A sitting posture correction method based on facial features and a sitting posture correction desk lamp

Abstract

The application provides a sitting posture correction method based on facial features and a sitting posture correction desk lamp. The desk lamp is provided with a camera module for collecting first image data, a feature extraction module, a sitting posture construction module, a judgment module, a reminding module and an adjustment module. The sitting posture construction module is used for constructing a standard sitting posture model. The feature extraction module is used for extracting facial feature data of a user from the first image data. The judgment module is used for judging whether the sitting posture of the user is standard according to the facial feature data and the standard sitting posture model. The reminding module is used for sending a reminding signal when the sitting posture of the user is not standard. The adjustment module is used for adjusting the working parameters of the desk lamp when the sitting posture of the user is not standard and the user does not adjust the sitting posture within a preset time after the reminding signal is sent. The application judges whether the sitting posture is standard by positioning facial features, which can save computing resources, is instant and saves power consumption compared with face recognition or bone shape recognition.

Description

Technical Field

[0001] This invention relates to the field of smart terminal technology, specifically to a posture correction method based on facial features and a posture correction desk lamp. Background Technology

[0002] With economic development, the pressure of life, work, and study is increasing, and prolonged desk work / study has become commonplace for many. Nearsightedness, cervical spondylosis, and hunchback are among the problems plaguing people. Because users often neglect posture when reading and writing, maintain an insufficient distance between their eyes and the desktop, and obstruct lighting, these are significant contributing factors to nearsightedness and skeletal deformities. Therefore, how to prevent nearsightedness, cervical spondylosis, and other related diseases has become an important issue for people.

[0003] In recent years, with the development of computer technology, especially computer vision technology, posture recognition methods based on vision and image processing technologies have emerged. Some methods use the size and positional relationship of the area occupied by a face in a video to identify posture, while others use depth sensors to collect human skeletal information and identify posture based on the angle of skeletal nodes. However, these posture recognition methods have poor adaptability to the diversity of human postures and are computationally complex. Currently, desk lamps are commonly used lighting tools for reading and writing, and are almost standard equipment for these activities.

[0004] There is an urgent need for a desk lamp that uses computer vision technology to correct sitting posture by extracting facial features. Summary of the Invention

[0005] Based on the above problems, this invention proposes a posture correction method and a posture correction desk lamp based on facial features. It determines whether the posture is standard by locating facial feature points. Compared with face recognition or skeletal shape recognition, it can save computing resources, is real-time and saves power.

[0006] In view of this, one aspect of the present invention proposes a posture correction desk lamp based on facial features, comprising: a camera module for acquiring first image data, a feature extraction module, a posture construction module, a judgment module, a reminder module, and an adjustment module;

[0007] The sitting posture construction module is used to construct a standard sitting posture model;

[0008] The feature extraction module is used to extract the user's facial feature data from the first image data;

[0009] The judgment module is used to determine whether the user's sitting posture is standard based on the facial feature data and the standard sitting posture model.

[0010] The reminder module is used to issue a reminder signal when the user's sitting posture is not standard;

[0011] The adjustment module is used to adjust the operating parameters of the desk lamp when the user's sitting posture is not standard and the user does not adjust their sitting posture within a preset time after the reminder signal is issued.

[0012] Optionally, the sitting posture construction module is used to construct a standard sitting posture model, specifically:

[0013] Obtain standard sitting posture requirements data for multiple application scenarios from cloud servers;

[0014] Based on the standard sitting posture requirements data in multiple application scenarios, a sitting posture input instruction for each scenario is generated and broadcast to the user, reminding the user to adjust to the standard sitting posture according to the requirements.

[0015] Record the standard sitting posture corresponding to each application scenario and establish a one-to-one correspondence between them to construct the user's standard sitting posture model, which includes standard facial feature vectors.

[0016] Optionally, the feature extraction module is used to extract the user's facial feature data from the first image data, including:

[0017] Extract facial feature point data from users;

[0018] The facial feature point data is processed to obtain facial feature data.

[0019] Optionally, the judgment module is used to determine whether the user's sitting posture is correct based on the facial feature data and the standard sitting posture model, specifically:

[0020] Determine the current application scenario in which the user is located;

[0021] Determine a first standard sitting posture model corresponding to the current application scenario from the standard sitting posture model;

[0022] The current facial feature vector is obtained based on the facial feature data and compared with the first standard facial feature vector in the first standard sitting posture model;

[0023] When the difference between the current facial feature vector and the first standard facial feature vector is within a first preset range, the user's sitting posture standard is determined.

[0024] When the difference between the current facial feature vector and the first standard facial feature vector is not within a first preset range, it is determined that the user's sitting posture is not standard.

[0025] Optionally, a working model building module is also included for:

[0026] Obtain working data of the smart desk lamp corresponding to various sitting postures from the cloud server;

[0027] Establish working models corresponding to various sitting postures;

[0028] The adjustment module is used to adjust the operating parameters of the desk lamp when the user's posture is incorrect and the user does not adjust their posture within a preset time after the reminder signal is issued. Specifically:

[0029] Obtain the user's current sitting posture data;

[0030] The corresponding first working model is determined from the working model based on the current sitting posture data;

[0031] Adjust the operating parameters of the smart desk lamp according to the first working model.

[0032] Another aspect of the present invention provides a posture correction method based on facial features. The posture correction system includes a camera module for acquiring first image data, a feature extraction module, a posture construction module, a judgment module, a reminder module, and an adjustment module. The posture correction method includes:

[0033] The sitting posture construction module constructs a standard sitting posture model;

[0034] The feature extraction module extracts the user's facial feature data from the first image data;

[0035] The judgment module determines whether the user's sitting posture is standard based on the facial feature data and the standard sitting posture model.

[0036] The reminder module sends a reminder signal when the user's sitting posture is not standard;

[0037] The adjustment module adjusts the operating parameters of the desk lamp when the user's sitting posture is not standard and the user does not adjust their posture within a preset time after the reminder signal is issued.

[0038] Optionally, the step of the sitting posture construction module constructing a standard sitting posture model includes:

[0039] Obtain standard sitting posture requirements data for multiple application scenarios from cloud servers;

[0040] Based on the standard sitting posture requirements data in multiple application scenarios, a sitting posture input instruction for each scenario is generated and broadcast to the user, reminding the user to adjust to the standard sitting posture according to the requirements.

[0041] Record the standard sitting posture corresponding to each application scenario and establish a one-to-one correspondence between them to construct the user's standard sitting posture model, which includes standard facial feature vectors.

[0042] Optionally, the step of the feature extraction module extracting the user's facial feature data from the first image data includes:

[0043] Extract facial feature point data from users;

[0044] The facial feature point data is processed to obtain facial feature data.

[0045] Optionally, the step of the judgment module determining whether the user's sitting posture is correct based on the facial feature data and the standard sitting posture model includes:

[0046] Determine the current application scenario in which the user is located;

[0047] Determine a first standard sitting posture model corresponding to the current application scenario from the standard sitting posture model;

[0048] The current facial feature vector is obtained based on the facial feature data and compared with the first standard facial feature vector in the first standard sitting posture model;

[0049] When the difference between the current facial feature vector and the first standard facial feature vector is within a first preset range, the user's sitting posture standard is determined.

[0050] When the difference between the current facial feature vector and the first standard facial feature vector is not within a first preset range, it is determined that the user's sitting posture is not standard.

[0051] Optionally, a working model building module is also included for:

[0052] Obtain working data of the smart desk lamp corresponding to various sitting postures from the cloud server;

[0053] Establish working models corresponding to various sitting postures;

[0054] The adjustment module is used to adjust the operating parameters of the desk lamp when the user's posture is incorrect and the user does not adjust their posture within a preset time after the reminder signal is issued. Specifically:

[0055] Obtain the user's current sitting posture data;

[0056] The corresponding first working model is determined from the working model based on the current sitting posture data;

[0057] Adjust the operating parameters of the smart desk lamp according to the first working model.

[0058] The technical solution of this invention includes a facial feature-based posture correction desk lamp with a camera module for acquiring first image data, a feature extraction module, a posture construction module, a judgment module, a reminder module, and an adjustment module. The posture construction module constructs a standard posture model. The feature extraction module extracts the user's facial feature data from the first image data. The judgment module determines whether the user's posture is standard based on the facial feature data and the standard posture model. The reminder module issues a reminder signal when the user's posture is not standard. The adjustment module adjusts the lamp's operating parameters when the user's posture is not standard and the user does not adjust their posture within a preset time after the reminder signal is issued. This invention determines posture standardness by locating facial feature points, which saves significant computational resources compared to face recognition or skeletal morphology recognition. Therefore, it does not require high-performance hardware, is instantaneous, and power-efficient. Attached Figure Description

[0059] Figure 1 This is a schematic block diagram of a table lamp provided in one embodiment of the present invention;

[0060] Figure 2 This is a flowchart of a posture correction method provided in another embodiment of the present invention. Detailed Implementation

[0061] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0062] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0063] The terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or apparatuses.

[0064] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0065] The following reference Figures 1 to 2 This invention describes a posture correction method based on facial features and a posture correction desk lamp, provided by some embodiments of the present invention.

[0066] like Figure 1 As shown, one embodiment of the present invention provides a posture correction desk lamp based on facial features, including: a camera module for acquiring first image data, a feature extraction module, a posture construction module, a judgment module, a reminder module, and an adjustment module;

[0067] The sitting posture construction module is used to construct a standard sitting posture model;

[0068] The feature extraction module is used to extract the user's facial feature data from the first image data;

[0069] The judgment module is used to determine whether the user's sitting posture is standard based on the facial feature data and the standard sitting posture model.

[0070] The reminder module is used to issue a reminder signal when the user's sitting posture is not standard;

[0071] The adjustment module is used to adjust the operating parameters of the desk lamp when the user's sitting posture is not standard and the user does not adjust their sitting posture within a preset time after the reminder signal is issued.

[0072] Understandably, to correct incorrect user posture, a standard sitting posture needs to be determined. However, the standard data for the standard sitting posture differs for different people and in different application scenarios. Therefore, in this embodiment, a standard sitting posture model is first constructed for different people and / or different application scenarios. The standard sitting posture model includes standard facial feature data, standard voice data, head angle data, and the positional correspondence between virtual target objects under the standard sitting posture.

[0073] The camera module can be a camera, such as a structured light-based 3D camera. The camera module can acquire first image data, which includes at least user facial image data and data of the target object that the user is interested in (such as a book, computer, mobile phone, television, tea set, etc.). The first image data carries the position information data of the object in the image.

[0074] The reminder module can be an integration of one or more of the following: a voice playback module, an image display module, and a light-emitting module.

[0075] The feature extraction module, the judgment module, the adjustment module, and other modules can be integrated into the processor of the desk lamp, or they can be used as independent modules. The embodiments of the present invention do not limit this.

[0076] The technical solution of this embodiment includes a desk lamp equipped with a camera module for acquiring first image data, a feature extraction module, a posture construction module, a judgment module, a reminder module, and an adjustment module. The posture construction module constructs a standard sitting posture model. The feature extraction module extracts facial feature data of the user from the first image data. The judgment module determines whether the user's sitting posture is standard based on the facial feature data and the standard sitting posture model. The reminder module issues a reminder signal when the user's posture is not standard. The adjustment module adjusts the desk lamp's operating parameters when the user's posture is not standard and the user does not adjust their posture within a preset time after issuing the reminder signal. This invention determines whether a sitting posture is standard by locating facial feature points, which saves significant computing resources compared to face recognition or skeletal morphology recognition. Therefore, it does not require high-performance hardware, is instantaneous, and saves power.

[0077] It should be known that, Figure 1 The block diagram of the facial feature-based posture correction desk lamp shown is for illustrative purposes only, and the number of modules shown does not limit the scope of protection of this invention.

[0078] In some possible embodiments of the present invention, the sitting posture construction module is used to construct a standard sitting posture model, specifically:

[0079] Obtain standard sitting posture requirements data for multiple application scenarios from cloud servers;

[0080] Based on the standard sitting posture requirements data in multiple application scenarios, a sitting posture input instruction for each scenario is generated and broadcast to the user, reminding the user to adjust to the standard sitting posture according to the requirements.

[0081] Record the standard sitting posture corresponding to each application scenario and establish a one-to-one correspondence between them to construct the user's standard sitting posture model, which also includes a standard facial feature vector.

[0082] It is understood that the desk lamp also includes a communication module (not shown in the figure) for receiving and sending data. Through the communication module, standard sitting posture requirements data for multiple application scenarios (such as using a computer, reading, doing homework, doing crafts, watching videos, etc.) can be obtained from the cloud server. For example, the standard sitting posture requirements for using a computer are "head facing the computer screen, 50 cm away from the screen, eyes looking straight ahead, shoulders level, etc."

[0083] Because each user's physiological characteristics differ, in order to construct an accurate standard sitting posture model that suits each user, after obtaining standard sitting posture requirement data for multiple application scenarios, a sitting posture input instruction for each scenario is generated. The sitting posture input instruction is then broadcast to the user in the form of multimedia such as audio, video, or text, reminding the user to adjust to the standard sitting posture according to the requirements. After the user adjusts to the standard sitting posture, the standard sitting posture data corresponding to each application scenario is recorded. After establishing a one-to-one correspondence between each standard sitting posture and each application scenario, the user's standard sitting posture model is constructed based on the standard sitting posture data, application scenarios, and the aforementioned correspondence between the two.

[0084] The standard sitting posture model also includes a standard facial feature vector generated from standard facial feature data to facilitate subsequent comparison.

[0085] To further ensure the accuracy of the constructed standard sitting posture model, during model construction, data on objects of interest to the user (including type, shape, location information, and positional relationship with the user) are acquired through the camera module of the desk lamp. This includes data on the type, shape, location, and positional relationship of objects such as books, computers, mobile phones, televisions, and tea sets. Using this data, along with the user's image data, a virtual 3D image of the object of interest and the user, as well as the positional relationship between these virtual 3D images, can be constructed. Combined with application scenario data, a standard sitting posture model for each application scenario can then be built.

[0086] In some possible embodiments of the present invention, the feature extraction module is used to extract the user's facial feature data from the first image data, including:

[0087] Extract facial feature point data from users;

[0088] The facial feature point data is processed to obtain facial feature data.

[0089] In this embodiment of the invention, the feature extraction module first performs image grayscale processing and data normalization on the first image data, and calculates the gradients in the horizontal and vertical directions of the image. Based on the gradients in the horizontal and vertical directions, it calculates the gradient magnitude and gradient direction. Using an undirected gradient and 9 histogram channels, the range of directions is divided into 180 / 9 = 20 degrees, i.e., the directions are divided into 9 bins. Then, the following steps are performed to extract the user's facial feature point data:

[0090] S1. Select a detection window from the image based on the window sliding step size;

[0091] S2. Select a block from the detection window based on the block sliding step size;

[0092] S3. Divide the block into sub-block units and perform box-direction projection within the sub-blocks;

[0093] S4. Calculate the projection sum and combine all sub-blocks into a large, spatially connected region to obtain the first feature.

[0094] S5. Determine if the last block has been reached. If not, return to S2.

[0095] S6. If the last block has been reached, determine whether the last window has been reached. If not, return to S1. If yes, output the facial feature point data.

[0096] The embodiments of the present invention determine whether the sitting posture is standard by locating facial feature points. Compared with face recognition or skeletal shape recognition, it can save a lot of computing resources. Therefore, it does not require hardware support with high computing power, and it is real-time and power-saving.

[0097] In some possible embodiments of the present invention, the judgment module is used to determine whether the user's sitting posture is correct based on the facial feature data and the standard sitting posture model, specifically:

[0098] Determine the current application scenario in which the user is located;

[0099] Determine a first standard sitting posture model corresponding to the current application scenario from the standard sitting posture model;

[0100] The current facial feature vector is obtained based on the facial feature data and compared with the first standard facial feature vector in the first standard sitting posture model;

[0101] When the difference between the current facial feature vector and the first standard facial feature vector is within a first preset range, the user's sitting posture standard is determined.

[0102] When the difference between the current facial feature vector and the first standard facial feature vector is not within a first preset range, it is determined that the user's sitting posture is not standard.

[0103] It should be noted that, in order to more flexibly correct sitting posture and avoid frequent reminders caused by subtle changes in the user's posture, a fault tolerance mechanism is set in this embodiment. That is, a first preset range is set (such as the user's side turning angle is -45 degrees to 45 degrees or the angle of head tilting up or down is 0 degrees to 10 degrees, etc.). When the difference between the current facial feature vector and the first standard facial feature vector is within the first preset range, the user's sitting posture standard is determined.

[0104] In some possible embodiments of the present invention, a working model construction module (not shown in the figures) is also included, for:

[0105] Obtain working data of the smart desk lamp corresponding to various sitting postures from the cloud server;

[0106] Establish working models corresponding to various sitting postures;

[0107] The adjustment module is used to adjust the operating parameters of the desk lamp when the user's posture is incorrect and the user does not adjust their posture within a preset time after the reminder signal is issued. Specifically:

[0108] Obtain the user's current sitting posture data;

[0109] The corresponding first working model is determined from the working model based on the current sitting posture data;

[0110] Adjust the operating parameters of the smart desk lamp according to the first working model.

[0111] In an embodiment of the present invention, by obtaining smart lamp working data corresponding to various sitting postures from the cloud server, a working model corresponding to each sitting posture is established. After the reminder module issues a reminder, the latest image data of the user is obtained through the camera module to obtain the user's current sitting posture data. It is analyzed whether the user has adjusted to a standard sitting posture. If the user does not adjust their sitting posture within a preset time (such as 5 minutes or 10 minutes) after the reminder signal is issued, a corresponding first working model is determined from the working model based on the current sitting posture data. The working parameters of the lamp are adjusted according to the working data contained in the first working model so that the working mode of the lamp conforms to the user's current sitting posture.

[0112] In some possible embodiments of the present invention, the posture correction system further includes a three-dimensional modeling module (not shown in the figure) for collecting the user's vital sign data, establishing facial feature models and body skeletal feature models of the user in different postures, and establishing the correlation between the two.

[0113] See Figure 2 Another embodiment of the present invention provides a sitting posture correction method based on facial features. The sitting posture correction system includes a camera module for acquiring first image data, a feature extraction module, a sitting posture construction module, a judgment module, a reminder module, and an adjustment module. The sitting posture correction method includes:

[0114] The sitting posture construction module constructs a standard sitting posture model;

[0115] The feature extraction module extracts the user's facial feature data from the first image data;

[0116] The judgment module determines whether the user's sitting posture is standard based on the facial feature data and the standard sitting posture model.

[0117] The reminder module sends a reminder signal when the user's sitting posture is not standard;

[0118] The adjustment module adjusts the operating parameters of the desk lamp when the user's sitting posture is not standard and the user does not adjust their posture within a preset time after the reminder signal is issued.

[0119] Understandably, to correct incorrect user posture, a standard sitting posture needs to be determined. However, the standard data for the standard sitting posture differs for different people and in different application scenarios. Therefore, in this embodiment, a standard sitting posture model is first constructed for different people and / or different application scenarios. The standard sitting posture model includes standard facial feature data, standard voice data, head angle data, and the positional correspondence between virtual target objects under the standard sitting posture.

[0120] The camera module can be a camera, such as a structured light-based 3D camera. The camera module can acquire first image data, which includes at least user facial image data and data of the target object that the user is interested in (such as a book, computer, mobile phone, television, tea set, etc.). The first image data carries the position information data of the object in the image.

[0121] The reminder module can be an integration of one or more of the following: a voice playback module, an image display module, and a light-emitting module.

[0122] The feature extraction module, the judgment module, the adjustment module, and other modules can be integrated into the processor of the desk lamp, or they can be used as independent modules. The embodiments of the present invention do not limit this.

[0123] The technical solution of this embodiment includes a desk lamp equipped with a camera module for acquiring first image data, a feature extraction module, a posture construction module, a judgment module, a reminder module, and an adjustment module. The posture construction module constructs a standard sitting posture model. The feature extraction module extracts facial feature data of the user from the first image data. The judgment module determines whether the user's sitting posture is standard based on the facial feature data and the standard sitting posture model. The reminder module issues a reminder signal when the user's posture is not standard. The adjustment module adjusts the desk lamp's operating parameters when the user's posture is not standard and the user does not adjust their posture within a preset time after issuing the reminder signal. This invention determines whether a sitting posture is standard by locating facial feature points, which saves significant computing resources compared to face recognition or skeletal morphology recognition. Therefore, it does not require high-performance hardware, is instantaneous, and saves power.

[0124] In some possible embodiments of the present invention, the step of the sitting posture construction module constructing a standard sitting posture model includes:

[0125] Obtain standard sitting posture requirements data for multiple application scenarios from cloud servers;

[0126] Based on the standard sitting posture requirements data in multiple application scenarios, a sitting posture input instruction for each scenario is generated and broadcast to the user, reminding the user to adjust to the standard sitting posture according to the requirements.

[0127] Record the standard sitting posture corresponding to each application scenario and establish a one-to-one correspondence between them to construct the user's standard sitting posture model, which includes standard facial feature vectors.

[0128] It is understood that the desk lamp also includes a communication module for receiving and sending data. Through the communication module, standard sitting posture requirements data for multiple application scenarios (such as using a computer, reading, doing homework, doing crafts, watching videos, etc.) can be obtained from the cloud server. For example, the standard sitting posture requirements for using a computer are "head facing the computer screen, 50 cm away from the screen, eyes looking straight ahead, shoulders level, etc."

[0129] Because each user's physiological characteristics differ, in order to construct an accurate standard sitting posture model that suits each user, after obtaining standard sitting posture requirement data for multiple application scenarios, a sitting posture input instruction for each scenario is generated. The sitting posture input instruction is then broadcast to the user in the form of multimedia such as audio, video, or text, reminding the user to adjust to the standard sitting posture according to the requirements. After the user adjusts to the standard sitting posture, the standard sitting posture data corresponding to each application scenario is recorded. After establishing a one-to-one correspondence between each standard sitting posture and each application scenario, the user's standard sitting posture model is constructed based on the standard sitting posture data, application scenarios, and the aforementioned correspondence between the two.

[0130] The standard sitting posture model also includes a standard facial feature vector generated from standard facial feature data to facilitate subsequent comparison.

[0131] To further ensure the accuracy of the constructed standard sitting posture model, during model construction, data on objects of interest to the user (including type, shape, location information, and positional relationship with the user) are acquired through the camera module of the desk lamp. This includes data on the type, shape, location, and positional relationship of objects such as books, computers, mobile phones, televisions, and tea sets. Using this data, along with the user's image data, a virtual 3D image of the object of interest and the user, as well as the positional relationship between these virtual 3D images, can be constructed. Combined with application scenario data, a standard sitting posture model for each application scenario can then be built.

[0132] In some possible embodiments of the present invention, the step of the feature extraction module extracting the user's facial feature data from the first image data includes:

[0133] Extract facial feature point data from users;

[0134] The facial feature point data is processed to obtain facial feature data.

[0135] In this embodiment of the invention, the feature extraction module first performs image grayscale processing and data normalization on the first image data, and calculates the gradients in the horizontal and vertical directions of the image. Based on the gradients in the horizontal and vertical directions, it calculates the gradient magnitude and gradient direction. Using an undirected gradient and 9 histogram channels, the range of directions is divided into 180 / 9 = 20 degrees, i.e., the directions are divided into 9 bins. Then, the following steps are performed to extract the user's facial feature point data:

[0136] S1. Select a detection window from the image based on the window sliding step size;

[0137] S2. Select a block from the detection window based on the block sliding step size;

[0138] S3. Divide the block into sub-block units and perform box-direction projection within the sub-blocks;

[0139] S4. Calculate the projection sum and combine all sub-blocks into a large, spatially connected region to obtain the first feature.

[0140] S5. Determine if the last block has been reached. If not, return to S2.

[0141] S6. If the last block has been reached, determine whether the last window has been reached. If not, return to S1. If yes, output the facial feature point data.

[0142] The embodiments of the present invention determine whether the sitting posture is standard by locating facial feature points. Compared with face recognition or skeletal shape recognition, it can save a lot of computing resources. Therefore, it does not require hardware support with high computing power, and it is real-time and power-saving.

[0143] In some possible embodiments of the present invention, the step of the judgment module determining whether the user's sitting posture is correct based on the facial feature data and the standard sitting posture model includes:

[0144] Determine the current application scenario in which the user is located;

[0145] Determine a first standard sitting posture model corresponding to the current application scenario from the standard sitting posture model;

[0146] The current facial feature vector is obtained based on the facial feature data and compared with the first standard facial feature vector in the first standard sitting posture model;

[0147] When the difference between the current facial feature vector and the first standard facial feature vector is within a first preset range, the user's sitting posture standard is determined.

[0148] When the difference between the current facial feature vector and the first standard facial feature vector is not within a first preset range, it is determined that the user's sitting posture is not standard.

[0149] It should be noted that, in order to more flexibly correct sitting posture and avoid frequent reminders caused by subtle changes in the user's posture, a fault tolerance mechanism is set in this embodiment. That is, a first preset range is set (such as the user's side turning angle is -45 degrees to 45 degrees or the angle of head tilting up or down is 0 degrees to 10 degrees, etc.). When the difference between the current facial feature vector and the first standard facial feature vector is within the first preset range, the user's sitting posture standard is determined.

[0150] In some possible embodiments of the present invention, a working model construction module is also included, for:

[0151] Obtain working data of the smart desk lamp corresponding to various sitting postures from the cloud server;

[0152] Establish working models corresponding to various sitting postures;

[0153] The adjustment module is used to adjust the operating parameters of the desk lamp when the user's posture is incorrect and the user does not adjust their posture within a preset time after the reminder signal is issued. Specifically:

[0154] Obtain the user's current sitting posture data;

[0155] The corresponding first working model is determined from the working model based on the current sitting posture data;

[0156] Adjust the operating parameters of the smart desk lamp according to the first working model.

[0157] In an embodiment of the present invention, by obtaining smart lamp working data corresponding to various sitting postures from the cloud server, a working model corresponding to each sitting posture is established. After the reminder module issues a reminder, the latest image data of the user is obtained through the camera module to obtain the user's current sitting posture data. It is analyzed whether the user has adjusted to a standard sitting posture. If the user does not adjust their sitting posture within a preset time (such as 5 minutes or 10 minutes) after the reminder signal is issued, a corresponding first working model is determined from the working model based on the current sitting posture data. The working parameters of the lamp are adjusted according to the working data contained in the first working model so that the working mode of the lamp conforms to the user's current sitting posture.

[0158] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0159] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0160] In the several embodiments provided in this application, it should be understood that the disclosed apparatus can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of the units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical or other forms.

[0161] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0162] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0163] If the integrated units described above are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage device (CMD). Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned memory includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0164] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0165] The embodiments of this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

[0166] While the present invention has been disclosed above, it is not limited thereto. Any person skilled in the art can easily conceive of variations or substitutions without departing from the spirit and scope of the present invention, and various modifications and alterations can be made, including combinations of the different functions and implementation steps described above, as well as software and hardware implementation methods, all of which are within the protection scope of the present invention.

Claims

1. A face feature based sitting posture correction desk lamp, characterized in that, include: The system includes a camera module, a feature extraction module, a posture construction module, a judgment module, a reminder module, and an adjustment module for acquiring first image data. The posture construction module is used to construct a standard posture model. Specifically, it retrieves standard posture requirement data for multiple application scenarios from a cloud server; generates posture input instructions for each scenario based on the standard posture requirement data for multiple application scenarios and broadcasts them to the user, reminding the user to adjust to the standard posture according to the posture requirements; records the standard posture corresponding to each application scenario and establishes a one-to-one correspondence between the two, constructing the user's standard posture model. The standard posture model includes standard facial feature data, standard voice data, head angle data, and the positional correspondence between virtual target objects under the standard posture. The feature extraction module is used to extract the user's facial feature data from the first image data; The judgment module is used to determine whether the user's sitting posture is standard based on the facial feature data and the standard sitting posture model. The reminder module is used to issue a reminder signal when the user's sitting posture is not standard; The adjustment module is used to adjust the working parameters of the desk lamp when the user's sitting posture is not standard and the user does not adjust the sitting posture within a preset time after the reminder signal is issued; The posture construction module acquires data on the target object of user interest through the camera module of the desk lamp during model construction. The target object data includes the type, shape, and location information of the target object, as well as the positional relationship data between the target object and the user. Based on the target object data and the user's image data, a virtual 3D image of the target object of user interest and the user's virtual 3D image, as well as the positional relationship data between the two virtual 3D images, are constructed. Combined with application scenario data, a standard posture model for each application scenario is constructed. The feature extraction module is used to extract the user's facial feature data from the first image data, including: The first image data is processed for grayscale, the data is normalized, and the gradients in the horizontal and vertical directions of the image are calculated. Based on the gradients in the horizontal and vertical directions, the magnitude and direction of the gradients are calculated. Using an undirected gradient and 9 histogram channels, the directional range is divided into 9 bins at intervals of 180 / 9 = 20 degrees. The process of extracting facial feature point data from the user is as follows: S1. Select a detection window from the image based on the window sliding step size; S2. Select a block from the detection window based on the block sliding step size; S3. Divide the block into sub-block units and perform box-direction projection within the sub-blocks; S4. Calculate the projection sum and combine all sub-blocks into a large, spatially connected region to obtain the first feature; S5. Determine if the last block has been reached. If not, return to S2; S6. If the last block has been reached, determine if the last window has been reached. If not, return to S1. If so, output the facial feature point data. The facial feature point data is processed to obtain facial feature data.

2. The posture correcting desk lamp according to claim 1, wherein The judgment module is configured to judge whether the sitting posture of the user is correct according to the facial feature data and the standard sitting posture model, and specifically to: determine a current application scenario in which the user is located; determine a first standard sitting posture model corresponding to the current application scenario from the standard sitting posture model; obtain a current facial feature vector from the facial feature data, and compare the current facial feature vector with a first standard facial feature vector in the first standard sitting posture model; when a difference value of the current facial feature vector and the first standard facial feature vector is within a first preset range, judge that the sitting posture of the user is standard; when the difference value of the current facial feature vector and the first standard facial feature vector is not within the first preset range, judge that the sitting posture of the user is not standard.

3. The sitting posture correcting desk lamp according to claim 2, further comprising a working model construction module configured to: obtain intelligent desk lamp working data corresponding to various sitting postures from the cloud server; establish working models corresponding to the various sitting postures; The adjustment module is configured to adjust the working parameters of the intelligent desk lamp when the sitting posture of the user is incorrect and the user does not adjust the sitting posture within a preset time after the reminder signal is sent, and specifically to: obtain current sitting posture data of the user; determine a corresponding first working model from the working models according to the current sitting posture data; adjust the working parameters of the intelligent desk lamp according to the first working model.

4. A sitting posture correction method based on facial features, characterized in that, The sitting posture correcting system comprises a camera module for collecting first image data, a feature extraction module, a sitting posture construction module, a judgment module, a reminder module, and an adjustment module. The sitting posture correcting method comprises: The sitting posture construction module constructs a standard sitting posture model, including: obtaining standard sitting posture requirement data under multiple application scenarios from a cloud server; generating a sitting posture input instruction under each scenario according to the standard sitting posture requirement data under the multiple application scenarios, and broadcasting the instruction to the user to remind the user to adjust to a standard sitting posture according to the sitting posture requirement; recording a standard sitting posture corresponding to each application scenario and establishing a one-to-one correspondence relationship therebetween to construct a standard sitting posture model of the user, the standard sitting posture model comprising standard facial feature data, standard sound data, and head angle data under a standard sitting posture, as well as a position correspondence relationship between the virtual target objects; The feature extraction module extracts facial feature data of the user from the first image data; The judgment module judges whether the sitting posture of the user is standard according to the facial feature data and the standard sitting posture model; The reminder module sends a reminder signal when the sitting posture of the user is not standard; The adjustment module adjusts the working parameters of the intelligent desk lamp when the sitting posture of the user is not standard and the user does not adjust the sitting posture within a preset time after the reminder signal is sent. The sitting posture construction module acquires data of a target object focused by the user through a camera module of the desk lamp when constructing a model; the data of the target object includes a type, a shape, position information of the target object, and position relationship data between the target object and the user; a virtual three-dimensional image of the target object focused by the user and a virtual three-dimensional image of the user are constructed through the data of the target object and image data of the user, and position relationship data between the two virtual three-dimensional images is constructed; a standard sitting posture model in each application scenario is constructed in combination with application scenario data; The feature extraction module extracts the facial feature data of the user from the first image data, and the step includes: The first image data is subjected to image grayscale processing, normalized data, and calculation of gradients in horizontal and vertical coordinate directions, and the gradient size and gradient direction are calculated according to the gradients in the horizontal and vertical coordinate directions; The direction range is divided into nine bins at an interval of 180 / 9=20 degrees by using the undirected gradient and nine histogram channels; The facial feature point data of the user is extracted, specifically: S1, a detection window is selected from the image according to a window sliding step; S2, a block is selected from the detection window according to a block sliding step; S3, a sub-block unit is divided in the block, and bin direction projection is performed in the sub-block; S4, projection and are calculated, and all sub-blocks are combined into a large, spatially connected region to obtain first features; S5, whether the last block is reached is judged, if not, return to S2; S6, if the last block has been reached, whether the last window is reached is judged, if not, return to S1, and if yes, facial feature point data is output. The facial feature point data is subjected to feature processing to obtain facial feature data.

5. The sitting posture correcting method according to claim 4, characterized by, The judgment module judges whether the sitting posture of the user is correct according to the facial feature data and the standard sitting posture model, and the step includes: Determining a current application scenario in which the user is located; Determining a first standard sitting posture model corresponding to the current application scenario from the standard sitting posture model; Obtaining a current facial feature vector from the facial feature data, and comparing the current facial feature vector with a first standard facial feature vector in the first standard sitting posture model; When a difference value between the current facial feature vector and the first standard facial feature vector is within a first preset range, it is judged that the sitting posture of the user is standard; When the difference value between the current facial feature vector and the first standard facial feature vector is not within the first preset range, it is judged that the sitting posture of the user is not standard.

6. The sitting posture correcting method according to claim 5, characterized by, Further comprising a working model construction module for: Obtaining intelligent desk lamp working data corresponding to various sitting postures from the cloud server; Establishing working models corresponding to the various sitting postures; The adjustment module is configured to adjust the working parameters of the desk lamp when the sitting posture of the user is incorrect and the user does not adjust the sitting posture within a preset time after the reminder signal is sent, and specifically: Obtaining current sitting posture data of the user; Determining a corresponding first working model from the working models according to the current sitting posture data; Adjusting the working parameters of the intelligent desk lamp according to the first working model.

Images