Table and chair remote management method based on smart home platform

Abstract

The invention provides a table and chair remote management method based on a smart home platform, and the method comprises the steps: building a table and chair three-dimensional model, and obtaining key structure points and outer contour line coordinate information of tables and chairs through the analysis of the three-dimensional model; a computer vision algorithm is adopted to analyze the home space image, main vision elements in the home space are extracted, the space relation and vision coordination between the vision elements are judged, and a home space vision feature graph is generated; dynamically adjusting the size, the shape and the color of the table and chair hidden contour according to the calculated optimal position and the shape parameter, so that the table and chair hidden contour is coordinated with the surrounding environment, and generating an initial design scheme of the hidden contour; the optimized hidden contour design scheme parameters are transmitted to an intelligent control system, training is carried out through a machine learning algorithm in combination with historical design scheme data, and dynamic adaptation of the table and chair hidden contour and the surrounding environment is achieved.

Description

Technical Field

[0001] The present invention relates to the field of information technology, and particularly to a method for remotely managing tables and chairs based on a smart home platform. Background Art

[0002] In the smart home environment, the core technical problem faced by the design of tables and chairs is how to achieve visual integration with the surrounding environment while ensuring functionality. In specific business scenarios, traditional table and chair designs often focus on functionality and structural stability, while ignoring the coordination with the smart home environment, resulting in a visually obtrusive feeling. In addition, due to the diversity and complexity of the home environment, after the tables and chairs are hidden in different positions, the coordination effect of their contour lines with the surrounding environment will also vary. Therefore, how to adjust the shape, size, color and other parameters of the hidden contour in real time according to the specific position of the tables and chairs to dynamically adapt to different environments is a technical difficulty that needs to be overcome urgently. At the same time, the design of the hidden contour also needs to consider the morphological characteristics and structural features of the tables and chairs, and while ensuring the hiding effect, it is also necessary to ensure that the normal use functions of the tables and chairs are not affected. Summary of the Invention

[0003] The present invention provides a method for remotely managing tables and chairs based on a smart home platform, mainly including:

[0004] Obtain the surrounding environment information of the position where the tables and chairs are located, including the color and material attribute data of the wall and floor environmental elements, and determine the initial color and material parameters of the hidden contour of the tables and chairs according to the obtained environmental color and material attribute data, including hue, saturation, lightness, and surface texture characteristics;

[0005] Establish a three-dimensional model of the tables and chairs, and obtain the key structure points and outer contour line coordinate information of the tables and chairs by analyzing the three-dimensional model;

[0006] Use computer vision algorithms to analyze the home space image, extract the main visual elements in the home space, judge the spatial relationship and visual coordination between the visual elements, and generate a visual feature map of the home space;

[0007] Match the key structure points and outer contour line coordinate information of the three-dimensional model of the tables and chairs with the visual feature map of the home space, and calculate the optimal position and shape parameters of the hidden contour of the tables and chairs in the home space through a spatial mapping algorithm;

[0008] Dynamically adjust the size, shape and color of the hidden contour of the tables and chairs according to the calculated optimal position and shape parameters to make it coordinated with the surrounding environment, and generate an initial design scheme for the hidden contour;

[0009] Use physical simulation methods to simulate the movement process of the table and chair, determine whether the hidden contour design affects the normal use function of the table and chair, optimize the hidden contour design according to the simulation results, and generate an optimized hidden contour design scheme;

[0010] Transfer the parameters of the optimized hidden contour design scheme to the intelligent control system, combine with the historical design scheme data, and train through machine learning algorithms to achieve the dynamic adaptation of the hidden contour of the table and chair to the surrounding environment.

[0011] The technical solution provided by the embodiment of the present invention may include the following beneficial effects:

[0012] The present invention discloses a remote management method for tables and chairs based on a smart home platform. The method first obtains the color and material data of the environment around the table and chair, establishes a three-dimensional model of the table and chair and extracts key structural information. Then, it analyzes the home space image through computer vision algorithms to generate a visual feature map. Matches the table and chair model with the visual feature map, calculates the optimal position and shape parameters of the hidden contour. Dynamically adjusts the design of the hidden contour according to the calculation results, and optimizes the design scheme through physical simulation. Finally, the present invention transfers the optimized design scheme to the intelligent control system to achieve the dynamic adaptation of the hidden contour to the environment through machine learning. The technical effect of the present invention is that it can automatically generate and optimize the hidden contour design of the table and chair according to the surrounding environment, making the table and chair better integrate into the home space, improving the visual coordination and beauty of the space. At the same time, it ensures the normal use function of the table and chair, and realizes the dynamic balance between the appearance and function of the table and chair in the smart home environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 It is a flowchart of a remote management method for tables and chairs based on a smart home platform of the present invention.

[0014] Figure 2 It is a schematic diagram of a remote management method for tables and chairs based on a smart home platform of the present invention.

[0015] Figure 3 It is another schematic diagram of a remote management method for tables and chairs based on a smart home platform of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0016] In order to enable those skilled in the art to better understand the technical solutions in this specification, the following will clearly and completely describe the technical solutions in the embodiments of this specification with reference to the accompanying drawings in the embodiments of this specification. Obviously, the described embodiments are only a part of the embodiments of this specification, rather than all the embodiments. Based on the embodiments in this specification, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of this specification.

[0017] As Figures 1-3 , a method for remotely managing tables and chairs based on a smart home platform in this embodiment may specifically include:

[0018] S101. Obtain the environmental information around the location where the tables and chairs are located, including the color and material attribute data of the wall and floor environmental elements. According to the obtained environmental color and material attribute data, determine the initial color and material parameters of the hidden contour of the tables and chairs, including hue, saturation, lightness, and surface texture characteristics.

[0019] Use a color sensor array to scan the wall area, obtain the main wall color value from the obtained wall color distribution data, and obtain the wall depth distribution map through a material texture analyzer; measure the wall reflection degree using a ray tracing model according to the main wall color value, obtain the indoor light intensity value from the photometric calculation unit, and perform quantization processing on the light intensity value to obtain the light intensity distribution map; process the light intensity distribution map and the wall depth distribution map through convolution operation to obtain the overall environmental brightness distribution map, and establish a three-dimensional space brightness model for the overall environmental brightness distribution map; use a depth camera to perform three-dimensional scanning and modeling of the table and chair surface, and calculate the lighting parameters of the table and chair contour area according to the three-dimensional space brightness model to obtain the lighting parameter set of the table and chair contour area.

[0020] Specifically, for the wall environment around the location of the tables and chairs, a color sensor array is used to scan the wall area. From the obtained wall environment color distribution data, the main wall color tone value is obtained through color space quantization processing, and then the surface depth distribution map is obtained by scanning the wall surface with a material texture analyzer. According to the obtained main wall color tone value, a ray tracing model is used to measure the wall reflection degree, and the indoor light intensity value is obtained from the photometric calculation unit, and the ambient illuminance is quantized to obtain the light intensity distribution map. According to the light intensity distribution map and the wall depth distribution map, the overall ambient brightness distribution map is obtained through convolution operation, and a three-dimensional space brightness model is established for the overall ambient brightness distribution map. A depth camera is used to perform three-dimensional scanning and modeling of the table and chair surfaces, and the reflected light data of the table and chair surfaces is obtained through the ray tracing model. Combining with the three-dimensional space brightness model, the illumination parameters of the table and chair contour areas are calculated. For the illumination parameters of the table and chair contour areas, a color space mapping algorithm is used for hue quantization calculation, and the hue parameter with the highest matching degree is extracted from the preset color database. A material texture analyzer is used to scan the table and chair surfaces to obtain the surface texture feature vectors, and the principal component analysis is performed on the texture feature vectors to obtain the main texture feature values. According to the hue parameters and the main texture feature values, the texture parameter combination with the highest similarity is retrieved from the preset material database to generate the initial color and material parameter set of the table and chair hidden contours. During the process of obtaining the table and chair environment information, the color sensor array scans the wall through the principle of photoelectric conversion. Each sensor unit collects spectral data, and the spectral data is converted into digital signals through color space quantization processing. For example, a certain sensor unit collects the visible spectrum with a wavelength range between 380 and 780 nanometers, and the converted RGB three-channel values are (200, 150, 100). Through color space conversion, the hue value in the HSV color space can be obtained as 25 degrees, the saturation is 0.5, and the lightness is 0.78. The material texture analyzer uses the structured light projection method to project a specific stripe pattern onto the wall, and the surface depth distribution is obtained by analyzing the stripe deformation degree. Suppose 20 equally spaced parallel stripes are projected within a 1-square-meter wall area, each stripe with a width of 2 millimeters. The stripe deformation image is collected by a high-resolution camera, the position offset of the stripe center line is extracted, and the depth values are calculated to be distributed within the range of 0 - 255. The ray tracing model is based on the principle of geometric optics and uses the Monte Carlo method to simulate the light propagation path. In a 5-meter × 4-meter × 3-meter indoor space, 500 light emission points are set, and each emission point emits 1000 rays. The intersection coordinates and incident angles of the rays and the wall are recorded. The photometric calculation unit measures that the indoor light intensity varies between 200 and 800 lux, and the illuminance values of each point on the wall are calculated by combining the light incident angles. By registering the three-dimensional point cloud data of the tables and chairs collected by the depth camera, a surface mesh model of the tables and chairs is established. The mesh model contains 10,000 vertices, and each vertex records the spatial coordinates and normal vector information.Project the ambient light data onto the surface of the grid model, and calculate the lighting parameters at each vertex, including the diffuse reflection coefficient and the specular reflection coefficient. Extract the texture features of the table and chair surfaces, and use the gray-level co-occurrence matrix method to calculate the texture statistical features. In a 5×5 pixel window, statistically analyze the gray value relationship of pixel pairs, and calculate the eigenvalue such as contrast, correlation, energy and homogeneity. Reduce the 16-dimensional texture feature vector to a 4-dimensional principal component space through principal component analysis, retaining more than 90% of the feature information. Finally, retrieve the best matching item in the preset database according to the hue parameter and the texture eigenvalue. The database contains parameter templates of 1,000 common materials, and each template contains three color parameters of hue, saturation, and lightness, and three texture parameters of contrast, directionality, and roughness. Perform similarity matching by calculating the Euclidean distance, and select the parameter combination with the smallest distance as the initial parameter of the hidden contour of the table and chair.

[0021] S102. Establish a three-dimensional model of the table and chair, and obtain the key structure points and the coordinate information of the outer contour line of the table and chair by analyzing the three-dimensional model.

[0022] Use a laser scanner to obtain the point cloud data set of the table and chair surface. The point cloud data set is fused with multi-angle data through a point cloud registration algorithm to obtain a complete point cloud model; according to the complete point cloud model, use the Poisson reconstruction algorithm to perform triangular mesh division, and optimize the grid vertex coordinates through the Laplacian smoothing method to obtain a smooth grid model; for the smooth grid model, use the Gaussian curvature calculation formula to extract the grid vertex curvature value, and identify the feature point set by setting the curvature threshold; according to the feature point set, use the Sobel operator to perform edge detection to obtain a binary image, and extract the key inflection point coordinate sequence from the binary image.

[0023] Specifically, a laser scanner is used to obtain the point cloud data of the table and chair surfaces. The spatial coordinate information is extracted from the original point cloud data. The multi-angle scan data is registered and fused through a point cloud registration algorithm to obtain a complete point cloud data set of the table and chair. A mesh topology structure is constructed based on the point cloud data set. The Poisson reconstruction algorithm is used to divide the point cloud into triangular meshes. The vertex coordinates of the mesh are optimized by the Laplacian smoothing method to obtain a smooth and continuous mesh model of the table and chair surfaces. For the vertices of the mesh model, the Gaussian curvature calculation formula is used to extract the curvature values. By setting a curvature threshold, a set of feature points is identified, and a distribution map of the characteristic regions on the table and chair surfaces is obtained. According to the distribution map of the characteristic regions, the Sobel operator is used for edge detection, and the detection result is binarized. Continuous edge segments are extracted through morphological operations. For the edge segments, the Douglas-Peucker algorithm is used to extract the coordinates of the key inflection points, and the inflection point sequence is linearly fitted by the least squares method. According to the fitted straight line segments, the cubic spline interpolation method is used to generate a continuous and smooth outer contour line, and the contour line coordinate sequence is obtained through uniform sampling. The coordinate sequence of the contour line is subjected to a spatial coordinate transformation to establish the corresponding relationship of the local coordinate system, and the coordinate data is stored in a structured database. In the 3D modeling of the table and chair, the laser scanner emits a laser beam with a wavelength of 905 nanometers, and the distance information of the spatial points is obtained by measuring the laser reflection time. A standard office chair is scanned 360 degrees, and a sample is taken every 15 degrees to obtain 24 groups of point cloud data, each group containing approximately 50,000 spatial point coordinates. These original point cloud data are registered through the iterative closest point algorithm, and the registration error is controlled within 0.5 millimeters, and finally a complete point cloud data set containing approximately 1 million points is formed. In the process of point cloud reconstruction, the Poisson reconstruction algorithm first calculates the normal vector field of the point cloud and estimates the local tangent plane for each point. By solving the Poisson equation ▽2χ = ▽·V, where V is the vector field and χ is the indicator function, an implicit function representation of the object surface is obtained. The algorithm sets the octree depth to 8, and the reconstructed mesh model contains approximately 200,000 triangular patches. The mesh is subjected to Laplacian smoothing for 5 iterations. In each iteration, the vertex coordinates are updated according to the weighted average of the surrounding vertices, and the weight is inversely proportional to the edge length. The surface curvature analysis uses the Gaussian curvature calculation formula K = κ1κ2, where κ1 and κ2 are the principal curvatures. For each vertex on the mesh model, a quadratic surface z = ax2 + by2 + cxy is fitted within the 3-ring neighborhood, and the coefficients a, b, and c are obtained by solving the least squares system of equations, and then the Gaussian curvature value of this point is calculated. The curvature threshold is set to 0.05, and the points with the absolute value of the curvature greater than the threshold are extracted as feature points, obtaining a set of approximately 5,000 feature points. The edge detection uses a 3×3 Sobel operator to calculate the gradients in the horizontal and vertical directions of the feature point distribution map. The gradient amplitude threshold is set to 128, and the points greater than the threshold are marked as edge points. The noise is removed through morphological opening operations to obtain continuous edge segments.Apply the Douglas - Peucker algorithm to simplify the edge segments, set the distance threshold to 2 mm, and retain approximately 200 key inflection points. Segment the sequence of inflection points. When the distance between adjacent inflection points is greater than 10 mm, it is divided into a new line segment. Use the least - squares method to fit the straight - line equation ax + by + c = 0, and control the fitting error within 1 mm. Generate a smooth contour line through cubic spline interpolation, and set the interpolation point spacing to 5 mm to ensure the continuity of the first - order and second - order derivatives of the curve at the nodes. Finally, establish a local coordinate system with the center of the seat as the origin, and store the spatial coordinates of the contour - line sampling points in the structured database after coordinate transformation.

[0024] S103. Use computer vision algorithms to analyze the home - space image, extract the main visual elements in the home space, judge the spatial relationship and visual coordination between the visual elements, and generate a home - space visual feature map.

[0025] Use a depth camera to obtain the original image of the home space, process the original image through a Gaussian filter and gamma correction to obtain an enhanced image; extract the gray - level co - occurrence matrix texture features, color - moment color features, and gradient - operator edge features from the enhanced image to obtain a feature vector, perform region - growing segmentation on the feature vector to obtain a visual - element labeling map; calculate the centroid coordinates of the visual elements for the visual - element labeling map, construct a spatial - distance matrix using the Euclidean - distance method, calculate the color similarity of the visual elements through the intersection of color histograms to obtain a visual - coordination matrix; use the spectral - clustering method to hierarchically divide the visual elements according to the visual - coordination matrix, construct a spatial - division structure using the quadtree method, and calculate and generate a visual - attention distribution map using brightness weights to obtain a home - space visual feature map.

[0026] Specifically, a depth camera is used to obtain images of the home space. The original images are denoised by a Gaussian filter. Gamma correction and histogram equalization are performed on the denoised images to obtain enhanced images. Feature extraction is carried out on the enhanced images. Texture features are calculated through the gray-level co-occurrence matrix, color features are extracted using the color moment method, and edge features are calculated via the gradient operator to obtain an image feature vector. For the image feature vector, a similarity threshold is set for region growing segmentation. A convolutional neural network is used to identify visual elements such as tables, chairs, cabinets, sofas, and walls in the home space to obtain a visual element labeling map. The centroid coordinates of the visual elements are calculated based on the visual element labeling map. A spatial distance matrix is constructed by the Euclidean distance method, and a spatial adjacency relationship is established for point pairs with a distance less than the preset threshold. For the visual element regions, color features are extracted using the color histogram, the color similarity is obtained by calculating the histogram intersection, and the visual coordination matrix is obtained in combination with the spatial adjacency relationship. Based on the visual coordination matrix, the spectral clustering method is used to hierarchically divide the visual elements, and the spatial arrangement rule is determined by calculating the main direction. For the result of the hierarchical division of the visual elements, a spatial division structure is constructed using the quadtree method, and a visual attention distribution map is generated by calculating the brightness weight to obtain the visual feature map of the home space. The depth camera uses the principle of structured light projection to obtain the RGB image and depth map of the home scene. The image resolution is 1920×1080 pixels, and the depth accuracy is ±1 mm. The RGB image is filtered by a 5×5 Gaussian kernel with a standard deviation set to 1.5 to remove random noise. Gamma correction is performed on the filtered image with the gamma value set to 0.8 to enhance the details in the dark areas. The histogram equalization method is used to remap the image gray values to the range of 0-255 to enhance the image contrast. In the feature extraction stage, the displacement vectors of the gray-level co-occurrence matrix are set to (0,1), (1,0), and (1,1), and the co-occurrence frequencies of pixel pairs at the gray levels of 0-255 are counted to calculate 4 texture features including contrast, correlation, energy, and homogeneity. The color moments of the image are calculated in the RGB space, including the mean moment, standard deviation moment, and skewness moment, to obtain a 9-dimensional color feature vector. The Sobel operator is used to calculate the horizontal and vertical gradients of the image, and an 8-dimensional edge feature vector is constructed through the gradient magnitude and direction. The region growing segmentation uses the central region of the image as the seed point, and the growth criterion is set as the Euclidean distance of the feature vector being less than the threshold of 10, and adjacent pixels are iteratively merged. The convolutional neural network uses a pre-trained model to classify the segmented regions, and the confidence threshold is set to 0.85 to obtain the visual element regions labeled with class labels. The centroid coordinates of each visual element region are calculated to construct an N×N distance matrix, where N is the number of visual elements. The spatial adjacency threshold is set to 0.2 times the length of the image diagonal, and an adjacency relationship is established when the centroid distance between two visual elements is less than the threshold. A 32-dimensional color histogram is calculated for each visual element region, and the color similarity between each pair is calculated by the histogram intersection, and the visual coordination matrix is obtained in combination with the adjacency relationship.The spectral clustering method is used to decompose the coordination matrix, and the first three eigenvectors are taken to construct the feature space. The visual elements are divided into three levels by the k-means algorithm. Calculate the main direction of the visual elements within each level, detect the line features through the Hough transform, and extract the spatial arrangement rules. Finally, according to the quadtree method, the image is recursively divided into 16×16 grids, and the average brightness value of each grid is calculated as the weight to generate a visual attention distribution map with a resolution of 256×256. The final visual feature map is obtained by superimposing the hierarchical division results.

[0027] S104. Match the key structural points and the coordinate information of the outer contour line of the three-dimensional model of the table and chair with the visual feature map of the home space, and calculate the optimal position and shape parameters of the hidden contour of the table and chair in the home space through the spatial mapping algorithm.

[0028] The key structural points of the three-dimensional model of the table and chair are projected onto the two-dimensional plane by the perspective projection method, and the standardized plane coordinate point set is obtained by calculating through the homography matrix; according to the standardized plane coordinate point set, the surf feature point extraction algorithm is used to obtain the feature descriptors, and the feature point correspondence set is obtained through the nearest neighbor matching method; a least squares optimization objective function is established for the feature point correspondence set, and the gradient descent method is used to iteratively optimize the position coordinates of the table and chair to obtain the initial placement position of the table and chair; according to the initial placement position of the table and chair, the contour is fitted by the cubic Bezier curve, the interval ratio of the curve control points and the tangent direction constraint are set, and the curvature value of the sampling points is calculated to obtain the contour shape feature vector.

[0029] Specifically, perform coordinate system transformation on the key structural points and outer contour line coordinates of the table and chair three-dimensional model. Use the perspective projection method to project the three-dimensional space coordinates onto a two-dimensional plane, and perform affine transformation on the projected coordinates through the calculation of the homography matrix to obtain a set of standardized plane coordinate points. According to the visual feature map of the home space, use the SURF feature point extraction algorithm to obtain feature descriptors, and match the contour points of the table and chair with the space feature points through the nearest neighbor matching method to obtain a set of corresponding relationships of feature points. According to the set of corresponding relationships of feature points, establish a least squares optimization objective function, and use the gradient descent method to iteratively optimize the position coordinates and size ratios of the table and chair to obtain the initial placement position of the table and chair. For the initial placement position of the table and chair, perform contour fitting through cubic Bezier curves, set the interval ratio of curve control points and the tangent direction constraint to obtain a continuous and smooth contour curve. Calculate the discrete sampling point sequence according to the smooth contour curve, obtain the curvature values of the sampling points through the curvature calculation formula, and perform normalization processing on the curvature values to obtain the contour shape feature vector. Use the homogeneous coordinate transformation method to calculate the position and direction of the contour curve in the actual space, and determine the actual size parameters of the table and chair through scale transformation to obtain the space mapping result. Package the data of the space mapping result, organize the position coordinates, size ratios, and curve parameters according to the preset data structure to generate a standard data packet, and transmit it to the smart home data center. During the space mapping process of the table and chair, the projection from three-dimensional coordinates to a two-dimensional plane uses a perspective projection matrix, and the matrix elements include the camera's internal parameters and external parameters. For a camera with a focal length of 50 mm and the principal point coordinates (320, 240), the projection matrix P can be represented as a 3×4 matrix, where the internal parameter matrix K contains the focal length and principal point information, and the external parameter matrix [R|t] describes the camera's pose. Perform affine correction on the projected coordinates through the homography matrix H to eliminate perspective distortion and keep parallel lines parallel after projection. The feature point matching uses the SURF algorithm to extract feature descriptors, and each feature point contains a 64-dimensional vector to describe its local features. Extract 1000 feature points from the visual feature map of the home space, and calculate the gradient direction histogram for each feature point. Use the nearest neighbor matching method, set the distance ratio threshold to 0.7, and consider the match valid when the ratio of the nearest neighbor distance to the second nearest neighbor distance is less than the threshold. Usually, 200 - 300 pairs of valid matching point pairs can be obtained. In the least squares optimization, the objective function J is composed of the sum of the squares of the Euclidean distances between the matching point pairs, J = Σ||pi - qi||2, where pi and qi are the coordinates of the corresponding feature point pairs. Use the gradient descent method to optimize the position parameters x, y, and the size ratio s, set the learning rate to 0.01, and stop the iteration when the value of the objective function drops less than 0.001. Usually, 50 - 100 iterations are required to reach convergence. The contour curve is described by a cubic Bezier curve, and the curve equation is P(t) = (1 - t)3P0 + 3t(1 - t)2P1 + 3t2(1 - t)P2 + t3P3, where the parameter t varies in the interval [0, 1].The selection of control points P0, P1, P2, and P3 is based on the positions of the contour inflection points. The distance between adjacent control points is set to 0.25 - 0.3 times the total length of the contour line, and the tangent direction is consistent with the local trend of the contour line. 100 sampling points on the contour curve are obtained by uniform sampling, and the curvature κ = |x'y” - y'x”| / (x'² + y'²)^(3 / 2) of each point is calculated, where x' and y' are the first-order derivatives of the parameter t, and x” and y” are the second-order derivatives. The curvature values are normalized and mapped to the [0, 1] interval to obtain the shape feature vector. The actual space mapping uses a 4×4 homogeneous transformation matrix, which includes the rotation matrix R and the translation vector t. The scale is determined according to the room size, and the normalized coordinates are converted into actual sizes. Finally, the calculation results are packaged in JSON format. The data includes the position coordinates (x, y, z), the direction angles (α, β, γ), the size ratio s, and the coordinates of 100 sampling points on the contour curve. The calculation process takes about 200 milliseconds on an ordinary computer, and the positioning accuracy is better than 5 cm.

[0030] S105. According to the calculated optimal position and shape parameters, dynamically adjust the size, shape, and color of the hidden contour of the table and chair to make it harmonious with the surrounding environment, and generate the initial design scheme of the hidden contour.

[0031] The size of the table and chair is scaled using the affine transformation method according to the contour of the table and chair. The contour of the table and chair is interpolated by cubic spline to obtain a smooth boundary curve; the main color tone is obtained from the smooth boundary curve using the color clustering method, and a linear gradient color band is obtained through hue saturation mapping; spatial interpolation is performed on the linear gradient color band using the polynomial radial basis function, and the color distribution map of the transition region is obtained by constraining the color difference threshold; the color distribution map of the transition region is superimposed with the environmental texture using the weight fusion method, and a hidden contour with a decreasing edge transparency is obtained through the alpha blending function.

[0032] Specifically, according to the optimal position parameters, the affine transformation method is used to scale the size and deform the shape of the hidden contour of the table and chair. The contour boundary is smoothed by cubic spline interpolation, and a sequence of boundary points is extracted from the smoothed boundary curve at fixed arc length intervals. For the color distribution in the environmental space, the color clustering method is used to extract the dominant color. The reference color of the hidden contour is calculated by hue-saturation mapping. A linear gradient color band is generated according to the brightness gradient. According to the gradient color band, the polynomial radial basis function is used to perform spatial interpolation on the regional color. The color distribution in the transition region is optimized by the chromatic aberration threshold constraint to obtain the color transition map. The color transition map is subjected to Gaussian smoothing, and the brightness and darkness distribution of the color is adjusted by the brightness histogram equalization method to obtain the smoothed color map. For the environmental texture features, the Markov random field method is used to generate a texture similar to the environment, and the texture coordination degree is calculated by texture feature vector matching. According to the distribution of the texture coordination degree, the weight fusion method is used to superimpose the texture pattern and the smoothed color map, and the transparency of the edge region is decreased by the alpha blending function. The contrast of the fusion result is adjusted, and the bilateral filtering method is used to smoothly transition the material boundary to obtain the initial design scheme of the hidden contour. During the dynamic adjustment of the table and chair contour, the affine transformation uses a 3×3 transformation matrix, which includes translation, rotation, and scaling parameters. For the contour adjustment of the office chair, the scaling coefficient is set to 0.8 - 1.2, and the rotation angle varies within the range of ±15 degrees. The cubic spline interpolation uses the basis function B(t)=(1 - t)^3 / 6. By setting the control point spacing to 20 mm, 200 boundary points are uniformly sampled on the contour curve. In color processing, the k-means clustering is used to extract the dominant color of the environment, and the number of clustering centers is set to 5. The pixels in the RGB color space are clustered to obtain the color values of each clustering center and their proportions. The clustering center with the highest proportion is selected as the reference color, and a gradient color band is generated in the HSV space, with the hue offset range of ±10 degrees and the saturation change range of 0.2. The radial basis function uses the polynomial form φ(r)=(r^2 + c^2)^(1 / 2), where r is the spatial distance and c is the shape parameter, with a value of 0.5. A 16×16 control point grid is set in the contour area, and the interpolation coefficients are obtained by solving the linear equations. The chromatic aberration threshold is set to 3 units, and local adjustment is performed when the chromatic aberration between adjacent regions exceeds the threshold. Color smoothing is filtered using a 5×5 Gaussian kernel, and the standard deviation is set to 1.2. The brightness value is remapped to the range of 40 - 220 by histogram equalization, avoiding being too dark or too bright while maintaining details. Texture generation is based on a 32×32 pixel sample block, and the Markov random field is used to establish the conditional probability relationship between pixels. The texture feature vector includes three components: contrast, directionality, and roughness, which are calculated by the gray-level co-occurrence matrix. The feature vectors of the environmental texture and the generated texture are calculated, and the cosine similarity is used to measure the texture coordination degree, with the coordination degree threshold set to 0.85. During weight fusion, the weight of the texture pattern linearly decreases from the center to the edge, with a decreasing range of 50 pixels.Edge processing uses the alpha blending function f(x) = 1 - x^2, where x is the normalized distance to the edge. The spatial standard deviation of bilateral filtering is set to 3, and the range standard deviation is set to 0.1 to achieve a smooth effect while preserving edges. By adjusting the contrast gain to 1.2, the material details are enhanced. The image resolution generated during the processing is 1024×1024 pixels, with a color depth of 24 bits. The processing time is approximately 300 milliseconds. The generated design solution includes four key elements: position, shape, color, and texture, with a data volume of approximately 2MB.

[0033] S106. Use the physical simulation method to simulate the movement process of the table and chair, determine whether the hidden contour design affects the normal use function of the table and chair, and optimize the hidden contour design according to the simulation results to generate an optimized hidden contour design solution.

[0034] The tetrahedral elements are used to mesh the structure of the table and chair. The positions of the mesh nodes are obtained through the element quality criterion, and the elastic modulus and Poisson's ratio parameters are extracted from the material database. Based on the positions of the mesh nodes and the material parameters, boundary conditions are established. The stiffness matrix is constructed through the nodal displacement interpolation function, and the static equilibrium equation is solved for the stiffness matrix to obtain the nodal displacement field. The element strain is calculated based on the nodal displacement field, and the element stress is solved through the constitutive equation. The von-mises equivalent stress criterion is used for the element stress to obtain the stress distribution contour map. The structure shape is topologically optimized based on the stress distribution contour map, and the natural frequency and vibration mode are calculated through harmonic response analysis. The optimized structural parameters are generated for the natural frequency and vibration mode.

[0035] Specifically, tetrahedral elements are used to perform finite element mesh division on the desk and chair structure. The mesh quality is controlled by the element quality criterion. Fixed constraints and displacement constraints are applied to the mesh nodes, and the elastic modulus and Poisson's ratio parameters are extracted from the material database. According to the structural mesh model, boundary conditions are established using gravity loads and external loads. The stiffness matrix is constructed through the nodal displacement interpolation function, and the structural static equilibrium equation is obtained. For the static equilibrium equation, the Lagrangian equation is used to establish the motion differential equation, and the Runge-Kutta method is used to numerically solve the motion equation to obtain the nodal displacement field. Based on the nodal displacement field, the element strain is calculated using the strain-displacement relationship, and the element stress is solved through the constitutive equation to obtain the structural stress field distribution. For the stress field distribution, the von-mises equivalent stress criterion is used for strength checking, and the mesh is refined in the area exceeding the strength limit value through the stress gradient method to obtain the stress distribution contour map. According to the stress distribution contour map, the topology optimization method is used to adjust the structure shape, and the optimization direction is determined through sensitivity analysis to obtain the improved structural scheme. The improved structural scheme is dynamically verified. The harmonic response analysis method is used to calculate the natural frequency and vibration mode, and the modal superposition method is used to verify the dynamic characteristics of the structure. According to the verification results, the optimized hidden contour curve equation is generated, the structural parameters are data encapsulated in json format, and the data packet is encrypted using the aes encryption algorithm and then transmitted to the smart home platform. In the finite element analysis of the desk and chair structure, tetrahedral elements are used to perform mesh dissection on the structure, the element size is set to 5-10 mm, and the mesh quality is controlled by the quality criterion with a Jacobian coefficient greater than 0.6. For the standard office chair structure, approximately 50,000 tetrahedral elements are generated, and the number of mesh nodes is approximately 15,000. The material parameters are selected as steel, with an elastic modulus of 210 GPa, a Poisson's ratio of 0.3, and a density of 7850 kg / m3. The static equilibrium equation is expressed as the linear equation system Kx = F, where K is the stiffness matrix, x is the nodal displacement vector, and F is the external load vector. The displacement within the element is interpolated through the shape function N(ξ, η, ζ) to construct the element stiffness matrix ke = ∫BTDBdV, where B is the strain matrix and D is the elastic matrix. The gravity load considers the self-weight of the chair, and the external load is set as a uniform load of 1000 N on the seat surface. The motion differential equation is in Lagrangian form M is the mass matrix and C is the damping matrix. The fourth-order Runge-Kutta method is used for solving, with the time step set to 0.001 seconds and the total calculation time being 1 second. For the seat movement condition, a horizontal thrust of 100 N is applied, and the calculated maximum displacement is approximately 20 mm. The stress calculation is based on Hooke's law σ = Dε. The element strain ε = Bu is calculated through the nodal displacement, and then the stress σ is solved. The material yield strength is set to 235 MPa, and the von-Mises criterion σvm = (σ12 - σ1σ2 + σ22)^(1 / 2) ≤ [σ] is used to judge the strength. There is stress concentration at the backrest connection, and the maximum equivalent stress reaches 180 MPa. The calculation accuracy is improved by refining the mesh size to 2 mm. The density method is used for structural optimization, with the design variable being the relative density ρ of the element and the objective function being the minimization of the structural flexibility. The relationship between density and elastic modulus E(ρ) = E0ρ^p is established through the SIMP model, where p = 3. The sensitivity analysis shows that the material utilization rate at the backrest support is low. The thickness of this area is reduced through iterative optimization, and the structural mass is reduced by 15% after optimization. The first six natural frequencies are calculated for dynamic verification. The first natural frequency is 8.5 Hz, and the vibration mode is forward and backward swing. Rayleigh damping C = αM + βK is adopted, with the coefficients α = 0.5 and β = 0.002. The harmonic response analysis is carried out in the frequency range of 0 - 50 Hz, and the resonance peak displacement amplitude is less than 5 mm. Finally, the optimization results are encapsulated in json format, including information such as nodal coordinates, element connection relationships, and material parameters. The size of the data packet is approximately 2 MB, and it is encrypted using the aes-256 encryption algorithm and then uploaded to the platform.

[0036] S107. Transmit the parameters of the optimized hidden contour design scheme to the intelligent control system, combine with the historical design scheme data, and train through machine learning algorithms to achieve the dynamic adaptation of the hidden contour of the desks and chairs to the surrounding environment.

[0037] According to the hidden contour design scheme data packet sent by the smart home platform, decrypt the data packet through the symmetric encryption algorithm, and obtain the valid data sequence by using the data integrity verification method; for the valid data sequence, preprocess the parameter time series by using the sliding window method, and filter out the noise data through wavelet transform to obtain the smooth parameter sequence; according to the smooth parameter sequence, construct a time series predictor by using the recurrent neural network, and extract the time series feature vector from the smooth parameter sequence through the long short-term memory unit; for the time series feature vector, train and fit it by using the radial basis kernel function support vector regressor. If the prediction error of the validation set is less than the preset threshold, collect the environmental data through the sliding time window for online incremental learning.

[0038] Specifically, obtain the optimized hidden contour design scheme data packet from the smart home platform data center, decrypt the data packet using the symmetric encryption algorithm, and verify the data validity using the data integrity verification method. According to the content of the data packet, preprocess the parameter time series using the sliding window method, filter out the noise data through wavelet transform, and obtain the smoothed parameter sequence. For the smoothed parameter sequence, construct a time series predictor using a recurrent neural network, extract features from historical data through long short-term memory units, and obtain the time series feature vector. According to the time series feature vector, normalize the feature data using the min-max normalization method, and divide the data subset according to the time ratio using the stratified sampling method. Perform five-fold cross-validation on the data subset, fit the training data using a radial basis kernel function support vector regressor, and optimize the kernel function parameters through the grid search method. Calculate the prediction error based on the validation set, update the network parameters using the adaptive moment estimation method, and control the parameter convergence through the learning rate annealing method. Perform online incremental learning on the trained network, continuously collect environmental data using the sliding time window, and achieve dynamic update through the parameter adaptation method. Package the updated network parameters and control instructions, and transmit the data to the intelligent control unit using the encrypted transmission protocol to achieve real-time dynamic adaptive control. The data transmission uses the AES-256 encryption algorithm to encrypt the design scheme data packet, with a key length of 256 bits and a block length of 128 bits. The data packet contains three parts: contour parameters, material parameters, and environmental parameters, with a total size of approximately 2MB. Verify the data integrity through the MD5 checksum to ensure that the data has not been tampered with during transmission. When preprocessing the parameter sequence, use a sliding window with a length of 24 and a window step size of 8. Use the db4 wavelet basis to decompose the data into 3 layers, and denoise the high-frequency coefficients using the soft threshold method, with the threshold set to 2.5 times the standard deviation. The signal-to-noise ratio after processing is increased by approximately 6dB, and the data smoothness is significantly improved. The recurrent neural network uses a bidirectional LSTM structure, including 2 hidden layers, with 128 neurons in each layer. The input features include 10 dimensions such as position coordinates, shape parameters, and environmental illumination. Control the retention ratio of historical information through the forget gate and the update ratio of new information through the input gate. The data normalization uses the formula x'=(x - xmin) / (xmax - xmin) to map the feature values to the [0,1] interval. The stratified sampling divides the data set into 5 time periods in chronological order, and randomly selects 80% of the samples as the training set and 20% as the test set in each time period. The ratio of positive samples to negative samples in the training set remains 1:1. The support vector regressor uses the Gaussian radial basis kernel function K(x,x') = exp(-γ||x - x'||2), where γ is the kernel parameter. Optimize the γ value in the range of [0.001,1] through the grid search method, and search for the penalty factor C in the range of [1,1000]. The cross-validation is performed in a 5-fold manner, and the average prediction error is less than 5%.The parameter update adopts the adaptive moment estimation algorithm. The initial value of the learning rate is set to 0.001, and it decays to 0.95 times the original value after every 1000 iterations. The momentum term is set to 0.9, which effectively prevents parameter oscillation. The root mean square error of the network on the validation set drops from the initial 0.15 to 0.03. For online incremental learning, environmental data is continuously collected using a 60-second sliding window, and the new and old data are fused through the exponential weighted average method with a weight decay coefficient of 0.95. The adaptive algorithm dynamically adjusts the network parameters according to the magnitude of the prediction error, and parameter update is triggered when the error exceeds the threshold of 0.1. The updated control command is transmitted to the execution unit at a frequency of 50Hz to achieve millisecond-level response.

[0039] The above content is only an example and explanation of the structure of the present invention. Those skilled in the art of this technology can make various modifications or supplements to the described specific embodiments or use similar methods for substitution. As long as they do not deviate from the structure of the invention or exceed the scope defined by this claim book, they should fall within the protection scope of the present invention.

Claims

1. A remote management method for tables and chairs based on a smart home platform, characterized in that: The method comprises: Obtaining the surrounding environment information of the table and chair, including the color and material attribute data of the wall and floor environment elements, and determining the initial color and material parameters of the hidden outline of the table and chair according to the obtained environmental color and material attribute data, including hue, saturation, brightness and surface texture characteristics; Establish a three-dimensional model of tables and chairs, and obtain the key structural points and outer contour coordinate information of the tables and chairs by analyzing the three-dimensional model; Use computer vision algorithms to analyze home space images, extract the main visual elements in the home space, determine the spatial relationship and visual coordination between the visual elements, and generate a visual feature map of the home space; Match the key structural points and outer contour coordinate information of the three-dimensional model of tables and chairs with the visual feature map of the home space, and calculate the optimal position and shape parameters of the hidden contours of the tables and chairs in the home space through the spatial mapping algorithm; According to the calculated optimal position and shape parameters, the size, shape and color of the hidden outline of the table and chair are dynamically adjusted to make it coordinate with the surrounding environment, and the initial design plan of the hidden outline is generated; Use physical simulation methods to simulate the movement process of tables and chairs, determine whether the hidden contour design affects the normal use function of tables and chairs, optimize the hidden contour design according to the simulation results, and generate an optimized hidden contour design scheme; The optimized hidden contour design parameters are passed to the intelligent control system, combined with historical design data and trained through a machine learning algorithm to achieve dynamic adaptation of the hidden contours of tables and chairs to the surrounding environment.

2. The method according to claim 1, characterized in that The method of obtaining the surrounding environment information of the table and chair, including the color and material attribute data of the wall and floor environment elements, and determining the initial color and material parameters of the hidden contours of the table and chair according to the obtained environmental color and material attribute data, including hue, saturation, brightness and surface texture characteristics, includes: The wall area is scanned by a color sensor array, the main color value of the wall is obtained from the wall color distribution data, and the wall depth distribution map is obtained by a material texture analyzer; The light reflection degree of the wall is measured by using a ray tracing model according to the main color value of the wall, the indoor light intensity value is obtained from the photometric calculation unit, and the light intensity value is quantified to obtain a light intensity distribution diagram; The light intensity distribution map and the wall depth distribution map are processed by convolution operation to obtain an overall brightness distribution map of the environment, and a three-dimensional space brightness model is established for the overall brightness distribution map of the environment; A depth camera is used to perform three-dimensional scanning and modeling of the surface of tables and chairs. The lighting parameters of the table and chair contour areas are calculated according to the three-dimensional space brightness model to obtain the lighting parameter set of the table and chair contour areas.

3. The method according to claim 1, characterized in that The three-dimensional model of the table and chair is established, and the key structural points and outer contour coordinate information of the table and chair are obtained by analyzing the three-dimensional model, including: A laser scanner is used to obtain a point cloud data set of the table and chair surface, and the point cloud data set is fused from multiple angles through a point cloud registration algorithm to obtain a complete point cloud model; According to the complete point cloud model, a Poisson reconstruction algorithm is used to perform triangular mesh division, and the mesh vertex coordinates are optimized by a Laplace smoothing method to obtain a smooth mesh model; The Gaussian curvature calculation formula is used to extract the curvature value of the mesh vertex for the smooth mesh model, and a set of feature points is obtained by setting a curvature threshold. A sobel operator is used to perform edge detection according to the feature point set to obtain a binary image, and a key inflection point coordinate sequence is extracted from the binary image.

4. The method according to claim 1, characterized in that: The method of using a computer vision algorithm to analyze the home space image, extract the main visual elements in the home space, determine the spatial relationship and visual coordination between the visual elements, and generate a home space visual feature map includes: A depth camera is used to obtain an original image of the home space, and the original image is processed by a Gaussian filter and gamma correction to obtain an enhanced image; Extracting gray-level co-occurrence matrix texture features, color square color features, and gradient operator edge features from the enhanced image to obtain a feature vector, performing region growing segmentation on the feature vector to obtain a visual element labeling map; Calculating the coordinates of the centroid of the visual elements for the visual element labeling diagram, constructing a spatial distance matrix using the Euclidean distance method, calculating the color similarity of the visual elements by the intersection of the color histograms, and obtaining a visual coordination matrix; According to the visual coordination matrix, the spectral clustering method is used to hierarchically divide the visual elements, the quadtree method is used to construct the space division structure, the brightness weight calculation is used to generate the visual attention distribution map, and the home space visual feature map is obtained.

5. The method according to claim 1, characterized in that The key structural points and outer contour line coordinate information of the three-dimensional model of the table and chair are matched with the visual feature map of the home space, and the optimal position and shape parameters of the hidden contour of the table and chair in the home space are calculated by the space mapping algorithm, including: The key structural points of the three-dimensional model of tables and chairs are projected onto a two-dimensional plane using a perspective projection method, and a standardized plane coordinate point set is obtained through homography matrix calculation; A surf feature point extraction algorithm is used to obtain a feature descriptor according to the standardized plane coordinate point set, and a feature point correspondence relationship set is obtained by a nearest neighbor matching method; A least squares optimization objective function is established for the feature point correspondence set, and the position coordinates of the tables and chairs are iteratively optimized using the gradient descent method to obtain the initial placement positions of the tables and chairs; According to the initial placement of the table and chair, the contour is fitted by a cubic Bezier curve, the curve control point interval ratio and tangent direction constraints are set, and the curvature value of the sampling point is calculated to obtain the contour shape feature vector.

6. The method according to claim 1, characterized in that The method of dynamically adjusting the size, shape and color of the hidden outline of the table and chair according to the calculated optimal position and shape parameters to coordinate with the surrounding environment and generating an initial design scheme of the hidden outline includes: The affine transformation method is used to scale the table and chair contours, and the contours of the tables and chairs are interpolated by cubic spline to obtain smooth boundary curves; A color clustering method is used to obtain the main color tone from the smooth boundary curve, and a linear gradient color band is obtained through hue-saturation mapping; According to the linear gradient color band, a polynomial radial basis function is used to perform spatial interpolation, and a color distribution map of the transition area is obtained through a color difference threshold constraint; A weighted fusion method is used to superimpose the color distribution map of the transition area with the environment texture, and an alpha blending function is used to obtain a hidden contour with decreasing edge transparency.

7. The method according to claim 1, characterized in that The physical simulation method is used to simulate the movement process of the table and chair, determine whether the hidden contour design affects the normal use function of the table and chair, optimize the hidden contour design according to the simulation result, and generate an optimized hidden contour design scheme, including: Tetrahedron elements are used to mesh the table and chair structure, the mesh node positions are obtained through the element quality criterion, and the elastic modulus and Poisson's ratio parameters are extracted according to the material database; Establishing boundary conditions according to the grid node positions and material parameters, constructing a stiffness matrix through a node displacement interpolation function, and solving a static equilibrium equation for the stiffness matrix to obtain a node displacement field; Calculate unit strain according to the node displacement field, solve unit stress by constitutive equation, and obtain stress distribution cloud diagram by adopting von-mises equivalent stress criterion for the unit stress; The structural shape is topologically optimized according to the stress distribution cloud map, the natural frequency and vibration mode are calculated through harmonic response analysis, and the optimized structural parameters are generated according to the natural frequency and vibration mode.

8. The method according to claim 1, characterized in that The optimized hidden contour design scheme parameters are transmitted to the intelligent control system, combined with the historical design scheme data, and trained through a machine learning algorithm to achieve dynamic adaptation of the hidden contour of the table and chair to the surrounding environment, including: According to the hidden contour design scheme data packet sent by the smart home platform, the data packet is decrypted by using a symmetric encryption algorithm, and a valid data sequence is obtained by using a data integrity verification method; For the effective data sequence, the sliding window method is used to preprocess the parameter time series, and the noise data is filtered out by wavelet transform to obtain a smooth parameter sequence; According to the smoothing parameter sequence, a time series predictor is constructed by using a recurrent neural network, and a time series feature vector is obtained by extracting features of the smoothing parameter sequence through a long short-term memory unit; For the time series feature vector, a radial basis kernel function support vector regressor is used for training and fitting. If the prediction error of the validation set is less than a preset threshold, environmental data is collected through a sliding time window for online incremental learning.

Images