Sleeping position detection method, pillow and storage medium

Abstract

The application relates to the sleep field and discloses a sleep posture detection method, a pillow and a storage medium. The sleep posture detection method comprises the following steps: collecting a sampling signal of a sensor array in real time; determining a pressure distribution diagram of a head supporting area according to the sampling signal, wherein the pressure distribution diagram is used for representing the pressure size and the pressure area distribution of each position in the head supporting area; and judging whether an ear contour feature exists in the pressure distribution diagram, wherein if yes, it is determined that the current sleep posture is side sleeping, and if no, it is determined that the current sleep posture is supine sleeping. According to the application, the array sensor is used to acquire the pillow surface pressure distribution information, the unique pressure feature formed by the inherent anatomical structure of the ear can be captured, the feature has high stability and recognition degree, is not disturbed by the non-sleep posture switching actions such as turning over and moving, is not affected by the shift of the user's head placement position, and has the advantages of high recognition accuracy, strong anti-interference ability and strong adaptability.

Description

Technical Field

[0001] This application relates to the field of sleep, and more particularly to a method for detecting sleeping posture, as well as a pillow and storage medium. Background Technology

[0002] Currently, sleep monitoring using pillows is gradually being applied in everyday sleep scenarios. Existing technologies typically use pressure sensors or intensity sensors placed under the front of the pillow to detect the pressure or pressure changes exerted by the body's shoulders and back on the sensors to determine sleeping posture. However, this traditional approach simply relies on changes in pressure or intensity to determine posture, making it susceptible to misjudgments due to normal physiological movements such as turning over or shifting, resulting in low accuracy.

[0003] Therefore, improvements to existing technologies are necessary.

[0004] The above information is provided as background information only to aid in understanding this application and does not constitute an assertion or admission that any of the above content can be used as prior art relative to this application. Summary of the Invention

[0005] This application provides a sleeping posture detection method, a pillow, and a storage medium to solve the problem of low accuracy in sleeping posture recognition in the prior art.

[0006] To achieve the above objectives, this application provides the following technical solution:

[0007] In a first aspect, embodiments of this application provide a sleeping posture detection method applied to a pillow, wherein a sensor array is provided in the head support area of ​​the pillow; the sleeping posture detection method includes:

[0008] The sampling signals of the sensor array are acquired in real time;

[0009] Based on the sampling signal, a pressure distribution map of the head support area is determined. The pressure distribution map is used to characterize the pressure magnitude and pressure area distribution at each location within the head support area.

[0010] Determine whether there are ear contour features in the pressure distribution map. If so, determine that the current sleeping position is side sleeping; otherwise, determine that the current sleeping position is supine sleeping.

[0011] Optionally, determining whether ear contour features exist in the pressure distribution map includes:

[0012] Based on the pressure distribution map, the pressure-bearing area within the head support region is determined;

[0013] Within the pressure area, the features of the concha cavity, earlobe, and helix are identified;

[0014] Based on at least two of the concha cavity features, the earlobe features, and the helix features, an ear contour pattern is drawn.

[0015] The ear contour pattern is matched with a preset ear contour template. If they match, it is determined that the pressure distribution map contains ear contour features.

[0016] Optionally, the method for identifying the concha cavity features includes: extracting a first set of low-pressure regions within the pressure-bearing region, the first set of low-pressure regions including at least one low-pressure region whose pressure value is lower than a first preset pressure threshold, which are discretely distributed among each other; determining whether there exists a first target region in the first set of low-pressure regions whose geometry and size match the preset concha cavity shape and size; if so, determining the pressure distribution information corresponding to the first target region as the concha cavity feature; and / or,

[0017] The method for identifying the earlobe feature includes: extracting a second set of low-pressure regions within the pressure-bearing area, the second set of low-pressure regions including at least one discretely distributed low-pressure region whose pressure value is lower than a second preset pressure threshold; determining whether there exists a second target region in the second set of low-pressure regions whose area is smaller than a preset area threshold and whose geometric center is farther from the geometric center of the pressure-bearing area than a preset distance threshold; if so, determining the pressure distribution information corresponding to the second target region as the earlobe feature; and / or,

[0018] The method for identifying the helix feature includes: extracting a set of high-pressure regions from the pressure area, wherein the set of high-pressure regions includes at least one third target region whose pressure value is higher than a preset high-pressure threshold and is discretely distributed among each other; determining whether there is a third target region in the set of high-pressure regions that has an arc-shaped profile and whose curvature matches a preset range of helix curvature; if so, determining the pressure distribution information of the third target region as the helix feature.

[0019] Optionally, the step of performing similarity matching processing between the ear contour pattern and a preset ear contour template includes:

[0020] Extract the contour feature vector of the ear contour pattern. The contour feature vector includes at least one of the following feature parameters: geometric shape parameters of the contour, curvature distribution parameters of the contour boundary, area and aspect ratio parameters of the contour.

[0021] The similarity score is obtained by comparing the contour feature vector with the template feature vector of the preset ear contour template.

[0022] The similarity score is compared with a preset similarity threshold, and the ear contour pattern is judged to match the preset ear contour template based on the comparison result.

[0023] Optionally, the step of performing similarity matching processing between the ear contour pattern and the preset ear contour template further includes:

[0024] The preset similarity threshold is dynamically adjusted based on the overall area of ​​the compressed region at the current moment, and the preset similarity threshold is positively correlated with the overall area; and / or,

[0025] The preset similarity threshold is dynamically adjusted based on the pressure value fluctuation of the pressure distribution map at the current moment, and the preset similarity threshold is positively correlated with the pressure value fluctuation; and / or,

[0026] The preset similarity threshold is dynamically adjusted based on the similarity score in the previous detection period, and the preset similarity threshold is negatively correlated with the similarity score in the previous detection period.

[0027] Optionally, the sleeping posture detection method further includes:

[0028] When the current sleeping position is determined to be side sleeping, the offset direction of the ear contour pattern in the pressure area is determined;

[0029] If the geometric center of the ear contour pattern is offset in a first direction relative to the geometric center of the pressure area, and the offset in the first direction is greater than a preset offset threshold, then the current sleeping position is determined to be the left sleeping position.

[0030] If the geometric center of the ear contour pattern has an offset in a second direction relative to the geometric center of the pressure area, and the offset in the second direction is greater than the preset offset threshold, then the current sleeping position is determined to be the right sleeping position.

[0031] Wherein, the first direction is opposite to the second direction.

[0032] Secondly, embodiments of this application provide a pillow, including a pillow core, a sensor array arranged in the head support area of ​​the pillow core, and a controller;

[0033] The sensor array is arranged in the head support area of ​​the pillow core;

[0034] The controller is used to perform sleeping posture detection according to the sleeping posture detection method described above.

[0035] Optionally, the pillow core is further provided with a head airbag assembly and a neck airbag assembly;

[0036] The controller is also used to adjust the height of the head airbag assembly and / or neck airbag assembly based on the sleeping posture detection results.

[0037] Thirdly, embodiments of this application provide a computer-readable storage medium having computer-executable instructions stored thereon, which are executed by a computer processor to implement the sleeping posture detection method as described in any of the above claims.

[0038] Compared with the prior art, this application has the following beneficial effects:

[0039] Compared with traditional methods for detecting changes in shoulder and back pressure, the embodiments of this application obtain pillow pressure distribution information through array sensors, which can capture the unique pressure characteristics formed by the inherent anatomical structure of the ear. This feature has high stability and recognizability, is not affected by non-sleeping posture switching actions such as turning over or moving, and is not affected by the user's head position shift. It has the advantages of high recognition accuracy, strong anti-interference ability, and strong adaptability.

[0040] This application has other features and advantages that will be apparent from or will be set forth in detail in the accompanying drawings and following detailed description, which together serve to explain the particular principles of this application. Attached Figure Description

[0041] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0042] Figure 1 This is a flowchart of the sleeping posture detection method provided in the embodiments of this application.

[0043] Figure 2 This is a schematic diagram of the ear outline provided in an embodiment of this application. Detailed Implementation

[0044] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0045] To improve the accuracy of sleep posture detection results, this application provides a sleep posture detection method applied to a pillow with a sensor array installed in its head support area. Please refer to [reference needed]. Figure 1 The sleeping posture detection method in this embodiment includes:

[0046] S1. Real-time acquisition of sampling signals from the sensor array.

[0047] S2. Based on the sampling signal, determine the pressure distribution map of the head support area. The pressure distribution map is used to characterize the pressure magnitude and pressure area distribution at each location within the head support area.

[0048] S3. Determine whether there are ear contour features in the pressure distribution map. If so, determine that the current sleeping position is side sleeping; otherwise, determine that the current sleeping position is supine sleeping.

[0049] When sleeping on one's side, the ear rests against the pillow surface. Due to the unique concave structure of the auricle, the contact pattern between the ear and the pillow surface differs from other body parts, exhibiting a distinctive ear contour feature in the pressure distribution map. Conversely, when sleeping on one's back, the back of the head rests on the pillow, and the ear does not contact the pillow surface; therefore, the aforementioned ear contour feature is absent from the pressure distribution map. Based on this objective law determined by physiological structure, this application embodiment uses a sensor array positioned in the head support area of ​​the pillow to collect pressure signals in real time and generate a pressure distribution map. The presence or absence of ear contour features in the pressure distribution map determines the sleeping posture.

[0050] Compared with existing technologies that detect changes in shoulder and back pressure, this application embodiment obtains pillow pressure distribution information through an array of sensors, which can capture the unique pressure characteristics formed by the inherent anatomical structure of the ear. This characteristic has high stability and recognizability, is not affected by non-sleeping posture switching actions such as turning over or moving, and is not affected by the user's head position shift. It has the advantages of high recognition accuracy, strong anti-interference ability, and strong adaptability.

[0051] Please refer to Figure 2 The ear typically comprises three main structures: the concha, the earlobe, and the helix. When a person sleeps on their side on a pillow, the ear rests against the pillow surface. Due to the unique concave and convex structure of the auricle, the concha is recessed and cannot fully contact the pillow surface. The earlobe rests against the pillow surface, but the contact area is small. The helix, supported by cartilage, rests against the pillow surface, forming edge contact. These structures together create a unique pressure distribution pattern in the pressure distribution map, distinct from other parts of the body.

[0052] Therefore, in one optional implementation, step S3, determining whether there are ear contour features in the head support area, includes:

[0053] S31. Based on the pressure distribution information, determine the pressure-bearing area within the head support region;

[0054] S32. Within the pressure area, identify the features of the concha, earlobe, and helix;

[0055] S33. Based on at least two of the features of the concha cavity, the earlobe, and the helix, draw the ear outline pattern;

[0056] S34. Perform similarity matching between the ear contour pattern and the preset ear contour template. If they match, it is determined that there are ear contour features in the pressure distribution map.

[0057] By verifying at least two combinations of features from the concha, earlobe, and helix, the ear contour can be effectively distinguished from similar shapes that may be formed by other body parts (such as shoulders and cheeks) in the pressure distribution map, thereby further improving the accuracy and robustness of recognition.

[0058] In one optional implementation, this embodiment provides differentiated identification strategies based on the physiological structures and occipital surface contact characteristics of the concha, earlobe, and helix, respectively:

[0059] S321. Identify the features of the concha cavity.

[0060] The concha is a recessed structure in the center of the ear. When sleeping on your side, this area is suspended and does not contact the pillow surface, appearing as an isolated low-pressure area in a pressure distribution map. However, other non-ear-related factors on the pillow surface can also create low-pressure areas, such as localized areas of relaxation caused by uneven pillow material.

[0061] Therefore, this embodiment achieves precise positioning through a dual screening mechanism:

[0062] Extract a first set of low-pressure regions within the pressure zone. The first set of low-pressure regions includes at least one low-pressure region whose pressure value is lower than a first preset pressure threshold and is discretely distributed among each other.

[0063] Determine whether there exists a first target region in the first low-pressure region set whose geometry and size match the preset concha cavity shape and size. If so, determine the pressure distribution information corresponding to the first target region as a concha cavity feature.

[0064] Based on this, this embodiment ensures that the features of the concha are not missed, while effectively eliminating interference from low-pressure areas caused by non-ear factors, thus improving the accuracy and specificity of concha feature recognition.

[0065] S322. Identify earlobe features.

[0066] The earlobe is the soft tissue below the ear. When sleeping on your side, it rests against the pillow. However, due to its small contact area and soft tissue, the pressure it generates is relatively low and the area is small. Unlike the concha, the low-pressure area of ​​the earlobe is even smaller, located lower, and isolated.

[0067] Therefore, this embodiment achieves accurate identification through the following triple screening mechanism:

[0068] Extract a second set of low-pressure regions within the pressure zone. The second set of low-pressure regions includes at least one low-pressure region whose pressure value is lower than a second preset pressure threshold and is discretely distributed among each other.

[0069] Determine whether there exists a second target region in the set of second low-pressure regions whose area is smaller than a preset area threshold and whose geometric center is more than a preset distance threshold from the geometric center of the pressure region. If it exists, then the pressure distribution information corresponding to the second target region is determined as an earlobe feature.

[0070] This embodiment uses the triple constraints of "low pressure, small area and lower edge position" to accurately identify the earlobe from multiple low-pressure areas that may exist on the pillow surface, effectively eliminating low-pressure areas formed by non-earlobe factors such as unsupported cheeks and poor contact with the edge of the head.

[0071] S323. Identify the characteristics of the helix.

[0072] The helix is ​​a cartilaginous structure on the outer edge of the ear. When sleeping on your side, it presses against the pillow surface due to the support of the cartilage, forming a ring-shaped or arc-shaped edge contact pressure distribution. Unlike the usual circular high-pressure area, the helix is ​​special in that it is ring-shaped or C-shaped—surrounding the low-pressure cavity of the concha.

[0073] Therefore, this embodiment achieves accurate identification through the following method:

[0074] Extract a set of high-pressure regions from the pressure zone. The set of high-pressure regions includes at least one third target region whose pressure value is higher than a preset high-pressure threshold and is discretely distributed among each other.

[0075] Determine whether there exists a third target region in the high-pressure region set that has an arc-shaped profile and whose curvature matches the preset range of helix curvature. If so, determine the pressure distribution information of the third target region as a helix feature.

[0076] This embodiment uses the dual constraints of "high pressure and arc curvature" to accurately identify the helix from multiple high-pressure areas that may exist on the occipital surface, effectively eliminating high-pressure areas formed by non-helix structures such as cheeks and shoulders.

[0077] Furthermore, step S34, which involves performing similarity matching between the ear contour pattern and the preset ear contour template, may include:

[0078] S341. Extract the contour feature vector of the ear contour pattern. The contour feature vector includes at least one of the following feature parameters: the geometric shape parameter of the contour, the curvature distribution parameter of the contour boundary, and the area and aspect ratio parameter of the contour.

[0079] Among them, the geometric shape parameters of the contour can be used to describe the overall shape of the ear contour; the curvature distribution parameters of the contour boundary can reflect the detailed undulation characteristics of the auricle edge and distinguish the differences in auricle shape among different individuals; the area and aspect ratio parameters of the contour can be used to constrain the ear contour from the dimension of size and eliminate pseudo contours with obviously inconsistent sizes.

[0080] The aforementioned multidimensional features collectively characterize the ear contour from three levels: global morphology, local details, and overall size. Compared with single feature matching, it can more comprehensively and precisely describe the morphological features of the ear contour, avoiding mismatches caused by insufficient feature representation.

[0081] S342. Calculate the similarity between the contour feature vector and the template feature vector of the preset ear contour template to obtain a similarity score.

[0082] This step uses similarity calculation to ensure that the matching determination is objective and consistent.

[0083] When calculating similarity, different weights can be assigned to various feature parameters based on actual conditions such as confidence level, and the overall similarity score can be calculated based on different weights.

[0084] S343. Compare the similarity score with the preset similarity threshold, and determine whether the ear contour pattern matches the preset ear contour template based on the comparison result.

[0085] Through the above method, this embodiment can perform a more objective and quantitative matching and judgment on the drawn ear contour pattern, ensuring the reliability and accuracy of the sleeping posture detection results.

[0086] Considering that in actual sleep scenarios, the user's head position, pillow contact state, and sleep movements are constantly changing, a fixed preset similarity threshold is difficult to adapt to the recognition needs of different states. Setting the threshold too high may lead to missed detection of side sleeping state, while setting the threshold too low may easily misjudge noise or non-ear contours as side sleeping.

[0087] Therefore, in one optional implementation, step S34 further includes dynamically adjusting a preset similarity threshold based on the current detection state to achieve an adaptive balance between sensitivity and specificity. Specifically, this includes:

[0088] (1) The preset similarity threshold is dynamically adjusted according to the overall area of ​​the pressure area at the current moment. The preset similarity threshold is positively correlated with the overall area.

[0089] The area of ​​the pressure-bearing region is an important indicator of the degree of contact between the head and the pillow surface. When the area of ​​the pressure-bearing region is large, it indicates that the user's head is in full contact with the pillow surface, the ear is in sufficient contact with the pillow surface, the pressure distribution information is rich, and the extracted ear contour pattern has a high degree of completeness. In this case, it is advisable to increase the threshold to avoid misjudging noise or non-ear features as ear contours.

[0090] When the area of ​​pressure is small, it indicates that the user's head may be in a state of being lifted off or only partially in contact with the pillow surface, and the contact between the ear and the pillow surface is insufficient. The extracted ear contour pattern may be incomplete. In this case, the threshold can be lowered to avoid missing the side sleeping state due to insufficient contour information.

[0091] (2) The preset similarity threshold is dynamically adjusted according to the pressure value fluctuation of the pressure distribution map at the current time. The preset similarity threshold is positively correlated with the pressure value fluctuation.

[0092] The degree of pressure fluctuation reflects the stability of the current detection environment and the data quality. When the pressure fluctuation is large (e.g., the user is turning over, moving, or the pillow material is deformed), a large number of noise points caused by dynamic disturbances may appear in the pressure distribution map. These noise points may occasionally form shapes similar to the ear contour. In this case, increasing the threshold can reduce false triggers. When the pressure fluctuation is small, it indicates that the user is in a stable sleep state, and the pressure distribution data is relatively more realistic and reliable. In this case, lowering the threshold helps to improve the sensitivity of capturing side-sleeping states while ensuring accuracy. Based on this, the anti-interference ability can be effectively improved.

[0093] (3) The preset similarity threshold is dynamically adjusted based on the similarity score in the previous detection period. The preset similarity threshold is negatively correlated with the similarity score in the previous detection period.

[0094] Sleep posture exhibits continuity and stability over a short period, without frequent abrupt changes. A high similarity score in the previous detection period indicates that the user was previously identified as sleeping on their side, and the ear contour matches the template well, suggesting the user is likely continuing in this position. In this case, appropriately lowering the threshold helps maintain consistency in side-sleeping determination, preventing frequent jumps in posture assessment results due to minor perturbations in a single frame (such as slight hair movement or unconscious head movements), which would negatively impact user experience. Conversely, a low similarity score in the previous detection period indicates that the user was not previously identified as sleeping on their side. In this case, the threshold should be increased to prevent misjudgments based on false contours caused by occasional noise. Therefore, this approach can suppress jumps in posture assessment results to a certain extent, ensuring the temporal continuity and stability of the detection results.

[0095] Of course, it is understandable that in practical applications, a reasonable default value, dynamic adjustment range, and single adjustment magnitude can be set for the preset similarity threshold based on actual verification results, so as to ensure the stability and reliability of dynamic adjustment.

[0096] Furthermore, based on the aforementioned determination of side sleeping, this embodiment can further utilize the offset direction of the ear contour pattern within the pressure area to distinguish between left-sided and right-sided sleeping positions, specifically including:

[0097] When the current sleeping position is determined to be side sleeping, the offset direction of the ear contour pattern within the pressure area is determined; if the geometric center of the ear contour pattern has an offset in a first direction relative to the geometric center of the pressure area, and the offset in the first direction is greater than a preset offset threshold, then the current sleeping position is determined to be left-side sleeping; if the geometric center of the ear contour pattern has an offset in a second direction relative to the geometric center of the pressure area, and the offset in the second direction is greater than a preset offset threshold, then the current sleeping position is determined to be right-side sleeping; wherein, the first direction and the second direction are opposite.

[0098] By using the above-mentioned bias direction judgment, we can further accurately distinguish between left and right sleeping positions based on the side sleeping judgment, thus meeting the needs of refined sleep monitoring.

[0099] Secondly, embodiments of this application provide a pillow, including a pillow core, a sensor array arranged in the head support area of ​​the pillow core, and a controller;

[0100] A sensor array, arranged in the head support area of ​​the pillow core, can specifically be a piezoresistive sensor;

[0101] A controller for performing sleep posture detection according to any embodiment of the sleep posture detection method.

[0102] In addition, the pillow may also contain a head airbag assembly and a neck airbag assembly; the controller is also used to adjust the height of the head airbag assembly and / or the neck airbag assembly according to the sleeping posture detection results.

[0103] The specific adjustment strategy for the airbag assembly can be configured according to actual needs, and this embodiment does not impose specific limitations. For example, based on the user's head and neck support needs in different sleeping positions, target heights for the head airbag and neck airbag can be preset for supine sleeping, left-side sleeping, and right-side sleeping, respectively, and the airbag assembly can be adaptively adjusted based on the sleeping position detection results.

[0104] Thirdly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the sleeping posture detection method provided in all embodiments of this application.

[0105] Any combination of one or more computer-readable media may be used. A computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. A computer-readable storage medium can be, for example—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this document, a computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in connection with an instruction execution system, apparatus, or device.

[0106] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, capable of sending, propagating, or transmitting programs for use by or in connection with an instruction execution system, apparatus, or device.

[0107] Program code contained on a computer-readable medium may be transmitted using any suitable medium, including but not limited to wireless, wire, optical fiber, RF, or any suitable combination thereof.

[0108] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0109] Finally, it should be noted that although the above embodiments have been described in the text and drawings of this application, this should not limit the scope of patent protection of this application. Any technical solutions that are based on the essential concept of this application and utilize the content described in the text and drawings of this application, resulting in equivalent structural or procedural substitutions or modifications, as well as the direct or indirect application of the technical solutions of the above embodiments to other related technical fields, are all included within the scope of patent protection of this application.

Claims

1. A method for detecting sleeping posture, characterized in that, Applied to pillows, the head support area of ​​the pillow is provided with a sensor array; The sleeping posture detection method includes: The sampling signals of the sensor array are acquired in real time; Based on the sampling signal, a pressure distribution map of the head support area is determined. The pressure distribution map is used to characterize the pressure magnitude and pressure area distribution at each location within the head support area. Determine whether there are ear contour features in the pressure distribution map. If so, determine that the current sleeping position is side sleeping; otherwise, determine that the current sleeping position is supine sleeping.

2. The sleeping posture detection method according to claim 1, characterized in that, The determination of whether ear contour features exist in the pressure distribution map includes: Based on the pressure distribution map, the pressure-bearing area within the head support region is determined; Within the pressure area, the features of the concha cavity, earlobe, and helix are identified; Based on at least two of the concha cavity features, the earlobe features, and the helix features, an ear contour pattern is drawn. The ear contour pattern is matched with a preset ear contour template. If they match, it is determined that the pressure distribution map contains ear contour features.

3. The sleeping posture detection method according to claim 2, characterized in that, The method for identifying the concha cavity features includes: extracting a first set of low-pressure regions within the pressure-bearing region, the first set of low-pressure regions including at least one low-pressure region whose pressure value is lower than a first preset pressure threshold; determining whether there exists a first target region in the first set of low-pressure regions whose geometry and size match the preset concha cavity shape and size; if so, determining the pressure distribution information corresponding to the first target region as the concha cavity feature; and / or, The method for identifying the earlobe feature includes: extracting a second set of low-pressure regions within the pressure-bearing area, the second set of low-pressure regions including at least one discretely distributed low-pressure region whose pressure value is lower than a second preset pressure threshold; determining whether there exists a second target region in the second set of low-pressure regions whose area is smaller than a preset area threshold and whose geometric center is farther from the geometric center of the pressure-bearing area than a preset distance threshold; if so, determining the pressure distribution information corresponding to the second target region as the earlobe feature; and / or, The method for identifying the helix feature includes: extracting a set of high-pressure regions from the pressure area, wherein the set of high-pressure regions includes at least one third target region whose pressure value is higher than a preset high-pressure threshold and is discretely distributed among each other; determining whether there is a third target region in the set of high-pressure regions that has an arc-shaped profile and whose curvature matches a preset range of helix curvature; if so, determining the pressure distribution information of the third target region as the helix feature.

4. The sleeping posture detection method according to claim 2, characterized in that, The step of performing similarity matching between the ear contour pattern and the preset ear contour template includes: Extract the contour feature vector of the ear contour pattern. The contour feature vector includes at least one of the following feature parameters: geometric shape parameters of the contour, curvature distribution parameters of the contour boundary, area and aspect ratio parameters of the contour. The similarity score is obtained by comparing the contour feature vector with the template feature vector of the preset ear contour template. The similarity score is compared with a preset similarity threshold, and the ear contour pattern is judged to match the preset ear contour template based on the comparison result.

5. The sleeping posture detection method according to claim 4, characterized in that, The step of performing similarity matching between the ear contour pattern and the preset ear contour template further includes: The preset similarity threshold is dynamically adjusted based on the overall area of ​​the compressed region at the current moment, and the preset similarity threshold is positively correlated with the overall area; and / or, The preset similarity threshold is dynamically adjusted based on the pressure value fluctuation of the pressure distribution map at the current moment, and the preset similarity threshold is positively correlated with the pressure value fluctuation; and / or, The preset similarity threshold is dynamically adjusted based on the similarity score in the previous detection period, and the preset similarity threshold is negatively correlated with the similarity score in the previous detection period.

6. The sleeping posture detection method according to claim 2, characterized in that, The sleeping posture detection method also includes: When the current sleeping position is determined to be side sleeping, the offset direction of the ear contour pattern in the pressure area is determined; If the geometric center of the ear contour pattern is offset in a first direction relative to the geometric center of the pressure area, and the offset in the first direction is greater than a preset offset threshold, then the current sleeping position is determined to be the left sleeping position. If the geometric center of the ear contour pattern has an offset in a second direction relative to the geometric center of the pressure area, and the offset in the second direction is greater than the preset offset threshold, then the current sleeping position is determined to be the right sleeping position. Wherein, the first direction is opposite to the second direction.

7. A pillow, characterized in that, Includes a pillow core, a sensor array arranged in the head support area of ​​the pillow core, and a controller; The sensor array is arranged in the head support area of ​​the pillow core; The controller is used to perform sleeping posture detection according to any one of claims 1-6.

8. The pillow according to claim 7, characterized in that, The pillow core is also equipped with a head airbag assembly and a neck airbag assembly. The controller is also used to adjust the height of the head airbag assembly and / or neck airbag assembly based on the sleeping posture detection results.

9. The pillow according to claim 7, characterized in that, The sensor array includes multiple piezoresistive sensors.

10. A computer-readable storage medium having computer-executable instructions stored thereon, characterized in that, The computer-executable instructions are executed by a computer processor to implement the sleeping posture detection method as described in any one of claims 1-6.

Images