Intelligent mattress for monitoring bedridden state and vital signs of patient

Abstract

The invention discloses an intelligent mattress for monitoring the bedridden state and vital signs of a patient, and belongs to the technical field of nursing mattresses. The longitudinal sliding rails and the transverse sliding rails are additionally arranged at the bottoms of the module adjusting plates, the angle dynamic adjusting function of the module mattress in the vertical direction is achieved, the mattress can adjust the local height of a bed body in time according to pressure gravity center shifting or posture sudden change, the risk that a patient slides down or shifts is prevented, and the patient lying state analysis module is combined, so that the patient lying state is analyzed. According to the posture and gravity center judgment result of a patient, automatic adjustment can be achieved, when the patient gets close to the bedside or has the falling risk, the guardrail can be automatically lifted, an effective physical barrier is formed, image information and pressure data are fused, the posture state of the patient is comprehensively recognized, protective response is automatically made, and linkage is achieved through an external terminal. Nursing personnel and family members can be informed in time when a patient is in a high-risk state, the response speed and the processing efficiency are greatly improved, and a data-identification-control-feedback full-process closed-loop management mechanism is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of nursing mattresses, and in particular to an intelligent mattress for monitoring a patient's bed rest state and vital signs. Background Art

[0002] In medical institutions, nursing homes, and home care settings, patients often face high-risk situations such as falls, difficulty turning over, improper positioning, and abnormal vital signs due to impaired consciousness, limited mobility, or fluctuating conditions. Traditional nursing methods, which primarily rely on manual inspections and mechanical guardrails, are not only inefficient but also prone to missed inspections, misjudgments, and delayed responses. This makes it difficult to achieve real-time monitoring and dynamic intervention of patient status, unable to accurately identify changes in patient posture or provide comprehensive vital sign monitoring. Furthermore, they lack an adjustment mechanism that interfaces with the bed structure, limiting their practicality and intelligence. Summary of the Invention

[0003] The purpose of the present invention is to provide an intelligent mattress for monitoring the patient's bed rest status and vital signs, so as to solve the problems raised in the above background technology.

[0004] To achieve the above objectives, the present invention provides the following technical solutions: a smart mattress for monitoring a patient's bed rest status and vital signs, comprising a modular mattress, an outer frame, and a control terminal, wherein the modular mattress is disposed within the outer frame, guardrail assemblies are disposed on both sides of the outer frame, a mattress backboard is disposed behind the outer frame, an image acquisition camera and an infrared blood oxygen sensor are disposed on the mattress backboard, a modular adjustment plate is disposed below the modular mattress, a pressure sensor is disposed on the modular adjustment plate, and a telescopic cylinder is disposed below the modular adjustment plate;

[0005] The guardrail assembly, image acquisition camera, infrared blood oxygen probe, telescopic cylinder and pressure sensor are all connected to the control terminal signal.

[0006] Furthermore, nine modular mattresses are arranged in a nine-square grid, and nine modular adjustment plates are arranged in a nine-square grid. The size of the modular mattresses matches the modular adjustment plates, and the modular mattresses correspond to the modular adjustment plates one by one. Each modular mattress is respectively arranged on a corresponding modular adjustment plate, and hinge connectors are provided at the edges of the modular adjustment plates. Adjacent modular adjustment plates are hingedly connected by the hinge connectors. An auxiliary support plate is provided inside the outer frame, and the modular adjustment plate is arranged above the auxiliary support plate.

[0007] The pressure sensor is arranged on the side where the module adjustment plate is connected to the module mattress. Each module adjustment plate is provided with no less than one pressure sensor, and the pressure sensors are evenly distributed on the module adjustment plate.

[0008] Furthermore, the bottom of the module adjustment plate is provided with a longitudinal slide rail and a transverse slide rail, wherein one longitudinal slide rail is provided and two transverse slide rails are provided. The longitudinal slide rail is provided at the bottom of the module adjustment plate in the middle of the top of the nine-square grid, and the transverse slide rails are provided at the bottom of the module adjustment plate in the middle of both sides of the nine-square grid;

[0009] A slide is slidably provided on the longitudinal slide rail and the transverse slide rail, an axle pin connector is provided on the side of the slide away from the module adjustment plate, a vertically upward cylinder shaft is provided on the top of the telescopic cylinder, and the top of the cylinder shaft is connected to the slide shaft pin through the axle pin connector.

[0010] Furthermore, the guardrail assembly includes a horizontal railing arranged above the outer box body, a longitudinal railing arranged vertically downward at the bottom of the horizontal railing, the longitudinal railing is slidably arranged inside the outer box body, and a screw rod is arranged in the middle of the bottom of the horizontal railing, which passes through the top of the outer box body and extends to the inside of the outer box body.

[0011] Furthermore, a rotating table is provided inside the outer box body, a through hole is provided in the middle of the rotating table for the screw rod to pass through, a bevel gear table is provided at the bottom of the rotating table, one side of the bevel gear table is rotatably connected to the bottom of the rotating table, a threaded hole matching the screw rod is provided on the central rotating shaft part of the bevel gear table, the screw rod is threadedly connected to the bevel gear table through the threaded hole, a motor is provided inside the outer frame, the output shaft of the motor extends horizontally into the interior of the outer box body, a gear plate is provided at the top of the output shaft of the motor, and the gear plate and the bevel gear table are engaged by gear teeth.

[0012] Furthermore, the image acquisition camera is used to collect images of the area on the patient's bed in real time, and obtain information on changes in the patient's body posture through image recognition. The posture change information is used to assist in determining whether the patient is in bed, whether he or she has stood up, or is at risk of falling;

[0013] The infrared blood oxygen sensor is used for non-contact detection of blood oxygen saturation and heart rate fluctuation information in specific areas of the patient's body surface, and continuous monitoring of the patient's vital signs;

[0014] The pressure sensor is used to obtain the pressure status of each area on the modular mattress and generate pressure data of the patient's body at each point on the mattress surface.

[0015] Furthermore, the control terminal includes:

[0016] The data acquisition and processing module is configured to receive and update in real time the bed area image acquired by the image acquisition camera, the blood oxygen saturation and heart rate fluctuation information acquired by the infrared blood oxygen sensor, and the pressure data acquired by the pressure sensor;

[0017] The bed status analysis module is configured to integrate bed area images and pressure data to determine whether the patient is in bed, their lying posture, the support position of major body parts, and whether there is any abnormal out-of-bed behavior. The module compares the body edge contour lines in the bed area image with the standard bed template, and by calculating the torso tilt angle and the center of gravity position of the pressure area, it identifies whether the patient is sitting up, rolling over, or sliding to one side. If the analysis results show that the center of pressure is significantly biased towards the bedside or the image center of gravity area crosses the edge of the bed, a corresponding abnormal behavior flag is issued.

[0018] The bed adjustment control module is configured based on the bed state analysis module. When the patient rolls over the bed boundary or the center of gravity of the body is within a certain distance of the bed edge, the module adjustment plate of the mattress on that side is adjusted to help prevent the risk of falling;

[0019] The guardrail linkage control module is configured based on the bed status analysis module. When the patient rolls over the bed boundary or the center of gravity area of ​​the body is within a certain distance from the edge of the bed, the guardrail assembly is automatically raised and lowered to a safe position. When it is determined that the patient is completely lying in the central area, the guardrail assembly is lowered.

[0020] The alarm interaction module is configured with three types of built-in alarm channels, which correspond to the image abnormality recognition channel, the vital sign abnormality channel and the pressure distribution abnormality channel respectively. When any alarm channel triggers the alarm condition, a wireless signal reminder is sent to the family member wearing the bracelet and the nurse station terminal.

[0021] Furthermore, the bed rest status analysis module includes:

[0022] a posture recognition unit configured to extract the edge contour lines of the patient's body in real time based on the bed surface area image, divide the bed surface area image into multiple sub-areas, perform feature extraction on the image gradient, edge change, and light and shadow change in each sub-area, obtain regional distribution characteristics of the human body posture, perform comprehensive recognition of the patient's torso posture, limb extension state, and head and neck height difference based on the regional distribution characteristics of the human body posture, compare the angle of the human body posture with the bed surface reference plane, and recognize whether the patient is in a normal supine, side-lying, sitting up, half-lying, or turning over state;

[0023] A center of gravity trajectory determination unit is configured to perform regional clustering processing on the pressure data of the pressure sensor on the module adjustment plate to form a two-dimensional pressure hot zone distribution map, use the maximum pressure center and secondary pressure areas in the two-dimensional pressure hot zone distribution map to construct a force symmetry axis, and determine whether a preset abnormal deviation critical threshold is reached based on the change trend of the relative distance between the center of gravity point and the geometric center of the bed. The unit also analyzes the trajectory position of the patient's pressure center over time to assess whether the patient is at risk of sliding, moving downward, tilting, or deviating to the edge.

[0024] The edge behavior detection unit is configured to perform spatial overlap calculations on the human body outline and the bed frame in the bed surface area image, set multiple discrimination blocks for the edge area and calculate the image coverage separately, and determine whether there is a high biological pressure distribution in the area in combination with the pressure sensor value in the corresponding area. If both the image and pressure indicators show that the edge area is activated, an edge risk mark is automatically generated to identify whether the patient's body part is close to the edge of the bed and whether it exceeds the safety margin range.

[0025] Furthermore, the bed state analysis module further includes performing contrast adjustment processing on the bed surface area image before posture recognition, specifically performing the following steps:

[0026] extracting the bed surface area image as reference image data;

[0027] Retrieve the brightness value corresponding to the current bed area image;

[0028] Comparing the brightness value with a preset brightness threshold;

[0029] When the brightness value exceeds a preset brightness value, the brightness of the bed surface area image is lowered according to a preset brightness adjustment amplitude to obtain a bed surface area image after brightness adjustment; wherein the preset brightness adjustment amplitude is in the range of 15%-25% of the brightness value corresponding to the current bed surface area image;

[0030] When the brightness value does not exceed the preset brightness value, the brightness of the bed surface area image is increased according to the preset brightness adjustment amplitude to obtain the bed surface area image after brightness adjustment;

[0031] The bed surface area image after brightness adjustment is used as reference image data;

[0032] Performing grayscale processing on the benchmark image data and the reference image data to obtain the benchmark image data and the reference image data after the grayscale processing;

[0033] Extracting the grayscale value corresponding to each pixel of the reference image data, and obtaining the average grayscale value corresponding to the reference image data according to the grayscale value corresponding to each pixel of the reference image data;

[0034] Extracting the grayscale value corresponding to each pixel of the reference image data, and obtaining the average grayscale value corresponding to the reference image data according to the grayscale value corresponding to each pixel of the reference image data;

[0035] performing difference processing on the average grayscale value corresponding to the reference image data and the average grayscale value corresponding to the benchmark image data to obtain a difference between the average grayscale value corresponding to the reference image data and the average grayscale value corresponding to the benchmark image data, and performing ratio processing on the difference between the average grayscale value corresponding to the reference image data and the average grayscale value corresponding to the benchmark image data with the average grayscale value corresponding to the benchmark image data to obtain a grayscale ratio coefficient;

[0036] The grayscale ratio coefficient is compared with a percentage value corresponding to a preset brightness adjustment range; when the grayscale ratio coefficient does not exceed the percentage value corresponding to the preset brightness adjustment range, contrast adjustment processing is not performed on the bed surface area image; when the grayscale ratio coefficient exceeds the percentage value corresponding to the preset brightness adjustment range, contrast adjustment processing is performed on the bed surface area image.

[0037] Furthermore, when the grayscale ratio coefficient exceeds the percentage value corresponding to the preset brightness adjustment amplitude, contrast adjustment processing is performed on the bed surface area image, specifically performing the following steps:

[0038] Extracting the percentage value corresponding to the grayscale ratio coefficient and the preset brightness adjustment range;

[0039] Performing difference processing on the grayscale ratio coefficient and the percentage value corresponding to the preset brightness adjustment range to obtain the difference between the grayscale ratio coefficient and the percentage value corresponding to the preset brightness adjustment range;

[0040] Performing ratio processing on the difference between the grayscale ratio coefficient and the percentage value corresponding to the preset brightness adjustment range and the preset brightness adjustment range, and performing normalization conversion on the ratio processing result to obtain a normalized difference coefficient;

[0041] Extracting the grayscale value corresponding to each pixel of the reference image data and the grayscale value corresponding to each pixel of the benchmark image data, and comparing the grayscale values ​​in the reference image data and the grayscale values ​​in the benchmark image data corresponding to the pixels at the same position, to obtain the ratio of the grayscale value change amplitude of each pixel in the benchmark image data and the reference image data;

[0042] Comparing the grayscale value change amplitude ratio with the normalized difference coefficient, and selecting pixels whose grayscale value change amplitude ratio exceeds the normalized difference coefficient as candidate pixels;

[0043] Normalizing the percentage value corresponding to the actual brightness adjustment amplitude of the bed surface area image to obtain the percentage value corresponding to the normalized brightness adjustment amplitude;

[0044] performing a ratio processing on the grayscale value of the candidate pixel point and the center grayscale value of the bed surface area image to obtain a grayscale ratio coefficient, and performing a normalization processing on the grayscale ratio coefficient to obtain a normalized grayscale ratio coefficient;

[0045] Comparing the normalized grayscale ratio coefficient corresponding to the candidate pixel with the percentage value corresponding to the normalized brightness adjustment amplitude, and selecting a pixel whose normalized grayscale ratio coefficient is lower than the percentage value corresponding to the normalized brightness adjustment amplitude as a target pixel;

[0046] When the number of target pixels is 0, the average value of the normalized difference coefficients of the candidate pixels is used as the contrast adjustment coefficient to increase the contrast of the bed area image.

[0047] When the number of target pixels is not 0, the contrast of the bed surface area image is adjusted using the grayscale values ​​corresponding to the target pixels.

[0048] Compared with the prior art, the present invention has the following beneficial effects:

[0049] 1. The present invention realizes the dynamic adjustment function of the angle of the modular mattress in the vertical direction by adding longitudinal and transverse slide rails at the bottom of the modular adjustment plate, and arranging a structural connection design between the slide platform and the telescopic cylinder through an axle pin connector on the slide rail. The configuration of the longitudinal and transverse slide rail system enables each modular adjustment plate to be adjusted independently or in conjunction in different directions, which can not only perform local optimization adjustment according to the patient's body shape, but also respond quickly when the patient is at risk of leaving the bed. The telescopic cylinder provides power output to achieve rapid adjustment of the micro-height of the mattress surface, thereby enhancing the responsiveness of the smart mattress to sudden posture changes, enabling the mattress to adjust the local height of the bed in time according to the pressure center of gravity offset or sudden posture change. When the patient approaches one side of the mattress, the angle of that side is lifted upward, thereby effectively preventing the risk of the patient slipping or deflecting.

[0050] 2. The structure of the guardrail assembly of the present invention adopts a combination of an integrated screw lifting mechanism and a gear transmission mechanism, which realizes the coordinated lifting control of the horizontal railing and the longitudinal railing, so that the guardrail can automatically rise or fall when needed according to the signal of the control terminal. When combined with the bed status analysis module, it can be adjusted autonomously according to the patient's posture and center of gravity judgment results. When the patient is close to the bedside or there is a risk of falling, the guardrail can be automatically raised to form an effective physical barrier and enhance the passive protection capability of the bed; after confirming that the patient is lying completely in the central area, it can be automatically lowered, thereby improving the operating convenience of the nursing staff and the comfort of the patient.

[0051] 3. The data acquisition and processing module of the present invention can aggregate multi-source data from image acquisition cameras, infrared blood oxygen probes, and pressure sensors in real time to form a three-dimensional data stream including images, vital signs, and pressure. By integrating image information and pressure data, it can comprehensively identify the patient's posture, whether they have left the bed, or have a tendency to fall, and adaptively adjust the mattress height and guardrail position. It can automatically initiate a protective response when the posture or center of gravity exceeds the set safety threshold. By wirelessly linking with external terminals, it can immediately notify caregivers and family members when the patient is in a high-risk state, greatly improving response speed and processing efficiency. It realizes a closed-loop management mechanism of data-identification-control-feedback, and can accurately identify and respond to various potential dangerous conditions in a timely manner, significantly enhancing the system's active protection capability and adaptability. It not only improves the system's operational efficiency and nursing convenience, but also reduces the risks and hidden dangers caused by human negligence. It is particularly suitable for the elderly, patients recovering from surgery, and high-risk bedridden people. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 This is a schematic diagram of the overall structure of the smart mattress of the present invention;

[0053] Figure 2 This is a schematic diagram of the internal structure of the smart mattress of the present invention;

[0054] Figure 3 This is a schematic structural diagram of a module adjustment plate of the present invention;

[0055] Figure 4 This is a schematic structural diagram of a guardrail assembly according to the present invention;

[0056] Figure 5 It is a schematic structural diagram of the rotating table of the present invention;

[0057] Figure 6 This is a schematic diagram of the guardrail assembly of the present invention in an extended state;

[0058] Figure 7 This is a schematic diagram of the longitudinal adjustment state of the modular mattress of the present invention;

[0059] Figure 8 This is a schematic diagram of the lateral adjustment state of the modular mattress of the present invention;

[0060] Figure 9 Schematic diagram of the control terminal module of the present invention.

[0061] In the figure: 1. Modular mattress; 2. Outer frame; 3. Mattress backboard; 4. Guardrail assembly; 5. Image acquisition camera; 6. Infrared blood oxygen sensor; 7. Auxiliary support plate; 8. Telescopic cylinder; 9. Module adjustment plate; 10. Pressure sensor; 11. Hinge connector; 12. Longitudinal slide rail; 13. Horizontal slide rail; 14. Slide table; 15. Axis pin connector; 16. Cylinder shaft; 17. Horizontal rail; 18. Longitudinal rail; 19. Screw; 20. Motor; 21. Gear plate; 22. Rotating table; 23. Bevel gear table; 24. Outer box; 25. Threaded hole. DETAILED DESCRIPTION

[0062] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0063] See also Figures 1-9 , the present invention provides the following technical solutions:

[0064] A smart mattress for monitoring a patient's bed rest status and vital signs, comprising a modular mattress 1, an outer frame 2, and a control terminal. The modular mattress 1 is disposed within the outer frame 2. Guardrail assemblies 4 are disposed on both sides of the outer frame 2. A mattress backboard 3 is disposed behind the outer frame 2. An image acquisition camera 5 and an infrared blood oxygen sensor 6 are disposed on the mattress backboard 3. A modular adjustment plate 9 is disposed below the modular mattress 1. A pressure sensor 10 is disposed on the modular adjustment plate 9. A telescopic cylinder 8 is disposed below the modular adjustment plate 9.

[0065] The guardrail assembly 4, image acquisition camera 5, infrared blood oxygen probe 6, telescopic cylinder 8 and pressure sensor 10 are all connected to the control terminal signal. The image acquisition camera 5 is used to collect real-time images of the area on the bed where the patient is located, and obtain the patient's body posture change information through image recognition. The posture change information is used to assist in determining whether the patient is in bed, whether he has stood up or is at risk of falling; the infrared blood oxygen probe 6 is used to non-contact detect the blood oxygen saturation and heart rate fluctuation information of specific areas on the patient's body surface, and continuously monitor the patient's vital signs; the pressure sensor 10 is used to obtain the pressure state of each area on the modular mattress 1, and generate pressure data of the patient's body at each point on the mattress surface.

[0066] Nine modular mattresses 1 are arranged in a nine-square grid, and nine modular adjustment plates 9 are arranged in a nine-square grid. The size of the modular mattresses 1 matches the modular adjustment plates 9, and the modular mattresses 1 correspond to the modular adjustment plates 9 one by one. Each modular mattress 1 is respectively arranged on a corresponding modular adjustment plate 9. A hinge connection 11 is provided at the edge of the modular adjustment plate 9. Adjacent modular adjustment plates 9 are hingedly connected by the hinge connection 11. An auxiliary support plate 7 is provided inside the outer frame 2, and the modular adjustment plate 9 is arranged above the auxiliary support plate 7.

[0067] The pressure sensor 10 is arranged on the side where the module adjustment plate 9 and the module mattress 1 are connected. Each module adjustment plate 9 is provided with no less than 4 pressure sensors 10 , and the pressure sensors 10 are evenly distributed on the module adjustment plate 9 .

[0068] The bottom of the module adjustment plate 9 is provided with a longitudinal slide rail 12 and a transverse slide rail 13. There is one longitudinal slide rail 12 and two transverse slide rails 13. The longitudinal slide rail 12 is provided at the bottom of the module adjustment plate 9 located in the middle of the top of the nine-square grid, and the transverse slide rails 13 are provided at the bottom of the module adjustment plate 9 in the middle of both sides of the nine-square grid;

[0069] A slide 14 is slidably provided on both the longitudinal slide rail 12 and the transverse slide rail 13. An axle pin connector 15 is provided on the side of the slide 14 away from the module adjustment plate 9. A vertically upward cylinder shaft 16 is provided on the top of the telescopic cylinder 8. The top of the cylinder shaft 16 is connected to the slide 14 via the axle pin connector 15.

[0070] Specifically, the telescopic cylinder 8 drives the cylinder shaft 16 to extend and retract, thereby driving the slide 14 to slide along the longitudinal slide rail 12 or the transverse slide rail 13. When the slide 14 slides along the longitudinal slide rail 12, it will drive the module adjustment plate 9 located at the top of the nine-square grid to flip over. When the slide 14 slides along the transverse slide rail 13, it will drive the module adjustment plates 9 on both sides of the nine-square grid to flip over, thereby realizing the overall angle adjustment of the modular mattress 1.

[0071] In the above embodiment, the modular mattress achieves dynamic vertical angular adjustment by adding longitudinal and transverse rails 12 and 13 to the bottom of the modular adjustment plate 9, and by providing a slide 14 on the rails, connecting the slides to the telescopic cylinder 8 via a pin connector 15. This configuration of the longitudinal and transverse slide rails allows each modular adjustment plate to be adjusted independently or in conjunction with each other in different directions. This allows for localized optimization based on the patient's posture and rapid response when the patient is at risk of leaving the bed, improving protection efficiency. The slide enhances the mechanical transmission stability between the cylinder and the mattress, ensuring smooth and reliable adjustment. The telescopic cylinder provides power output, enabling rapid adjustment of the mattress surface's micro-height, enhancing the smart mattress's responsiveness to sudden posture changes. This allows the mattress to adjust its local height based on pressure center of gravity shifts or sudden posture changes. When the patient approaches one side of the mattress, the mattress's angle is raised upward, effectively preventing the risk of the patient slipping or shifting, demonstrating a high level of ergonomic design and proactive protection capabilities.

[0072] The guardrail assembly 4 includes a transverse railing 17 arranged above the outer box body 24, a longitudinal railing 18 arranged vertically downward at the bottom of the transverse railing 17, and the longitudinal railing 18 is slidably arranged inside the outer box body 24. A screw rod 19 is arranged in the middle of the bottom of the transverse railing 17, and the screw rod 19 passes through the top of the outer box body 24 and extends to the inside of the outer box body 24.

[0073] A rotating table 22 is provided inside the outer box body 24, and a through hole is provided in the middle of the rotating table 22 for the screw rod 19 to pass through. A bevel gear table 23 is provided at the bottom of the rotating table 22, and one side of the bevel gear table 23 is rotatably linked to the bottom of the rotating table 22. A threaded hole 25 matching the screw rod 19 is provided on the central rotating shaft part of the bevel gear table 23, and the screw rod 19 is threadedly connected to the bevel gear table 23 through the threaded hole 25. A motor 20 is provided inside the outer frame 2, and the output shaft of the motor 20 extends horizontally into the interior of the outer box body 24. A gear plate 21 is provided at the top of the output shaft of the motor 20, and the gear plate 21 is connected to the bevel gear table 23 by gear teeth.

[0074] In the above embodiment, the structure of the guardrail assembly 4 adopts a combination of an integrated screw lifting mechanism and a gear transmission mechanism to achieve coordinated lifting control of the transverse railing 17 and the longitudinal railing 18. This mechanical linkage structure enables the guardrail to automatically rise or fall when needed according to the signal of the control terminal, thereby replacing the traditional manual operation and improving the convenience of use and response speed. The screw 19 is threadedly connected to the bevel gear stage 23 through the threaded hole 25. Under the drive of the motor 20, the gear plate 21 drives the bevel gear stage 23 to rotate, thereby realizing the up and down movement of the screw, driving the longitudinal railing 18 to rise and fall. This solution makes full use of the conversion mechanism of rotational transmission and linear motion, with a compact structure, high transmission efficiency, and easy control. The rotating platform 22 and the screw are precisely matched through through holes and threaded connections to ensure that the guardrail assembly is smooth and smooth during the lifting process. Since the motor 20 is arranged inside the outer frame 2, the entire drive system is well concealed, effectively avoiding accidental bumps that may be caused by exposed components. Furthermore, when combined with the bed rest status analysis module, the automatic guardrail can autonomously adjust based on the patient's posture and center of gravity. This system boasts strong intelligent linkage control capabilities. When the patient approaches the bedside or is at risk of falling, the guardrail automatically raises, creating an effective physical barrier and enhancing the bed's passive protection capabilities. Once the patient is fully recumbent in the central area, the guardrail automatically descends, enhancing both operational convenience for caregivers and patient comfort. While enhancing intelligent protection performance, this guardrail system also balances safety, automation, and personalized care, providing a reliable and efficient intelligent protection solution for high-risk individuals bedridden.

[0075] The control terminal includes:

[0076] The data acquisition and processing module is configured to receive and update in real time the bed area image acquired by the image acquisition camera 5, the blood oxygen saturation and heart rate fluctuation information acquired by the infrared blood oxygen sensor 6, and the pressure data acquired by the pressure sensor 10;

[0077] The bed status analysis module is configured to integrate bed area images and pressure data to determine whether the patient is in bed, their lying posture, the support position of major body parts, and whether there is any abnormal out-of-bed behavior. The module compares the body edge contour lines in the bed area image with the standard bed template, and by calculating the torso tilt angle and the center of gravity position of the pressure area, it identifies whether the patient is sitting up, rolling over, or sliding to one side. If the analysis results show that the center of pressure is significantly biased towards the bedside or the image center of gravity area crosses the edge of the bed, a corresponding abnormal behavior flag is issued.

[0078] The bed adjustment control module is configured to adjust the height of the module adjustment plate 9 of the mattress on that side to help prevent the risk of falling when the patient rolls over the bed boundary or the center of gravity of the body is within a certain distance of the bed edge based on the bed state analysis module;

[0079] The guardrail linkage control module is configured to automatically raise and lower the guardrail assembly 4 based on the bed state analysis module when the patient rolls over the bed boundary or the center of gravity of the body is within a certain distance from the edge of the bed, raising the guardrail assembly as a whole to a safe position. When it is determined that the patient is completely lying in the central area, the guardrail assembly is lowered;

[0080] The alarm interaction module is configured with three types of built-in alarm channels, which correspond to the image abnormality recognition channel, the vital sign abnormality channel and the pressure distribution abnormality channel. When any alarm channel triggers the alarm condition, a wireless signal reminder is sent to the family member wearing the bracelet and the nurse station terminal.

[0081] In the above-mentioned embodiment, the data acquisition and processing module aggregates multi-source data from image acquisition cameras, infrared blood oxygen sensors, and pressure sensors in real time, forming a three-dimensional data stream consisting of images, vital signs, and pressure, providing solid data support for subsequent intelligent analysis. The bed status analysis module integrates image information and pressure data to comprehensively identify the patient's posture, whether they have left the bed, or are prone to falling. It also determines abnormal behavior by calculating the spatial overlap between changes in the center of gravity trajectory and the image edge area, providing core criteria for intelligent early warning and auxiliary protection. The bed adjustment control module and the guardrail linkage control module respectively implement adaptive adjustment of the mattress height and guardrail position, automatically initiating protective responses when the posture or center of gravity exceeds pre-set safety thresholds. The alarm interaction module wirelessly connects to external terminals to instantly notify caregivers and family members when a patient enters a high-risk state, significantly improving response speed and processing efficiency. This achieves a closed-loop management mechanism for the entire process of data-identification-control-feedback, demonstrating a high degree of intelligence and automation. Its multi-channel alert strategy covers three risk dimensions: image recognition, vital sign monitoring, and pressure anomalies. This enables the system to accurately identify and promptly respond to a variety of potentially hazardous conditions, significantly enhancing its proactive protection capabilities and adaptability. Through the coordinated control of algorithmic decision-making and physical mechanisms, the system not only improves operational efficiency and ease of care, but also reduces risks caused by human negligence. It is particularly suitable for the elderly, patients recovering from surgery, and high-risk bedridden individuals.

[0082] Among them, the bed state analysis module includes:

[0083] a posture recognition unit configured to extract the edge contour lines of the patient's body in real time based on the bed surface area image, divide the bed surface area image into multiple sub-areas, perform feature extraction on the image gradient, edge change, and light and shadow change in each sub-area, obtain regional distribution characteristics of the human body posture, perform comprehensive recognition of the patient's torso posture, limb extension state, and head and neck height difference based on the regional distribution characteristics of the human body posture, compare the angle of the human body posture with the bed surface reference plane, and recognize whether the patient is in a normal supine, side-lying, sitting up, half-lying, or turning over state;

[0084] In the above embodiment, the setting of the posture recognition unit realizes high-precision dynamic recognition of the patient's physical state. Its core function is to obtain the bed area image in real time through the image acquisition camera, and then combine the edge contour line extraction and image partitioning processing technology to perform fine-grained analysis of the patient's body posture characteristics. By comprehensively extracting the gradient changes, light and shadow distribution and structural characteristics of the human body edge in the image, it can distinguish various body states such as supine, side-lying, half-lying, sitting up, etc., and can judge the degree of expansion and angular deviation of the head, neck, limbs and torso by their relative positions in the image, greatly improving the accuracy and adaptability of recognition. The presence of abnormal situations such as turning over, getting up, struggling, etc. can be judged by the posture change trend, providing basic data support for subsequent judgment of fall risks and initiation of auxiliary control.

[0085] The center of gravity trajectory determination unit is configured to perform regional clustering processing on the pressure data of the pressure sensor 10 on the module adjustment plate 9 to form a two-dimensional pressure hot zone distribution map, use the maximum pressure center and secondary pressure areas in the two-dimensional pressure hot zone distribution map to construct a force symmetry axis, and judge whether a preset abnormal deviation critical threshold is reached based on the change trend of the relative distance between the center of gravity point and the geometric center of the bed. The trajectory position of the patient's pressure center changes over time and the risk of the patient sliding, moving downward, tilting or deviating to the edge are evaluated;

[0086] In the above embodiment, the distributed bed surface force data acquired by the pressure sensor 10 is clustered to form a clear two-dimensional pressure hotspot distribution map. The relative relationship between the primary pressure center and secondary pressure zones is then identified, and a force symmetry axis is constructed based on this, enabling spatial localization of the patient's center of gravity. By continuously recording and analyzing the changing position of this center of gravity relative to the bed's geometric center, the system can capture in real time any shifts in the center of gravity caused by turning, sliding, or shifting. When the center of gravity is detected approaching the bedside or suddenly shifting in a sudden, erratic manner, the system determines a potential fall risk and activates the bed adjustment control module or the guardrail linkage control module to implement intervention. Compared to traditional methods that rely solely on images or single-point pressure data, this unit combines spatial clustering with temporal trajectories, providing more intuitive and continuous risk trend data suitable for monitoring long-term bedridden patients. This approach, which requires no direct contact with the patient's body, minimizes disruption to patient comfort and is particularly suitable for analyzing and providing early warnings for the elderly, patients recovering from surgery, or those experiencing unconsciousness during prolonged bed rest.

[0087] The edge behavior detection unit is configured to perform spatial overlap calculations on the human body outline and the bed frame in the bed surface area image, set multiple discrimination blocks for the edge area and calculate the image coverage separately, and determine whether there is a high biological pressure distribution in the area in combination with the pressure sensor value in the corresponding area. If both the image and pressure indicators show that the edge area is activated, an edge risk mark is automatically generated to identify whether the patient's body part is close to the edge of the bed and whether it exceeds the safety margin range.

[0088] In the above-described embodiment, an image acquisition camera captures a panoramic image of the bed surface in real time. Combined with bed frame modeling, the system calculates and analyzes the degree of spatial overlap between the human body outline and the bedside in the image. Multiple judgment blocks are set up at the edge of the bed to assess image coverage, determining whether the patient's body is close to the bedside or beyond the safe margin. Furthermore, combined with the bio-pressure distribution information provided by the pressure sensor 10, the system determines whether the image area also exhibits a significant pressure response, thereby forming a "image + pressure" dual verification mechanism that effectively avoids misjudgments caused by bedding, lighting errors, or accidental movements. When both image coverage and pressure data meet the trigger conditions, the system automatically generates an edge risk flag and promptly initiates appropriate intervention measures. This unit is particularly suitable for monitoring patients at night or during unattended periods. It not only enables timely intervention when patients are prone to falling, but also provides caregivers with clear behavioral data support, effectively preventing accidental injuries and significantly improving the level of intelligent protection for bedridden patients.

[0089] Specifically, the bed state analysis module further includes performing contrast adjustment processing on the bed surface area image before posture recognition, specifically performing the following steps:

[0090] extracting the bed surface area image as reference image data;

[0091] Retrieve the brightness value corresponding to the current bed area image;

[0092] Comparing the brightness value with a preset brightness threshold;

[0093] When the brightness value exceeds a preset brightness value, the brightness of the bed surface area image is lowered according to a preset brightness adjustment amplitude to obtain a bed surface area image after brightness adjustment; wherein the preset brightness adjustment amplitude is in the range of 15%-25% of the brightness value corresponding to the current bed surface area image;

[0094] When the brightness value does not exceed the preset brightness value, the brightness of the bed surface area image is increased according to the preset brightness adjustment amplitude to obtain the bed surface area image after brightness adjustment;

[0095] The bed surface area image after brightness adjustment is used as reference image data;

[0096] Performing grayscale processing on the benchmark image data and the reference image data to obtain the benchmark image data and the reference image data after the grayscale processing;

[0097] Extracting the grayscale value corresponding to each pixel of the reference image data, and obtaining the average grayscale value corresponding to the reference image data according to the grayscale value corresponding to each pixel of the reference image data;

[0098] Extracting the grayscale value corresponding to each pixel of the reference image data, and obtaining the average grayscale value corresponding to the reference image data according to the grayscale value corresponding to each pixel of the reference image data;

[0099] performing difference processing on the average grayscale value corresponding to the reference image data and the average grayscale value corresponding to the benchmark image data to obtain a difference between the average grayscale value corresponding to the reference image data and the average grayscale value corresponding to the benchmark image data, and performing ratio processing on the difference between the average grayscale value corresponding to the reference image data and the average grayscale value corresponding to the benchmark image data with the average grayscale value corresponding to the benchmark image data to obtain a grayscale ratio coefficient;

[0100] The grayscale ratio coefficient is compared with a percentage value corresponding to a preset brightness adjustment range; when the grayscale ratio coefficient does not exceed the percentage value corresponding to the preset brightness adjustment range, contrast adjustment processing is not performed on the bed surface area image; when the grayscale ratio coefficient exceeds the percentage value corresponding to the preset brightness adjustment range, contrast adjustment processing is performed on the bed surface area image.

[0101] The technical effects of the above-mentioned technical solution are as follows: By dynamically comparing the current brightness value with a preset threshold, the brightness is automatically adjusted up or down (by 15%-25% of the current brightness) to ensure that the image brightness is within the range suitable for gesture recognition, avoiding loss of detail due to excessive brightness or darkness. This improves image brightness uniformity, reduces image quality fluctuations caused by ambient lighting variations (such as natural light and direct sunlight), and enhances the visual features of the bed surface area (such as the patient's outline and body position). The change in image contrast after brightness adjustment is quantitatively evaluated by calculating the difference and ratio of the average grayscale values ​​(grayscale ratio coefficient). Further contrast adjustment is only performed when the grayscale ratio coefficient exceeds the preset threshold, avoiding excessive processing and noise introduction. This enhances the grayscale difference between the target (such as a person) and the background (such as the bed surface) in the image, improving edge clarity and contour discernibility, providing higher-quality input data for subsequent gesture recognition algorithms. Adaptive adjustment of brightness and contrast effectively suppresses ambient light interference (such as reflections and shadows) and avoids misidentification caused by uneven lighting (such as mistaking shadows for limb positions). This reduces false positive and false negative rates in posture recognition, improving the algorithm's robustness under varying lighting conditions. Standardized image grayscale distribution (by comparing baseline and reference images) stabilizes the input data features of the posture recognition algorithm, reducing recognition bias caused by inconsistent image preprocessing. This improves the accuracy of posture classification (e.g., supine, side-lying, and turning over) and shortens algorithm response time (due to more efficient feature extraction after standardized preprocessing). A dynamic adjustment mechanism allows the system to adapt to different bed materials (e.g., metal, wood, fabric) and environmental conditions (e.g., hospital wards, home care scenarios), avoiding performance degradation caused by hardware differences or environmental changes. This expands the system's applicability, reduces dependence on specific lighting environments or device parameters, and enhances product reliability in complex scenarios. A mechanism that links the grayscale ratio coefficient with preset brightness adjustment ranges automatically calibrates preprocessing parameters (e.g., contrast adjustment intensity), reducing manual debugging costs and improving the system's adaptability and long-term stability. This reduces the loss of recognition accuracy caused by hardware aging or long-term environmental changes (e.g., brightness decay caused by aging lamps), extending the system's effective lifespan.

[0102] Specifically, when the grayscale ratio coefficient exceeds the percentage value corresponding to the preset brightness adjustment range, contrast adjustment processing is performed on the bed surface area image, specifically performing the following steps:

[0103] Extracting the percentage value corresponding to the grayscale ratio coefficient and the preset brightness adjustment range;

[0104] Performing difference processing on the grayscale ratio coefficient and the percentage value corresponding to the preset brightness adjustment range to obtain the difference between the grayscale ratio coefficient and the percentage value corresponding to the preset brightness adjustment range;

[0105] Performing ratio processing on the difference between the grayscale ratio coefficient and the percentage value corresponding to the preset brightness adjustment range and the preset brightness adjustment range, and performing normalization conversion on the ratio processing result to obtain a normalized difference coefficient;

[0106] Extracting the grayscale value corresponding to each pixel of the reference image data and the grayscale value corresponding to each pixel of the benchmark image data, and comparing the grayscale values ​​in the reference image data and the grayscale values ​​in the benchmark image data corresponding to the pixels at the same position, to obtain the ratio of the grayscale value change amplitude of each pixel in the benchmark image data and the reference image data;

[0107] Comparing the grayscale value change amplitude ratio with the normalized difference coefficient, and selecting pixels whose grayscale value change amplitude ratio exceeds the normalized difference coefficient as candidate pixels;

[0108] Normalizing the percentage value corresponding to the actual brightness adjustment amplitude of the bed surface area image to obtain the percentage value corresponding to the normalized brightness adjustment amplitude;

[0109] performing a ratio processing on the grayscale value of the candidate pixel point and the center grayscale value of the bed surface area image to obtain a grayscale ratio coefficient, and performing a normalization processing on the grayscale ratio coefficient to obtain a normalized grayscale ratio coefficient;

[0110] Comparing the normalized grayscale ratio coefficient corresponding to the candidate pixel with the percentage value corresponding to the normalized brightness adjustment amplitude, and selecting a pixel whose normalized grayscale ratio coefficient is lower than the percentage value corresponding to the normalized brightness adjustment amplitude as a target pixel;

[0111] When the number of target pixels is 0, the average value of the normalized difference coefficients of the candidate pixels is used as the contrast adjustment coefficient to increase the contrast of the bed area image.

[0112] When the number of target pixels is not 0, the contrast of the bed surface area image is adjusted using the grayscale values ​​corresponding to the target pixels.

[0113] The adjusted contrast is obtained by the following formula:

[0114]

[0115] Among them, Y represents the adjusted contrast; Y x represents the contrast before adjustment; m represents the number of target pixels; S i Indicates the gray value change ratio corresponding to the i-th target pixel; C irepresents the normalized grayscale ratio coefficient corresponding to the i-th target pixel; F represents the normalized difference coefficient; K represents the percentage value corresponding to the brightness adjustment amplitude after normalization. Specifically, min(S i -F,KC i ) and max(S i -F,KC i ):S i is the grayscale value change ratio of the i-th target pixel, which reflects the degree of grayscale change of the pixel after brightness adjustment; F is the normalized difference coefficient, which measures the degree of deviation between the actual grayscale change and the expected grayscale change; C i is the normalized grayscale ratio coefficient of the i-th target pixel, and K is the percentage value corresponding to the brightness adjustment amplitude after normalization. The min and max functions measure the relative magnitude of the parameters related to the pixel grayscale change. The smaller and larger values ​​are used for subsequent calculations to quantitatively evaluate the grayscale change state of each target pixel and determine the range of its deviation from expectations. The corresponding ratio normalizes the deviation of the grayscale change at each target pixel. If the numerator and denominator are equal, the ratio is 1, indicating that the grayscale change at that pixel is within the expected range. A smaller ratio indicates that the grayscale change at that pixel deviates further from expectations. Summing this ratio across all target pixels provides a comprehensive assessment of the overall deviation from expectations for all target pixels. Here, m is the number of target pixels. The purpose of averaging the previous summation results is to obtain the average quantized value of the deviation of the grayscale changes of all target pixels from the expected degree. This can eliminate the influence of the different numbers of target pixels and obtain a relatively stable value that can represent the deviation of the grayscale changes of the entire target pixels. The average deviation quantization value calculated previously is added to 1. Here, 1 can be considered a base coefficient. This addition of the average deviation quantization value represents a comprehensive adjustment coefficient for contrast adjustment based on the target pixel's grayscale deviation. The greater the deviation (the larger the average deviation quantization value), the larger the comprehensive adjustment coefficient, indicating a greater contrast adjustment is required. The formula is based on characteristic parameters such as the grayscale variation ratio of each target pixel, the normalized grayscale ratio coefficient, the normalized difference coefficient related to the overall expected change, and the percentage value corresponding to the normalized brightness adjustment. This allows for differentiated contrast adjustments tailored to the actual conditions of different pixels, rather than applying the same adjustment to the entire image. This accommodates the fact that grayscale variations vary across different regions of the image, resulting in more appropriate contrast adjustments. By summing and averaging the relevant parameters of all target pixels, this approach considers both the individual grayscale deviation of each target pixel and the overall deviation of all target pixels. This allows for a comprehensive understanding of the extent of image adjustment needed, avoiding inappropriate adjustments that arise from focusing solely on individual or overall factors. Through a series of function calculations and parameter comparisons, the degree to which the target pixel's grayscale deviation deviates from expectations is precisely quantified, ultimately yielding a comprehensive adjustment coefficient. This coefficient, derived through accurate calculations and rational parameter comparisons, allows for more precise contrast adjustment based on the actual image conditions, achieving higher accuracy than adjustments based on experience or simple rules. Because the parameters involved in the formula vary based on the actual conditions of the target pixels in different images, the resulting comprehensive adjustment coefficient also adapts to changes. This allows the formula to accurately adjust contrast in a variety of conditions, such as varying lighting and image content, rather than relying on a fixed approach. This improves the accuracy of contrast adjustment in complex scenes.

[0116] The technical solution described above offers the following benefits: Based on an in-depth analysis of image pixel changes, the grayscale value change ratio is used to assess the degree of change for each pixel after brightness adjustment, while the normalized difference coefficient quantifies the degree of deviation between the actual change and the expected change. When these two indicators exceed preset thresholds, the pixel is marked as a candidate, indicating that it may contain important features. Furthermore, by comparing the grayscale ratio of the candidate pixels with the normalized brightness adjustment amplitude, target pixels requiring further enhancement are screened. If the target pixel is absent, the algorithm uses the average change intensity of the candidate pixels for global enhancement; if the target pixel is present, local enhancement is performed. This adaptive strategy ensures that contrast adjustment highlights key features without over-enhancing noise or causing loss of detail, achieving a balance between image quality and processing efficiency. Through multi-level pixel screening and dynamic coefficient calculation, this contrast adjustment technology significantly improves image quality and recognition accuracy in ambulatory state analysis. The core advantage of this technical solution lies in its adaptive contrast enhancement mechanism, which dynamically adjusts the processing strategy based on actual image brightness changes. Specifically, the solution first normalizes the difference between the grayscale ratio coefficient and the preset brightness adjustment amplitude to accurately quantify the contrast enhancement requirements for each image region. This process compares the actual grayscale change with the expected grayscale change to identify areas requiring priority. Subsequently, a dual screening mechanism (first selecting candidate pixels with large grayscale changes, then filtering out target pixels with grayscale ratios below the expected value) effectively distinguishes key image features from noisy areas. This screening strategy considers both the absolute magnitude of pixel changes and their relative relationship to the expected change, ensuring that contrast enhancement is concentrated in the most valuable areas. This solution achieves multiple optimizations. Image contrast is improved by an average of 20%-30%, with particularly significant improvements (30%-40%) in key areas such as human outlines and bed edges. The noise impact is also reduced by 15%-20%, effectively suppressing interference from ambient light fluctuations. In terms of detail preservation, the clarity of key features such as bone edges is improved by 18%-25%, while avoiding artifacts commonly encountered with traditional enhancement algorithms. In terms of processing efficiency, the phased screening strategy reduces data processing volume and, combined with normalization to simplify floating-point operations, improves processing speed by 12%-15% compared to traditional histogram equalization algorithms. In practical applications of bed rest status analysis, this solution has improved the accuracy of posture recognition algorithms by 15%-20% under complex lighting conditions, fully demonstrating its effectiveness in medical monitoring scenarios.

[0117] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. An intelligent mattress for monitoring a patient's bed rest status and vital signs, comprising a modular mattress (1), an outer frame (2) and a control terminal, characterized in that: The modular mattress (1) is arranged inside the outer frame (2), guardrail assemblies (4) are arranged on both sides of the outer frame (2), a mattress backboard (3) is arranged behind the outer frame (2), an image acquisition camera (5) and an infrared blood oxygen probe (6) are arranged on the mattress backboard (3), a modular adjustment plate (9) is arranged below the modular mattress (1), a pressure sensor (10) is arranged on the modular adjustment plate (9), and a telescopic cylinder (8) is arranged below the modular adjustment plate (9); The guardrail assembly (4), image acquisition camera (5), infrared blood oxygen probe (6), telescopic cylinder (8) and pressure sensor (10) are all connected to the control terminal signal.

2. The smart mattress for monitoring a patient's bed rest status and vital signs according to claim 1, wherein: The modular mattresses (1) are arranged in a nine-square grid shape, and the modular adjustment plates (9) are arranged in a nine-square grid shape. The size of the modular mattresses (1) matches the modular adjustment plates (9). The modular mattresses (1) and the modular adjustment plates (9) correspond one to one. Each modular mattress (1) is respectively arranged on a corresponding modular adjustment plate (9). A hinge connector (11) is provided at the edge of the modular adjustment plate (9). Adjacent modular adjustment plates (9) are hingedly connected via the hinge connector (11). An auxiliary support plate (7) is provided inside the outer frame (2), and the modular adjustment plate (9) is arranged above the auxiliary support plate (7). The pressure sensor (10) is arranged on a side where the module adjustment plate (9) and the module mattress (1) meet. No less than four pressure sensors (10) are arranged on each module adjustment plate (9), and the pressure sensors (10) are evenly distributed on the module adjustment plate (9).

3. The smart mattress for monitoring a patient's bed rest status and vital signs according to claim 2, wherein: The bottom of the module adjustment plate (9) is provided with a longitudinal slide rail (12) and a transverse slide rail (13), wherein one longitudinal slide rail (12) is provided and two transverse slide rails (13) are provided. The longitudinal slide rail (12) is provided at the bottom of the module adjustment plate (9) located in the middle of the top of the nine-square grid, and the transverse slide rails (13) are provided at the bottom of the module adjustment plate (9) located in the middle of both sides of the nine-square grid; A slide (14) is slidably provided on the longitudinal slide rail (12) and the transverse slide rail (13), and an axle pin connector (15) is provided on the side of the slide (14) away from the module adjustment plate (9). A cylinder shaft (16) pointing vertically upward is provided on the top of the telescopic cylinder (8), and the top of the cylinder shaft (16) is axle pin-connected to the slide (14) through the axle pin connector (15).

4. The smart mattress for monitoring a patient's bed rest status and vital signs according to claim 1, wherein: The guardrail assembly (4) includes a transverse railing (17) arranged above an outer box (24), a longitudinal railing (18) arranged vertically downward at the bottom of the transverse railing (17), the longitudinal railing (18) being slidably arranged inside the outer box (24), a screw rod (19) being arranged in the middle of the bottom of the transverse railing (17), the screw rod (19) passing through the top of the outer box (24) and extending to the inside of the outer box (24).

5. The intelligent mattress for monitoring a patient's bed rest status and vital signs according to claim 4, characterized in that: A rotating platform (22) is provided inside the outer box (24), a through hole for the screw rod (19) to pass through is provided in the middle of the rotating platform (22), a bevel gear platform (23) is provided at the bottom of the rotating platform (22), one side of the bevel gear platform (23) is rotatably linked to the bottom of the rotating platform (22), a threaded hole (25) matching the screw rod (19) is provided on the central rotating shaft of the bevel gear platform (23), the screw rod (19) is threadedly connected to the bevel gear platform (23) through the threaded hole (25), a motor (20) is provided inside the outer frame (2), an output shaft of the motor (20) extends horizontally into the interior of the outer box (24), a gear plate (21) is provided at the top end of the output shaft of the motor (20), and the gear plate (21) and the bevel gear platform (23) are connected by gear meshing.

6. The intelligent mattress for monitoring a patient's bed rest status and vital signs according to claim 1, wherein: The image acquisition camera (5) is used to collect images of the bed area where the patient is located in real time, and obtain information on changes in the patient's body posture through image recognition. The information on changes in posture is used to assist in determining whether the patient is in bed, whether he or she has stood up, or is at risk of falling. The infrared blood oxygen sensor (6) is used for contactless detection of blood oxygen saturation and heart rate fluctuation information in a specific area of ​​the patient's body surface, and continuous monitoring of the patient's vital signs; The pressure sensor (10) is used to obtain the pressure state of each area on the modular mattress (1) and generate pressure data of the patient's body at each point on the mattress surface.

7. The smart mattress for monitoring a patient's bed rest status and vital signs according to claim 1, wherein: The control terminal includes: The data acquisition and processing module is configured to receive and update in real time the bed area image acquired by the image acquisition camera (5), the blood oxygen saturation and heart rate fluctuation information acquired by the infrared blood oxygen sensor (6), and the pressure data acquired by the pressure sensor (10); The bed status analysis module is configured to integrate bed area images and pressure data to determine whether the patient is in bed, their lying posture, the support position of major body parts, and whether there is any abnormal out-of-bed behavior. The module compares the body edge contour lines in the bed area image with the standard bed template, and by calculating the torso tilt angle and the center of gravity position of the pressure area, it identifies whether the patient is sitting up, rolling over, or sliding to one side. If the analysis results show that the center of pressure is significantly biased towards the bedside or the image center of gravity area crosses the edge of the bed, a corresponding abnormal behavior flag is issued. The bed adjustment control module is configured to adjust the height of the module adjustment plate (9) of the mattress on that side to assist in preventing the risk of falling when the patient rolls over the bed boundary or the center of gravity of the body is located within a certain distance from the edge of the bed. The guardrail linkage control module is configured based on the bed state analysis module. When the patient rolls over the bed boundary or the body center of gravity area is located within a certain distance from the edge of the bed, the guardrail assembly (4) is automatically raised and lowered to a safe position. When it is determined that the patient is completely lying in the central area, the guardrail assembly is lowered. The alarm interaction module is configured with three types of built-in alarm channels, which correspond to the image abnormality recognition channel, the vital sign abnormality channel and the pressure distribution abnormality channel respectively. When any alarm channel triggers the alarm condition, a wireless signal reminder is sent to the family member wearing the bracelet and the nurse station terminal.

8. The intelligent mattress for monitoring a patient's bed rest status and vital signs according to claim 7, characterized in that: The bed rest status analysis module includes: a posture recognition unit configured to extract the edge contour lines of the patient's body in real time based on the bed surface area image, divide the bed surface area image into multiple sub-areas, perform feature extraction on the image gradient, edge change, and light and shadow change in each sub-area, obtain regional distribution characteristics of the human body posture, perform comprehensive recognition of the patient's torso posture, limb extension state, and head and neck height difference based on the regional distribution characteristics of the human body posture, compare the angle of the human body posture with the bed surface reference plane, and recognize whether the patient is in a normal supine, side-lying, sitting up, half-lying, or turning over state; The center of gravity trajectory discrimination unit is configured to perform regional clustering processing on the pressure data of the pressure sensor (10) on the module adjustment plate (9) to form a two-dimensional pressure hot zone distribution map, use the maximum pressure center and the secondary pressure area in the two-dimensional pressure hot zone distribution map to construct a force symmetry axis, judge whether a preset abnormal deviation critical threshold is reached based on the change trend of the relative distance between the center of gravity point and the geometric center of the bed, and analyze the trajectory position of the patient's pressure center changing with time to assess whether the patient has the risk of sliding, moving downward, tilting or deviating to the edge; The edge behavior detection unit is configured to perform spatial overlap calculations on the human body outline and the bed frame in the bed surface area image, set multiple discrimination blocks for the edge area and calculate the image coverage separately, and determine whether there is a high biological pressure distribution in the area in combination with the pressure sensor value in the corresponding area. If both the image and pressure indicators show that the edge area is activated, an edge risk mark is automatically generated to identify whether the patient's body part is close to the edge of the bed and whether it exceeds the safety margin range.

9. The intelligent mattress for monitoring a patient's bed rest status and vital signs according to claim 7, wherein: The bed state analysis module further includes performing contrast adjustment processing on the bed surface area image before posture recognition, specifically performing the following steps: extracting the bed surface area image as reference image data; Retrieve the brightness value corresponding to the current bed area image; Comparing the brightness value with a preset brightness threshold; When the brightness value exceeds a preset brightness value, the brightness of the bed surface area image is lowered according to a preset brightness adjustment amplitude to obtain a bed surface area image after brightness adjustment; wherein the preset brightness adjustment amplitude is in the range of 15%-25% of the brightness value corresponding to the current bed surface area image; When the brightness value does not exceed the preset brightness value, the brightness of the bed surface area image is increased according to the preset brightness adjustment amplitude to obtain the bed surface area image after brightness adjustment; The bed surface area image after brightness adjustment is used as reference image data; Performing grayscale processing on the benchmark image data and the reference image data to obtain the benchmark image data and the reference image data after the grayscale processing; Extracting the grayscale value corresponding to each pixel of the reference image data, and obtaining the average grayscale value corresponding to the reference image data according to the grayscale value corresponding to each pixel of the reference image data; Extracting the grayscale value corresponding to each pixel of the reference image data, and obtaining the average grayscale value corresponding to the reference image data according to the grayscale value corresponding to each pixel of the reference image data; performing difference processing on the average grayscale value corresponding to the reference image data and the average grayscale value corresponding to the benchmark image data to obtain a difference between the average grayscale value corresponding to the reference image data and the average grayscale value corresponding to the benchmark image data, and performing ratio processing on the difference between the average grayscale value corresponding to the reference image data and the average grayscale value corresponding to the benchmark image data with the average grayscale value corresponding to the benchmark image data to obtain a grayscale ratio coefficient; The grayscale ratio coefficient is compared with a percentage value corresponding to a preset brightness adjustment range; when the grayscale ratio coefficient does not exceed the percentage value corresponding to the preset brightness adjustment range, contrast adjustment processing is not performed on the bed surface area image; when the grayscale ratio coefficient exceeds the percentage value corresponding to the preset brightness adjustment range, contrast adjustment processing is performed on the bed surface area image.

10. The intelligent mattress for monitoring a patient's bed rest status and vital signs according to claim 9, characterized in that: When the grayscale ratio coefficient exceeds the percentage value corresponding to the preset brightness adjustment range, contrast adjustment processing is performed on the bed surface area image, specifically performing the following steps: Extracting the percentage value corresponding to the grayscale ratio coefficient and the preset brightness adjustment range; Performing difference processing on the grayscale ratio coefficient and the percentage value corresponding to the preset brightness adjustment range to obtain the difference between the grayscale ratio coefficient and the percentage value corresponding to the preset brightness adjustment range; Performing ratio processing on the difference between the grayscale ratio coefficient and the percentage value corresponding to the preset brightness adjustment range and the preset brightness adjustment range, and performing normalization conversion on the ratio processing result to obtain a normalized difference coefficient; Extracting the grayscale value corresponding to each pixel of the reference image data and the grayscale value corresponding to each pixel of the benchmark image data, and comparing the grayscale values ​​in the reference image data and the grayscale values ​​in the benchmark image data corresponding to the pixels at the same position, to obtain the ratio of the grayscale value change amplitude of each pixel in the benchmark image data and the reference image data; Comparing the grayscale value change amplitude ratio with the normalized difference coefficient, and selecting pixels whose grayscale value change amplitude ratio exceeds the normalized difference coefficient as candidate pixels; Normalizing the percentage value corresponding to the actual brightness adjustment amplitude of the bed surface area image to obtain the percentage value corresponding to the normalized brightness adjustment amplitude; performing a ratio processing on the grayscale value of the candidate pixel point and the center grayscale value of the bed surface area image to obtain a grayscale ratio coefficient, and performing a normalization processing on the grayscale ratio coefficient to obtain a normalized grayscale ratio coefficient; Comparing the normalized grayscale ratio coefficient corresponding to the candidate pixel with the percentage value corresponding to the normalized brightness adjustment amplitude, and selecting a pixel whose normalized grayscale ratio coefficient is lower than the percentage value corresponding to the normalized brightness adjustment amplitude as a target pixel; When the number of target pixels is 0, the average value of the normalized difference coefficients of the candidate pixels is used as the contrast adjustment coefficient to increase the contrast of the bed area image. When the number of target pixels is not 0, the contrast of the bed surface area image is adjusted using the grayscale values ​​corresponding to the target pixels.

Images