Intelligent pressure-sensitive bedding adjustment method and device based on sleep image data

Abstract

The present invention discloses an intelligent pressure-sensitive bedding adjustment method and device based on sleep image data, belonging to the field of intelligent control technology. The method is used to solve the technical problem that existing bedding adjustment methods rely on pressure data to judge sleeping posture, which is prone to misjudgment, resulting in unreasonable adjustment or energy waste. The method includes: superimposing a panoramic bedding image with a pre-stored bedding airbag distribution map to obtain a bedding superimposed image; associating pressure data at the same time with the corresponding folding airbag positions in the bedding superimposed image; segmenting the bedding superimposed image based on the pressure data to obtain a bedding pressure area image; determining non-human body areas in the bedding pressure area image, and clearing the pressure data associated with the folding airbags in the non-human body areas; converting the processed pressure data into a pressure matrix, and performing analysis and calculation to determine the user's sleeping posture information; and inflating or deflating the airbags in the intelligent pressure-sensitive bedding based on the sleeping posture information.

Description

Technical Field

[0001] The present invention relates to the field of intelligent control technology, and in particular to an intelligent pressure-sensitive bedding adjustment method and device based on sleep image data. Background Art

[0002] Sleep quality is directly linked to sleeping posture. Sleeping or lying in bed is an effective way to relax the entire body, especially the spine, because it allows for full contact between the body and the bed, reducing the burden of resisting gravity when standing upright. However, improper sleeping or lying posture can exacerbate spinal and muscle contraction, or overstretch local ligaments, leading to conditions such as stiff neck and lower back pain, while also compromising sleep comfort.

[0003] Currently, some beds or mattresses with intelligent height adjustment have emerged. By adjusting the height of different parts of the bed or mattress, they provide different support forces for different parts of the user's body, thereby maintaining a normal physiological curve during sleep. However, existing intelligent adjustment methods mostly rely on collecting pressure data from the user on the mattress to determine the user's posture, or using image acquisition to identify the user's posture and then provide support accordingly. However, relying solely on pressure data detected by the mattress is prone to misjudgment. For example, a heavy object (such as a schoolbag or quilt) placed on the mattress may be identified as part of the human body, resulting in an incorrect human posture and unreasonable height adjustment, affecting normal mattress use. Alternatively, a heavy object may be identified as a human body when the person is not lying on the mattress, leading to unnecessary automatic adjustment, which may cause the object to fall and waste energy. Summary of the Invention

[0004] Embodiments of the present invention provide an intelligent pressure-sensitive bedding adjustment method and device based on sleep image data, which is used to solve the following technical problems: In existing bedding adjustment methods, sleeping posture judgment relies on pressure data, which is prone to misjudgment, resulting in unreasonable adjustment or energy waste.

[0005] The embodiment of the present invention adopts the following technical solutions:

[0006] On the one hand, an embodiment of the present invention provides a method for adjusting intelligent pressure-sensitive bedding based on sleep image data. The intelligent pressure-sensitive bedding includes at least: an image collector, an intelligent air pump, a plurality of folding airbags vertically mounted in an array on a bedding base, and a plurality of fiber sensors mounted on the surface of each folding airbag; the method includes: receiving a panoramic image of the bedding taken by the image collector, and superimposing the panoramic image of the bedding with a pre-stored bedding airbag distribution map to obtain a superimposed bedding image; receiving pressure data returned by each fiber sensor, and comparing the pressure data at the same moment with the pressure of the corresponding folding airbag in the superimposed bedding image; The method is configured to associate the bedding with a user's position; segment the bedding superimposed image based on the pressure data to obtain an image of the pressure area of the bedding; perform human body part recognition on the image of the pressure area of the bedding to determine a non-human area, and reset the pressure data values associated with the folded airbags in the non-human area to zero; convert the zeroed pressure data into a pressure matrix, and analyze and calculate the pressure matrix to determine the user's sleeping posture information; and control the intelligent air pump based on the sleeping posture information to inflate or deflat the airbags in the intelligent pressure-sensitive bedding, so that the intelligent pressure-sensitive bedding can be intelligently adjusted according to the user's sleeping posture.

[0007] In a feasible embodiment, before receiving the panoramic image of the bedding captured by the image collector, the method further includes: drawing and saving the bedding airbag distribution map according to the screen ratio of the image collector and the airbag position design data when the smart pressure-sensitive bedding is produced; wherein the bedding airbag distribution map is a black and white line map; in the bedding airbag distribution map, extracting and saving the four vertex coordinates of the smart pressure-sensitive bedding; receiving the captured image of the image collector, and detecting whether the length-to-width ratio of the first bedding in the captured image is consistent with the length-to-width ratio of the bedding airbag; The length-to-width ratio of the second bedding in the capsule distribution map is the same as that of the first bedding; if not, the image collector is controlled to move along the slide rail until the length-to-width ratio of the first bedding is the same as that of the second bedding; wherein the image collector is installed in the slide rail directly above the smart pressure-sensitive bedding and can move translationally along the slide rail; the slide rail is parallel to the perpendicular bisectors of the two short sides of the smart pressure-sensitive bedding; the image collector is controlled to zoom and locate the four vertices of the smart pressure-sensitive bedding in the captured image to the four saved vertex coordinates for image calibration.

[0008] In a feasible embodiment, the panoramic bedding image is superimposed with a pre-stored bedding airbag distribution map to obtain the bedding superimposed image, specifically comprising: adjusting the black pixels in the bedding airbag distribution map to a semi-transparent state and the white pixels to a transparent state to obtain a bedding airbag distribution line map; aligning the four vertices of the smart pressure-sensitive bedding in the bedding airbag distribution line map with the four vertices of the smart pressure-sensitive bedding in the panoramic bedding image; after the alignment, superimposing the bedding airbag distribution line map on an upper layer of the panoramic bedding image, and merging the two layers to obtain the bedding superimposed image.

[0009] In a feasible embodiment, the pressure data at the same moment is associated with the corresponding folding airbag position in the bedding overlay image, specifically including: a one-to-one correspondence between the position coordinates of each folding airbag in the bedding overlay image and the position of each folding airbag in the pre-stored intelligent pressure-sensitive bedding; setting the same number for the folding airbags located at the same position in the bedding overlay image and the intelligent pressure-sensitive bedding, and assigning the number to the fiber sensor corresponding to the surface of the folding airbag; and according to the number of each fiber sensor and the number of each folding airbag in the bedding overlay image, associating the pressure data returned by each fiber sensor at the current moment with the corresponding folding airbag in the bedding overlay image.

[0010] In one feasible embodiment, the bedding superimposed image is segmented according to the value of the pressure data to obtain an image of the compressed area of the bedding, specifically comprising: determining, in the bedding superimposed image, all compressed folded airbags whose pressure data values are not 0; determining the pixels within a preset range around each compressed folded airbag, and setting the pixel values of the remaining pixels to 255 to complete the segmentation of the bedding superimposed image to obtain the image of the compressed area of the bedding; wherein the preset range is the range enclosed by a square with the center point of the folded airbag cross-section and a side length of twice the diameter of the folded airbag cross-section.

[0011] In a feasible embodiment, human body part recognition is performed on the bedding pressure area image to determine the non-human area, specifically comprising: determining a target area threshold based on the grayscale value distribution of the bedding pressure area image; determining an area in the bedding pressure area image with a grayscale value greater than the target area threshold as a target area; dividing the target area into a plurality of target blocks, using the gradient value of each pixel point in each target block as a weight vector, and calculating a gradient histogram of different gradient directions in each target block; normalizing the gradient histogram of different gradient directions in each target block to obtain a gradient feature vector for each target block; combining the gradient feature vectors of each target block into a gradient feature vector of the target area, and inputting the gradient feature vectors into a classifier to classify each target block separately; connecting the target blocks classified as human targets to determine them as human body areas; and determining the area in the bedding pressure area image other than the human body area as the non-human body area.

[0012] In a feasible embodiment, the target area threshold is determined based on the grayscale value distribution of the bedding pressure area image, specifically including: determining the grayscale value distribution interval in the bedding pressure area image; randomly selecting multiple grayscale values in the grayscale value distribution interval as multiple initial cluster centers, and performing k-means clustering on the bedding pressure area image; after clustering is completed, analyzing the grayscale value change data of each cluster center during the clustering process to determine the cluster center where the grayscale value change data has a sudden change and the grayscale value of the mutation point; and determining the grayscale value of the mutation point as the target area threshold.

[0013] In a feasible implementation, the pressure matrix is analyzed and calculated to determine the sleeping posture of the user, specifically including: in the pressure matrix, determining the area where the matrix elements with values greater than 0 are located as the pressure area, and counting the number of matrix elements in the pressure area; according to the number of matrix elements in the pressure area and the element scale, as the current pressure area of the smart pressure-sensitive bedding; wherein the element scale is the proportional relationship between a matrix element and the area occupied by a folding airbag in the smart pressure-sensitive bedding; comparing the current pressure area with the preset pressure area intervals of different posture types to preliminarily determine the posture type of the user; wherein the posture types include sitting, lying flat, and lying on the side; the lying flat specifically includes lying on the back and lying on the stomach, and the Side-lying specifically includes lying on the left side and lying on the right side; when the posture type is lying on the side, the peak values of several matrix element values in the pressure area are obtained, and the peak values of the several matrix element values are fitted into a straight line to be determined as the force axis; the number of matrix elements on both sides of the force axis in the pressure area is analyzed, and the side with more matrix elements is determined as the facing side of the user; when the posture type is lying flat, the average value of the element values in the pressure area is calculated, and the area formed by elements smaller than the average value of the element values is determined as a low-force area; if there is at least one low-force area in the longitudinal middle area of the pressure area, the user is determined to be in a supine posture, otherwise the user is determined to be in a prone posture.

[0014] In a feasible embodiment, according to the sleeping posture, the intelligent air pump is controlled to inflate or deflat the airbags in the intelligent pressure-sensitive bedding, so that the intelligent pressure-sensitive bedding can be intelligently adjusted according to the sleeping posture of the user. Specifically, the following steps are performed: judging the height of the user according to the user's posture; determining the position of each body part of the user according to the height of the user and the distribution of pressure data; dividing the pressure area into at least a trunk area, a left upper limb area, a right upper limb area, a left lower limb area, and a right lower limb area according to the position of each body part; obtaining the minimum value of the element value in each area and the pressure data; The maximum value is obtained, and the average value is calculated to obtain the pressure balance value corresponding to each partition; the folding airbags corresponding to the elements with a pressure greater than the pressure balance value in the intelligent pressure-sensitive bedding are slowly deflating, and the folding airbags corresponding to the elements with a pressure less than the pressure balance value are slowly inflated; the pressure value of each folding airbag in the pressurized area is monitored in real time during the inflation or deflating process; when the error between the pressure value borne by the folding airbag and the corresponding pressure balance value enters the allowable error range, or the real-time deformation value reaches the boundary value of the expansion and contraction interval corresponding to the current adjustment gear, the inflation or deflating process is stopped; wherein, different adjustment gears correspond to different expansion and contraction intervals of the folding airbag.

[0015] On the other hand, an embodiment of the present invention also provides an intelligent pressure-sensitive bedding adjustment device based on sleep image data, the device including: 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, so that the at least one processor can execute an intelligent pressure-sensitive bedding adjustment method based on sleep image data according to any of the above embodiments.

[0016] Compared with the prior art, the intelligent pressure-sensitive bedding adjustment method and device based on sleep image data provided by the present invention has the following beneficial effects:

[0017] The present invention relies on a specially designed intelligent pressure-sensitive bedding system. It combines pressure data collected by the bedding system with image data captured by an image collector to generate an accurate human body pressure data matrix, reducing the possibility of misidentifying pressure data generated by non-human body parts as human body parts. The pressure data matrix analyzes the user's sleeping position and adjusts the user's body parts accordingly. Each zone adjusts the folding airbag based on different pressure balance values, achieving balanced force and support for all parts of the user's body. This method can improve the accuracy of the intelligent pressure-sensitive bedding's sleeping position detection results, allowing it to better fit each user's body and provide better support. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments described in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work. In the drawings:

[0019] Figure 1 A top view of the airbag layer structure of an intelligent pressure-sensitive bedding provided by an embodiment of the present invention;

[0020] Figure 2 A top view of the base layer structure of an intelligent pressure-sensitive bedding provided by an embodiment of the present invention;

[0021] Figure 3 A flow chart of a method for adjusting intelligent pressure-sensitive bedding based on sleep image data provided by an embodiment of the present invention;

[0022] Figure 4 A schematic structural diagram of an intelligent pressure-sensitive bedding adjustment device based on sleep image data provided by an embodiment of the present invention;

[0023] Description of reference numerals:

[0024] 1. Folding airbag; 2. Head strip airbag; 3. Foot strip airbag; 4. Intelligent air pump; 5. Folding airbag mounting slot; 6. Head strip airbag mounting slot; 7. Foot strip airbag mounting slot. DETAILED DESCRIPTION

[0025] In order to enable those skilled in the art to better understand the technical solutions of the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.

[0026] First, an embodiment of the present invention provides a method for adjusting an intelligent pressure-sensitive bedding based on sleep image data, which is applied to an intelligent pressure-sensitive bedding. The intelligent pressure-sensitive bedding in the present invention can be understood as an intelligent mattress, which is composed of an airbag layer and a base layer.

[0027] As a feasible implementation method, Figure 1 This is a top view of the airbag layer structure of an intelligent pressure-sensitive bedding provided in an embodiment of the present application, as shown in FIG. Figure 1 As shown, the airbag layer has a number of folded airbags 1 installed vertically in an array. A fiber sensor (not shown) is installed on the surface of each folded airbag to send the pressure value of each folded airbag to the processor. Different numbers of head strip airbags 2 and foot strip airbags 3 are also installed at the head and tail of the mattress. The arrangement of the folded airbags can be Figure 1 The staggered arrangement shown can also be arranged in rows and columns.

[0028] Furthermore, the airbag layer is installed on the base layer. Figure 2 This is a top view of the base layer structure of an intelligent pressure-sensitive bedding provided in an embodiment of the present application, as shown in FIG. Figure 2 As shown, the base layer includes a smart air pump 4, a foldable airbag mounting slot 5 for mounting a foldable airbag 1, a head strip airbag mounting slot 6 for mounting a head strip airbag 2, and a foot strip airbag mounting slot 7 for mounting a foot strip airbag 3. All foldable and strip airbags are connected to the smart air pump 4 via pipes. The smart air pump is connected to a processor and inflates or deflates the connected airbags according to the processor's instructions. The processor is connected to a fiber sensor on the surface of each airbag via a wire to receive pressure data from the fiber sensor. The smart air pump and processor are both installed at the foot end of the mattress.

[0029] Based on the structure of the aforementioned intelligent pressure-sensitive bedding, this application adds a matching image collector and a slide rail for mounting the image collector. The slide rail is installed directly above and parallel to the perpendicular midline of the two short sides of the intelligent pressure-sensitive bedding. This ensures that the image collector remains on the midline of the intelligent pressure-sensitive bedding as it moves on the slide rail, maintaining symmetry in the field of view. Preferably, the image collector can be an infrared camera or an infrared camera head.

[0030] Based on the above intelligent pressure-sensitive bedding structure, Figure 3 A flow chart of a method for adjusting intelligent pressure-sensitive bedding based on sleep image data provided by an embodiment of the present invention is shown in FIG. Figure 3 As shown, the method specifically includes:

[0031] S301: The processor receives a panoramic image of bedding captured by an image collector, and superimposes the panoramic image of bedding with a pre-stored bedding airbag distribution map to obtain a superimposed bedding image.

[0032] Specifically, after the smart pressure-sensitive bedding is produced, an airbag distribution map is created based on the image acquisition device's screen ratio and the airbag location design data for the current model of smart pressure-sensitive bedding. This map is then stored in the bedding processor. The airbag distribution map is a black and white line drawing, with black lines and a white background.

[0033] Furthermore, in the bedding airbag distribution map, a plane rectangular coordinate system is established with the lower left vertex of the map as the origin and the two edges connected to the origin as coordinate axes. Then, the coordinates of the four vertices of the smart pressure-sensitive bedding in the bedding airbag distribution map are extracted and saved in the processor.

[0034] Furthermore, after the intelligent pressure-sensitive bedding is installed and powered on, the processor receives the captured image from the image collector and detects whether the first bedding length-to-width ratio in the captured image is the same as the second bedding length-to-width ratio in the bedding airbag distribution map.

[0035] If they are different, it proves that the image collector is not located at the exact center of the intelligent pressure-sensitive bedding. Therefore, the image collector is controlled to move along the slide rail until the length-to-width ratio of the first bedding is the same as that of the second bedding.

[0036] Furthermore, the image acquisition device is controlled to zoom and locate the four vertices of the intelligent pressure-sensitive bedding in the captured image at the four saved vertex coordinates for image calibration. The purpose of image calibration is to ensure that the bedding image captured by the image acquisition device completely overlaps with the pre-stored bedding airbag distribution map, thereby increasing the accuracy of subsequent image overlay.

[0037] Furthermore, in the bedding airbag distribution map, the transparency values of all pixels with a pixel value of 0 are set to 100, and the transparency values of all pixels with a pixel value of 255 are set to 0, so as to adjust the black pixels to a semi-transparent state and the white pixels to a transparent state, thereby obtaining a bedding airbag distribution line map.

[0038] As a feasible implementation, the alpha channel value of black pixels with a pixel value of 0 in the bedding airbag distribution map is set to 100, making the black pixels semi-transparent. Furthermore, the alpha channel value of white pixels with a pixel value of 255 is set to 0, making the white pixels transparent. This converts the bedding airbag distribution map into a line drawing. Except for the lines, everything else is transparent to avoid obstructing the image details of the panoramic bedding image after superposition.

[0039] Furthermore, the four vertices of the smart pressure-sensitive bedding in the bedding airbag distribution line diagram are aligned with the four vertices of the smart pressure-sensitive bedding in the bedding panoramic image. After alignment, the bedding airbag distribution line diagram is superimposed on the upper layer of the bedding panoramic image, and the two layers are merged to obtain the bedding superimposed image.

[0040] S302: The processor receives the pressure data transmitted by each fiber sensor, and associates the pressure data at the same moment with the corresponding folded airbag position in the bedding overlay image.

[0041] Specifically, the processor maps the position of each folded airbag in the overlaid bedding image to the position of each folded airbag in the intelligent pressure-sensitive bedding. It then assigns the same number to the corresponding fiber sensor on the surface of the folded airbag, located in the same position in the overlaid bedding image and the intelligent pressure-sensitive bedding.

[0042] Furthermore, according to the number of each fiber sensor and the number of each folded airbag in the bedding superimposed image, the pressure data returned by each fiber sensor at the current moment is associated with the corresponding folded airbag in the bedding superimposed image.

[0043] In one embodiment, Figure 1 The placement direction shown is the positive direction. Numbering begins with the foldable airbag in the upper left corner, with the first foldable airbag numbered 1. Numbering increases as you go right. At the end of a row, numbering resumes from the leftmost edge of the next row until the last foldable airbag has been numbered. Following this method, the foldable airbags in the bedding overlay image and the foldable airbags in the intelligent pressure-sensitive bedding are numbered. Finally, based on matching numbers, the pressure data is associated with the foldable airbags in the bedding overlay image.

[0044] S303 : Segment the bedding superimposed image according to the pressure data value to obtain a bedding pressure area image.

[0045] Specifically, within the bedding overlay image, all compressed folded airbags with non-zero pressure data are identified. Pixels within a preset range surrounding each compressed folded airbag are identified, and the remaining pixels are set to 255 to complete segmentation of the overlay image, resulting in an image of the compressed bedding area. The preset range is defined as the area enclosed by a square centered at the center of the folded airbag's cross-section and with a side length of twice the diameter of the cross-section. The "segmentation" described here is not segmentation in the traditional sense; it simply retains the pixels in the compressed area and sets the remaining pixels to white. The information contained in the resulting image of the compressed bedding area can be considered a segmented portion of the original image.

[0046] As a feasible implementation method, since the bedding superimposed image contains both the folded airbag image and the bedding panoramic image, the area where the folded airbag has a non-zero pressure data value (i.e., the compressed folded airbag) is located is retained, and the pixels in the area where the folded airbag has a zero pressure data value (i.e., the uncompressed folded airbag) are located are changed to white or directly cut out. The remaining area is the image of the actual compressed area of the bedding.

[0047] S304 , performing human body part recognition on the image of the pressure area of the bedding, determining the non-human body area, and clearing the pressure data associated with the folding airbag in the non-human body area to zero.

[0048] Specifically, the target area threshold is determined according to the grayscale value distribution of the bedding pressure area image, which specifically includes: determining the grayscale value distribution interval in the bedding pressure area image; randomly selecting multiple grayscale values in the grayscale value distribution interval as multiple initial clustering centers, and performing k-means clustering on the bedding pressure area image; after clustering is completed, analyzing the grayscale value change data of each cluster center during the clustering process, determining the cluster center where the grayscale value change data has a mutation and the grayscale value of the mutation point; and determining the grayscale value of the mutation point as the target area threshold.

[0049] As a feasible implementation method, since in infrared images, the grayscale values of targets of the same category are not much different, while the grayscale values of targets of different categories are quite different, after clustering is completed, the cluster centers of each subclass belonging to the same category of target objects still change approximately linearly, while the cluster centers of subclasses at the critical points of different target objects will inevitably have a significant turning point. Therefore, by determining the grayscale value at the turning point, the threshold value of image segmentation can be determined. The processor in the present invention records the grayscale value of the initial cluster center after each clustering process, and calculates the difference between the grayscale value of the cluster center after each clustering and the grayscale value of the cluster center after the previous clustering. If the error between the calculated differences each time is always less than a certain smaller threshold value, until the error between the difference of a certain time and the previous difference is greater than the threshold value, then this time is the mutation point, and the grayscale value corresponding to the mutation point is the required target area threshold value.

[0050] Furthermore, the target area is identified as any region within the bedding compression area image whose grayscale value exceeds the target area threshold. The target area is then divided into several target blocks. The gradient values of each pixel within each target block are used as weight vectors to calculate the gradient histograms of different gradient orientations within each target block. The gradient histograms of different gradient orientations within each target block are normalized to obtain the gradient feature vectors of each target block. The gradient feature vectors of each target block are combined into the gradient feature vector of the target area and input into the classifier to classify each target block separately. Target blocks classified as human targets are connected and identified as human body regions. Regions other than human body regions within the bedding compression area image are then identified as non-human body regions.

[0051] Furthermore, all pressure data values associated with the folded airbag in the determined non-human area are set to zero. After the pressure data is processed, the non-zero portion is all human area. Except for the human area, the remaining pressure data values are all 0.

[0052] The above operation eliminates some interference factors in the pressure data, assigns the value of the pressure data not generated by the human body to 0, and then performs subsequent sleeping posture recognition, which can greatly improve the accuracy of sleeping posture recognition.

[0053] S305: Convert the processed pressure data into a pressure matrix, and analyze and calculate the pressure matrix to determine the user's sleeping posture information.

[0054] Specifically, the processor stores the processed pressure data in the form of a matrix. The specific method is: mark the row number i and column number j of each folding airbag in the top view of the intelligent pressure-sensitive bedding, and then display the pressure value Aij corresponding to the folding airbag (i, j) in the i-th row and j-th column of the pressure matrix, and finally obtain a complete pressure matrix.

[0055] In the pressure matrix, the area containing matrix elements with values greater than 0 is identified as the pressure area, and the number of matrix elements in the pressure area is counted. The number of matrix elements in the pressure area and the element scale are used as the current pressure area of the smart pressure-sensitive bedding. The element scale is the ratio between one matrix element and the area occupied by one folding airbag in the smart pressure-sensitive bedding. The current pressure area is then compared with the preset pressure area ranges for different posture types to preliminarily determine the user's posture type. Posture types include sitting, supine, and side-lying. Supine specifically includes supine and prone, and side-lying specifically includes left and right side lying.

[0056] In one embodiment, each element in the pressure matrix corresponds to the pressure value felt by a folded airbag. Because there are gaps between folded airbags, the present invention uses the length of the line connecting the centers of the two folded airbag cross-sections as a side length, A. The area of the square with side length A is calculated as the actual area occupied by the folded airbag cross-section. For example, if the calculated actual area occupied by a folded airbag cross-section is 25 cm², the element scale is 1:25. In the element scale, the value of the matrix element is always 1.

[0057] In one embodiment, according to the different pressure areas of different postures, the general relationship between the pressure areas is: lying flat > lying on the side > sitting. Therefore, different pressure area intervals are set for different posture types in this application. The actual pressure area can be preliminarily determined as the corresponding posture type based on which interval it is within. After determining the posture type, the specific posture can be further determined.

[0058] Furthermore, when the posture type is side-lying, the peak values of several matrix element values in the pressure area are obtained and fitted to a straight line to determine the force axis. The number of matrix elements on both sides of the force axis in the pressure area is analyzed, and the side with the largest number of elements is determined as the user's facing direction. When lying on the side, the main force area is on the torso. By fitting the element peaks to a straight line that roughly coincides with the torso and analyzing the number of elements on both sides of this line, the side with the largest number of elements is determined as the user's facing direction.

[0059] Furthermore, if the posture type is supine, the average value of the elements in the pressure area is calculated, and the area formed by the elements with less than the average value is identified as a low-pressure area. If at least one low-pressure area exists in the longitudinal middle section of the pressure area, the user is determined to be supine; otherwise, the user is determined to be prone. The principle behind this design is that when the human body is supine, the waist is arched, and the pressure between the body and the mattress is relatively low. When the user is prone, the pressure between the abdomen and the mattress is relatively high. Therefore, by calculating the average pressure of the entire body and selecting the area with lower pressure, if there is a low-pressure area in the middle section of the pressure area, the user is determined to be supine; otherwise, the user is prone.

[0060] S306 : Controlling the intelligent air pump according to the sleeping posture to inflate or deflat the airbags in the intelligent pressure-sensitive bedding, so that the intelligent pressure-sensitive bedding can be intelligently adjusted according to the user's sleeping posture.

[0061] Specifically, the user's height is determined based on their posture. Based on the user's height and the pressure data distribution, the location of each body part is determined. Then, based on the location of each body part, the pressure area is divided into at least a trunk zone, a left upper limb zone, a right upper limb zone, a left lower limb zone, and a right lower limb zone. The minimum and maximum values of the elements in each zone are then obtained and averaged to obtain the corresponding pressure balance value for each zone.

[0062] In one embodiment, when the user is in a supine or side-lying position, the maximum vertical span of the pressure zone is first determined as the user's body length. The locations of the shoulders and hips are then determined based on the areas with the greatest pressure within the pressure zone. Furthermore, the locations of other parts of the body, such as the neck, waist, and limbs, are determined based on the distribution of pressure within the body, thereby dividing the pressure zone into multiple zones.

[0063] Different parts of the human body have different weights. For example, there is a large difference in weight between the torso and the limbs. If the human body is adjusted as a whole, the weight differences between different parts of the body will be ignored. Therefore, in this application, each part is adjusted in a zone so that each part of the body can get appropriate support. When a human body lies on a non-adjustable mattress, due to the physiological curvature, the pressure exerted on the mattress by each part is uneven. For example, the torso and buttocks exert greater pressure on the mattress, while the waist exerts less pressure on the mattress. The uneven force causes the waist to be unable to get effective support, and the buttocks are squeezed, which can also cause discomfort. Therefore, this application calculates the average of the maximum and minimum pressure values in different zones to obtain the pressure balance value within the zone, and then adjusts the expansion and contraction value of the folding air cushion at the corresponding position according to the current actual pressure value, so that the pressure between the body and the mattress in the same zone is more balanced.

[0064] Furthermore, in each partition, the folding airbags corresponding to the elements with a pressure greater than the pressure balance value are slowly deflating, and the folding airbags corresponding to the elements with a pressure less than the pressure balance value are slowly inflated, and the pressure value of each folding airbag in the pressurized area during the inflation or deflating process is monitored in real time.

[0065] When the error between the pressure value of the folding airbag and the corresponding pressure balance value falls within the allowable error range, or when the real-time deformation value reaches the boundary value of the expansion and contraction range corresponding to the current adjustment gear, the inflation or deflating process is stopped. Different adjustment gears correspond to different expansion and contraction ranges of the folding airbag.

[0066] As a feasible implementation method, the processor will set different adjustment gears for the different weight ranges that all folding airbags can bear, and different adjustment gears correspond to different expansion and contraction ranges of the folding airbags. First, the values of the elements in the obtained pressure matrix that are greater than the minimum pressure threshold are added together to obtain the total weight currently borne by all folding airbags, where the minimum pressure threshold is used to set the minimum pressure that the human body can exert on the intelligent pressure-sensitive adjustment bedding. If the total weight is less than the preset weight threshold, the folding airbags are not adjusted. If the total weight is greater than or equal to the preset weight threshold, the corresponding adjustment gear is determined based on the weight range to which the total weight belongs.

[0067] In addition, the embodiment of the present invention also provides an intelligent pressure-sensitive bedding adjustment device based on sleep image data, such as Figure 4 As shown, the intelligent pressure-sensing bedding adjustment device based on sleep image data specifically includes:

[0068] at least one processor; and a memory communicatively connected to the at least one processor; wherein,

[0069] The memory stores instructions executable by at least one processor, so as to enable the at least one processor to perform:

[0070] receiving a panoramic image of the bedding captured by the image collector, and superimposing the panoramic image of the bedding with a pre-stored bedding airbag distribution map to obtain a superimposed bedding image;

[0071] Receive the pressure data returned by each fiber sensor and associate the pressure data at the same moment with the corresponding folded airbag position in the bedding overlay image;

[0072] Segmenting the bedding superimposed image according to the value of the pressure data to obtain an image of the bedding pressure area;

[0073] Performing human body part recognition on the image of the pressure area of the bedding to determine a non-human body area, and clearing the pressure data associated with the folded airbag in the non-human body area to zero;

[0074] Converting the processed pressure data into a pressure matrix, and analyzing and calculating the pressure matrix to determine the user's sleeping position;

[0075] According to the sleeping posture, the intelligent air pump is controlled to inflate or deflat the airbags in the intelligent pressure-sensitive bedding, so that the intelligent pressure-sensitive bedding can be intelligently adjusted according to the user's sleeping posture.

[0076] The various embodiments of the present invention are described in a progressive manner. Similar portions between the various embodiments can be referred to in conjunction with each other. Each embodiment focuses on the differences between the other embodiments. In particular, the device embodiments are generally similar to the method embodiments, so their description is relatively simple. For relevant portions, refer to the description of the method embodiments.

[0077] The above description of specific embodiments of the present invention is provided. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0078] The foregoing is merely an embodiment of the present invention and is not intended to limit the present invention. It will be apparent to those skilled in the art that various modifications and variations may be made to the embodiments of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the embodiments of the present invention are intended to be included within the scope of protection of the present invention.

Claims

1. A method for adjusting intelligent pressure-sensitive bedding based on sleep image data, characterized in that: The intelligent pressure-sensitive bedding comprises at least: an image collector, an intelligent air pump, a plurality of folding airbags vertically mounted in an array on a bedding base, and a plurality of fiber sensors mounted on the surface of each folding airbag; the method comprises: According to the screen ratio of the image collector and the airbag position design data during the production of the intelligent pressure-sensitive bedding, a bedding airbag distribution map is drawn and saved; wherein the bedding airbag distribution map is a black and white line drawing; Extracting and saving the coordinates of four vertices of the intelligent pressure-sensitive bedding from the bedding airbag distribution map; receiving a captured image from the image collector, and detecting whether a first bedding length-to-width ratio in the captured image is the same as a second bedding length-to-width ratio in the bedding airbag distribution map; If they are different, controlling the image collector to move along the slide rail until the length-to-width ratio of the first bedding is the same as the length-to-width ratio of the second bedding; wherein the image collector is installed in the slide rail directly above the smart pressure-sensitive bedding and can move translationally along the slide rail; the slide rail is parallel to the perpendicular midline of the two short sides of the smart pressure-sensitive bedding; Controlling the image collector to zoom, and positioning the four vertices of the intelligent pressure-sensitive bedding in the captured image to the four saved vertex coordinates, so as to perform image calibration; receiving a panoramic image of the bedding captured by the image collector, and superimposing the panoramic image of the bedding with a pre-stored bedding airbag distribution map to obtain a superimposed bedding image; receiving pressure data transmitted by each fiber sensor, and associating the pressure data at the same moment with the corresponding folded airbag position in the bedding overlay image; Segmenting the bedding superimposed image according to the pressure data to obtain an image of the bedding pressure area; Performing human body part recognition on the image of the pressure area of the bedding to determine a non-human body area, and setting the pressure data value associated with the folded airbag in the non-human body area to zero; Converting the zeroed pressure data into a pressure matrix, and analyzing and calculating the pressure matrix to determine the user's sleeping posture information; According to the sleeping posture information, the intelligent air pump is controlled to inflate or deflat the airbags in the intelligent pressure-sensitive bedding, so that the intelligent pressure-sensitive bedding can be intelligently adjusted according to the sleeping posture of the user.

2. The intelligent pressure-sensitive bedding adjustment method based on sleep image data according to claim 1, characterized in that: The bedding panoramic image is superimposed with a pre-stored bedding airbag distribution map to obtain a bedding superimposed image, specifically comprising: Adjusting the black pixels in the bedding airbag distribution map to a semi-transparent state and the white pixels to a transparent state to obtain a bedding airbag distribution line map; Aligning four vertices of the smart pressure-sensitive bedding in the bedding airbag distribution line diagram with four vertices of the smart pressure-sensitive bedding in the bedding panoramic image; After alignment, the bedding airbag distribution line diagram is superimposed on the upper layer of the bedding panoramic image, and the two layers are merged to obtain the bedding superimposed image.

3. The intelligent pressure-sensitive bedding adjustment method based on sleep image data according to claim 1, characterized in that: Correlate the pressure data at the same moment with the corresponding folded airbag position in the bedding overlay image, specifically including: Matching the position coordinates of each folding airbag in the bedding overlay image with the position of each folding airbag in the pre-stored intelligent pressure-sensitive bedding; For the folding airbags located at the same position in the bedding superimposed image and the intelligent pressure-sensitive bedding, the same number is set, and the number is assigned to the fiber sensor corresponding to the surface of the folding airbag; According to the number of each fiber sensor and the number of each folded airbag in the bedding superimposed image, the pressure data returned by each fiber sensor at the current moment is associated with the corresponding folded airbag in the bedding superimposed image.

4. The intelligent pressure-sensitive bedding adjustment method based on sleep image data according to claim 3, characterized in that: Segmenting the bedding superimposed image according to the value of the pressure data to obtain an image of the bedding pressure area specifically includes: In the bedding superimposed image, determining all pressurized folded airbags whose pressure data values are not 0; Determine the pixels within a preset range around each compressed folded airbag, and set the pixel values of the remaining pixels to 255 to complete the segmentation of the bedding superimposed image to obtain the compressed area image of the bedding; The preset range is a range enclosed by a square with the center of the cross-section of the folded airbag as the center point and a side length of twice the diameter of the cross-section of the folded airbag as the side length.

5. The intelligent pressure-sensitive bedding adjustment method based on sleep image data according to claim 1, characterized in that: Performing human body part recognition on the image of the pressure area of the bedding to determine the non-human body area specifically includes: determining a target area threshold according to a grayscale value distribution of the image of the pressure area of the bedding; Determining, in the image of the bedding pressure area, an area whose grayscale value is greater than the target area threshold as a target area; Divide the target area into several target blocks, use the gradient value of each pixel in each target block as a weight vector, and calculate the gradient histogram of different gradient directions in each target block; Normalize the gradient histograms of different gradient directions in each target block to obtain the gradient feature vector of each target block; Combining the gradient feature vectors of each target block into a gradient feature vector of the target region and inputting the gradient feature vectors into a classifier to classify each target block separately; The target blocks classified as human targets are connected to determine them as human body areas; and the area other than the human body area in the bedding pressure area image is determined as the non-human body area.

6. The intelligent pressure-sensitive bedding adjustment method based on sleep image data according to claim 5, characterized in that: Determining the target area threshold according to the grayscale value distribution of the image of the bedding pressure area specifically includes: Determining a grayscale value distribution interval in the image of the bedding pressure area; In the grayscale value distribution interval, a plurality of grayscale values are randomly selected as a plurality of initial cluster centers, and k-means clustering is performed on the image of the bedding pressure area; After clustering is completed, the grayscale value change data of each cluster center during the clustering process is analyzed to determine the cluster center where the grayscale value change data has a mutation and the grayscale value of the mutation point; The grayscale value of the mutation point is determined as the target area threshold.

7. The intelligent pressure-sensitive bedding adjustment method based on sleep image data according to claim 1, characterized in that: Analyzing and calculating the pressure matrix to determine the user's sleeping position specifically includes: In the pressure matrix, the area where the matrix elements with values greater than 0 are located is determined as the pressure area, and the number of matrix elements in the pressure area is counted; The number of matrix elements and the element scale of the pressure-sensitive area are used as the current pressure-sensitive area of the smart pressure-sensitive bedding; wherein the element scale is the ratio between one matrix element and the area occupied by one folding airbag in the smart pressure-sensitive bedding; Comparing the current pressure area with preset pressure area intervals for different posture types to preliminarily determine the posture type of the user; wherein the posture types include sitting, lying flat, and lying on the side; lying flat specifically includes lying on the back and lying on the stomach, and lying on the side specifically includes lying on the left side and lying on the right side; When the posture type is side-lying, obtaining peak values of several matrix element values in the pressure area, and fitting the peak values of the several matrix element values into a straight line to determine it as the force axis; Analyzing the number of matrix elements on both sides of the force axis in the pressure area, and determining the side with more matrix elements as the user facing side; When the posture type is lying flat, the average value of the element values in the pressure area is calculated, and the area formed by elements smaller than the average value of the element values is determined as a low-force area; if there is at least one low-force area in the longitudinal middle area of the pressure area, the user is determined to be in a supine posture, otherwise the user is determined to be in a prone posture.

8. The intelligent pressure-sensitive bedding adjustment method based on sleep image data according to claim 1, characterized in that: According to the sleeping posture, the intelligent air pump is controlled to inflate or deflat the airbag in the intelligent pressure-sensitive bedding, so that the intelligent pressure-sensitive bedding is intelligently adjusted according to the sleeping posture of the user, specifically comprising: Determining the height of the user according to the user's posture; Determining the position of each body part of the user according to the user's height and pressure data distribution; According to the positions of the various body parts, the compressed area is divided into at least a trunk area, a left upper limb area, a right upper limb area, a left lower limb area, and a right lower limb area; Obtain the minimum and maximum values of the elements in each partition respectively, and calculate the average value to obtain the pressure balance value corresponding to each partition; Slowly deflating the folding airbags corresponding to the elements with a pressure greater than the pressure balance value in the intelligent pressure-sensitive bedding, and slowly inflating the folding airbags corresponding to the elements with a pressure less than the pressure balance value; Real-time monitoring of the pressure value of each folded airbag in the pressurized area during inflation or deflating; When the error between the pressure value borne by the folding airbag and the corresponding pressure balance value enters the allowable error range, or the real-time deformation value reaches the boundary value of the expansion and contraction range corresponding to the current adjustment gear, the inflation or exhaust process is stopped; wherein different adjustment gears correspond to different expansion and contraction ranges of the folding airbag.

9. An intelligent pressure-sensitive bedding adjustment device based on sleep image data, characterized in that: The device comprises: 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, so that the at least one processor can execute the intelligent pressure-sensitive bedding adjustment method based on sleep image data according to any one of claims 1 to 8.

Images