an intelligent adaptive projection lamp. Figure 2 As shown, the intelligent adaptive projection lamp 100 may include a processor 110 and a memory 120, wherein the memory 120 is used to store a computer program, and when the processor 110 executes the computer program, the above-mentioned projection control method is implemented.

[0176] It should be noted that the projection control method of the embodiment of the present application can also be implemented as a computer device, a computer-readable storage medium or a computer program product. Among them, the computer device includes a processor and a memory, the memory stores a computer program, and the processor executes the program to implement the various steps of the above-mentioned projection control method. The computer-readable storage medium stores computer instructions, which are used to enable the computer to execute the above-mentioned projection control method. The computer instructions contained in the computer program product implement the above-mentioned method steps when executed by the processor.

[0177] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. For the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0178] The above description is only a preferred embodiment of the present application and is not intended to limit the protection scope of the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application are included in the protection scope of the present application.

[0179] The system of the embodiments of the present disclosure can execute the method provided by the embodiments of the present disclosure, and the implementation principles are similar. The actions performed by each module in the system of each embodiment of the present disclosure correspond to the steps in the method of each embodiment of the present disclosure. For the detailed functional description of each module of the system, please refer to the description in the corresponding method shown in the previous text, which will not be repeated here.

[0180] The above is only an optional implementation method for some implementation scenarios of the present disclosure. It should be pointed out that for ordinary technicians in this technical field, without departing from the technical concept of the scheme of the present disclosure, other similar implementation methods based on the technical ideas of the present disclosure are also within the protection scope of the embodiments of the present disclosure.

Claims

1. An intelligent adaptive projection control method, characterized in that: include: Collecting multi-source environmental data, wherein the multi-source environmental data includes at least one of environmental lighting data, projection surface characteristic data, and user operation data; Preprocessing, feature extraction and feature fusion of the multi-source environmental data to generate an environmental feature vector; Projecting the environment feature vector using a parameter prediction model based on rate-distortion optimization to generate projection parameters; Performing image deblurring processing and adaptive k-space sampling reconstruction on the projection parameters to generate an optimized projection image; Collecting projection effect feedback data for the optimized projection image; The parameter prediction model is updated and optimized online according to the projection effect feedback data.

2. The projection control method according to claim 1, characterized in that: The collecting of multi-source environmental data includes: Collecting the ambient light data by an ambient light sensor, wherein the ambient light data includes ambient light intensity data; Acquiring the projection surface characteristic data through a projection surface detector, wherein the projection surface characteristic data includes reflectivity and texture feature data of the projection surface; Through the user interaction interface, the user's adjustment operation data on the projection parameters is collected to obtain the user operation data.

3. The projection control method according to claim 1, characterized in that: The preprocessing, feature extraction and feature fusion of the multi-source environmental data to generate an environmental feature vector includes: Performing noise filtering and outlier detection processing on the ambient light data to obtain cleaned ambient light data; Performing geometric correction and texture feature extraction on the projection surface characteristic data to obtain standardized surface characteristic data; Performing time sequence alignment and behavior pattern analysis on the user operation data to obtain standardized user operation data; A convolutional neural network is used to extract standardized surface feature data to obtain image features; Use recurrent neural networks to analyze normalized user operation data and obtain time series features; The image features, the time series features and the cleaned ambient light data are fused through an attention mechanism, and the fused data is the ambient feature vector.

4. The projection control method according to claim 1, characterized in that: The parameter prediction model is a multi-layer deep neural network, and the multi-layer deep neural network includes a feature receiving layer, a multi-layer fully connected layer based on a rectified linear unit activation function, and a parameter prediction layer; The step of performing projection mapping on the environmental feature vector to generate projection parameters by using a parameter prediction model based on rate distortion optimization includes: Receiving the environment feature vector through the feature receiving layer; In the fully connected layer, a deep representation corresponding to the environment feature vector is generated through a nonlinear transformation based on a ReLU activation function and a residual connection; In the parameter prediction layer, projection mapping is performed on the deep representation corresponding to the environmental feature vector based on multiple independent prediction units to generate the projection parameters, wherein one prediction unit corresponds to one projection parameter.

5. The projection control method according to claim 1 or 4, characterized in that: The parameter prediction model based on rate-distortion optimization is obtained in the following manner; Construct training data sets and validation sets based on pre-collected historical environmental data and corresponding optimal projection parameters; Constructing a loss function based on rate-distortion theory, wherein the loss function includes a parameter prediction error term and a rate constraint term; Using an improved stochastic gradient descent algorithm, based on the loss function and the training data set, to train the preset neural network model, and during the training process, using an adaptive moment estimation optimizer and a dynamic learning rate adjustment strategy; The model effect is evaluated based on the validation set, and hyperparameters including batch size and learning rate are dynamically adjusted to generate the parameter prediction model.

6. The projection control method according to claim 1, characterized in that: The performing image deblurring processing and adaptive k-space sampling reconstruction on the projection parameters to generate an optimized projection image comprises: A neural network encoder based on prior knowledge encodes the projection image corresponding to the projection parameters to obtain an implicit representation of the feature encoding of the projection image; A decoder network based on a multi-scale structure restores the implicit representation of the feature encoding by layer-by-layer feature reconstruction to obtain a reconstructed image corresponding to the projected image; The reconstructed image is fused with the projection image through a residual learning strategy to generate a clear projection image; Based on a learnable sampling pattern generator, an adaptive acquisition operation is performed to determine the optimal sampling trajectory according to the content characteristics of the clear projection image; A reconstruction model based on a deep convolutional network of the U-Net architecture is used to preserve image details of different scales through skip connections. Based on Fourier consistency constraints, the optimal sampling trajectory and image details of different scales are operated consistently to generate the initial optimized projection image; An image registration network is used to perform precise alignment, and geometric correction is performed on the initial optimized projection image to generate the optimized projection image.

7. The projection control method according to claim 6, characterized in that: In the step of obtaining a reconstructed image corresponding to the projected image, the projection control method further includes: adopting a spatial attention mechanism to enhance the detail performance of important areas, and adding an adaptive weight adjustment strategy to dynamically adjust the deblurring intensity according to the blur degree of the image area.

8. The projection control method according to claim 1, characterized in that: The collecting of projection effect feedback data for the optimized projection image includes: The brightness uniformity, color reproduction and contrast of the projected image are collected through optical sensors to form a time-series data stream; Determining quantitative indicators of the projection image by a preset evaluation algorithm, wherein the quantitative indicators include peak signal-to-noise ratio and structural similarity; Collecting user evaluation data through a user interaction interface, wherein the evaluation data includes satisfaction ratings and feedback information; Record the user's parameter adjustment operation behavior data; The time series data stream, the quantitative index, the user evaluation data and the operation behavior data are cleaned and standardized to obtain standardized projection effect feedback data.

9. The projection control method according to claim 1 or 8, characterized in that: The online updating and optimization of the parameter prediction model according to the projection effect feedback data includes: Based on the normalized feedback data, an incremental learning strategy is adopted to gradually update the parameters of the parameter prediction model through a small batch gradient descent method. In the parameter updating process, the parameter prediction model is adjusted according to the consistency of the feedback data based on a dynamic learning rate adjustment mechanism, and the empirical knowledge of the historical model is transferred to the updated parameter prediction model based on the knowledge distillation technology; By comparing the projection qualities of the projection images before and after the parameter prediction model is updated, and based on the comparison result, verifying the performance improvement of the parameter prediction model, and deciding whether to accept the update based on the comparison result.

10. An intelligent adaptive projection lamp, characterized in that: The method comprises a processor and a memory, wherein the memory is used to store a computer program, and when the processor executes the computer program, the projection control method according to any one of claims 1 to 9 is implemented.

Images