CN-116912886-B - Sitting posture monitoring method and device, intelligent desk lamp, electronic equipment and storage medium

CN116912886BCN 116912886 BCN116912886 BCN 116912886BCN-116912886-B

Abstract

The application discloses a sitting posture monitoring method, a device, an intelligent table lamp, electronic equipment and a storage medium, wherein the method comprises the steps of obtaining an image to be identified containing a user, detecting human body key points of the image to be identified to obtain human body key points in the image to be identified, determining a human body deflection angle of the user in the image to be identified based on the human body key points in the image to be identified and the human body key points in a standard sitting posture image corresponding to the user, performing deflection correction on the human body key points in the image to be identified based on the human body deflection angle to obtain the human body key points after deflection correction, and determining a first sitting posture identification result of the user based on the human body key points after deflection correction. The method and the device disclosed by the application improve the accuracy of sitting posture identification of the placement position of the sitting posture monitoring device under the condition that different deflection angles exist relative to the front direction of the user.

Inventors

PU GUIYANG
ZHANG WENJING
WU XINGWEN
LI YING
WANG LIUQING

Assignees

中移（杭州）信息技术有限公司
中国移动通信集团有限公司

Dates

Publication Date: 20260505
Application Date: 20230727

Claims (11)

1. A sitting posture monitoring method, comprising: Acquiring an image to be identified containing a user; Detecting human body key points of the image to be identified to obtain human body key points in the image to be identified; Determining a human body deflection angle of the user in the image to be identified based on the human body key points in the image to be identified and the human body key points in the standard sitting posture image corresponding to the user, wherein the human body deflection angle is a horizontal included angle of the sitting posture of the user in the image to be identified relative to the sitting posture of the user in the standard sitting posture image in the aspect of human body front orientation; performing deflection correction on human body key points in the image to be identified based on the human body deflection angle to obtain the human body key points after deflection correction; and determining a first sitting posture recognition result of the user based on the deflection corrected human body key points.
2. The sitting posture monitoring method according to claim 1, wherein the determining the human body deflection angle of the user in the image to be recognized based on the human body key points in the image to be recognized and the human body key points in the standard sitting posture image corresponding to the user comprises: Determining a first connecting line between shoulder key points in human body key points in the image to be identified; determining a second line between shoulder keypoints of human body keypoints in the standard sitting position image; And determining the human body deflection angle of the user in the image to be identified based on the included angle between the first connecting line and the second connecting line.
3. A sitting posture monitoring method according to claim 2, wherein the performing deflection correction on the human body key points in the image to be identified based on the human body deflection angle to obtain the human body key points after deflection correction includes: determining a shoulder width of the user and a pixel position of a human body key point in the middle of the shoulder of the user based on a second line between shoulder key points in the human body key points in the standard sitting posture image; Determining a pixel displacement deflection value of a human body key point on the left side of the spine of the user based on the human body deflection angle, a distance between the human body key point on the left shoulder of the user and the human body key point in the middle of the shoulder, and the shoulder width; Determining a pixel displacement deflection value of a human body key point on the right side of the spine of the user based on the human body deflection angle, a distance between the human body key point on the right shoulder of the user and the human body key point in the middle of the shoulder, and the shoulder width; And performing pixel position adjustment on the human body key points on the left side of the spine of the user based on the pixel displacement deflection values of the human body key points on the left side of the spine of the user, and performing pixel position adjustment on the human body key points on the right side of the spine of the user based on the pixel displacement deflection values of the human body key points on the right side of the spine of the user, so as to obtain the human body key points after deflection correction.
4. A method of monitoring a sitting posture according to claim 1, wherein the determining a first sitting posture recognition result of the user based on the deflection corrected human body keypoints comprises: determining corresponding key points of the human body key points after deflection correction in the human body key points in the standard sitting posture image; determining the distance between the human body key point after deflection correction and the corresponding key point; Determining the similarity of the user between the sitting posture of the image to be identified and the standard sitting posture based on the distance between the human body key points after deflection correction and the corresponding key points; and determining a first sitting posture recognition result of the user based on the similarity between the sitting posture of the image to be recognized and the standard sitting posture of the user.
5. A method of monitoring a sitting posture according to claim 1, wherein after determining the first sitting posture recognition result of the user based on the deflection corrected human body keypoints, the method further comprises: detecting the key points of the human face of the image to be identified to obtain the key points of the human face in the image to be identified; Determining the current face attitude angle of the user based on the face key points; And determining a second sitting posture recognition result of the user based on the current face posture angle of the user and the face posture angle in the standard sitting posture image corresponding to the user.
6. A method of monitoring a sitting posture according to claim 1, wherein after determining the first sitting posture recognition result of the user based on the deflection corrected human body keypoints, the method further comprises: generating sitting posture prompt information under the condition that the first sitting posture identification result is an incorrect sitting posture; Counting the sitting posture holding time of the user under the condition that the first sitting posture identification result is a correct sitting posture; And generating sedentary prompt information under the condition that the sitting posture holding time is greater than or equal to a preset time threshold value.
7. A sitting posture monitoring method according to any of claims 1 to 6, wherein the standard sitting posture image is determined based on the steps of: Placing a sitting posture monitoring device in front of the user to acquire a sitting posture image of the user; Carrying out face attitude angle detection and human body sitting posture recognition on the sitting posture image of the user to obtain a face attitude angle detection result and a human body sitting posture recognition result; Based on the face attitude angle detection result and the human body sitting posture recognition result, determining that the sitting posture of the user meets a standard sitting posture; Determining the sitting posture image as the standard sitting posture image.
8. A sitting posture monitoring device, comprising: The acquisition module is used for acquiring an image to be identified containing a user; the detection module is used for detecting human body key points of the image to be identified to obtain the human body key points in the image to be identified; The system comprises a determination module, a human body deflection angle determination module and a display module, wherein the determination module is used for determining a human body deflection angle of a user in an image to be recognized based on human body key points in the image to be recognized and human body key points in a standard sitting posture image corresponding to the user, wherein the human body deflection angle is a horizontal included angle of the sitting posture of the user in the image to be recognized relative to the sitting posture of the user in the standard sitting posture image in the aspect of the front orientation of the human body; The correction module is used for carrying out deflection correction on the human body key points in the image to be identified based on the human body deflection angle to obtain the human body key points after deflection correction; And the identification module is used for determining the sitting posture identification result of the user based on the human body key points after deflection correction.
9. An intelligent table lamp is characterized by comprising a table lamp body, a camera acquisition module, a voice reminding module, an operation analysis module, a connection control module and the sitting posture monitoring device of claim 8, wherein the camera acquisition module, the voice reminding module, the operation analysis module and the connection control module are arranged on the table lamp body; the camera acquisition module is used for acquiring images to be identified of a user; the voice reminding module is used for playing sitting posture prompt information and sedentary reminding information generated by the sitting posture monitoring device; The operation analysis module is used for generating desk lamp use information of the user based on the sitting posture identification result, the sitting posture prompt information and the sedentary prompt information of the user; the connection control module is in communication connection with the remote control terminal and is used for configuring the sitting posture monitoring device based on a control instruction sent by the remote control terminal.
10. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements a sitting posture monitoring method according to any one of claims 1 to 7 when the program is executed by the processor.
11. A non-transitory computer readable storage medium having stored thereon a computer program, wherein the computer program when executed by a processor implements a sitting posture monitoring method according to any one of claims 1 to 7.

Description

Sitting posture monitoring method and device, intelligent desk lamp, electronic equipment and storage medium Technical Field The application relates to the technical field of computers, in particular to a sitting posture monitoring method and device, an intelligent table lamp, electronic equipment and a storage medium. Background Along with the progress of technology, the intelligent desk lamp that possesses intelligent warning function is produced, can carry out position of sitting discernment and healthy warning to the user, has played health protection's effect. In the prior art, an intelligent table lamp uses a camera to collect images of the sitting posture of a human body. When the intelligent desk lamp is used, the intelligent desk lamp is required to be placed at a specific position, so that the camera is positioned at an optimal visual angle to collect images of the sitting posture of a human body. When the intelligent desk lamp deviates from a specific position, the sitting posture image acquired by the camera is difficult to identify the sitting posture, so that the sitting posture identification effect is poor. The place of intelligent desk lamp is limited, and use experience is relatively poor. Disclosure of Invention The application provides a sitting posture monitoring method, a sitting posture monitoring device, an intelligent table lamp, electronic equipment and a storage medium, which are used for solving the technical problem of poor sitting posture recognition effect when image recognition is carried out on the sitting posture of a human body. The application provides a sitting posture monitoring method, which comprises the following steps: Acquiring an image to be identified containing a user; Detecting human body key points of the image to be identified to obtain human body key points in the image to be identified; Determining a human body deflection angle of the user in the image to be identified based on the human body key points in the image to be identified and the human body key points in the standard sitting posture image corresponding to the user; performing deflection correction on human body key points in the image to be identified based on the human body deflection angle to obtain the human body key points after deflection correction; and determining a first sitting posture recognition result of the user based on the deflection corrected human body key points. In some embodiments, the determining the human body deflection angle of the user in the image to be identified based on the human body key points in the image to be identified and the human body key points in the standard sitting posture image corresponding to the user includes: Determining a first connecting line between shoulder key points in human body key points in the image to be identified; determining a second line between shoulder keypoints of human body keypoints in the standard sitting position image; And determining the human body deflection angle of the user in the image to be identified based on the included angle between the first connecting line and the second connecting line. In some embodiments, the performing deflection correction on the human body key points in the image to be identified based on the human body deflection angle to obtain the human body key points after deflection correction includes: determining a shoulder width of the user and a pixel position of a human body key point in the middle of the shoulder of the user based on a second line between shoulder key points in the human body key points in the standard sitting posture image; Determining a pixel displacement deflection value of a human body key point on the left side of the spine of the user based on the human body deflection angle, a distance between the human body key point on the left shoulder of the user and the human body key point in the middle of the shoulder, and the shoulder width; Determining a pixel displacement deflection value of a human body key point on the right side of the spine of the user based on the human body deflection angle, a distance between the human body key point on the right shoulder of the user and the human body key point in the middle of the shoulder, and the shoulder width; And performing pixel position adjustment on the human body key points on the left side of the spine of the user based on the pixel displacement deflection values of the human body key points on the left side of the spine of the user, and performing pixel position adjustment on the human body key points on the right side of the spine of the user based on the pixel displacement deflection values of the human body key points on the right side of the spine of the user, so as to obtain the human body key points after deflection correction. In some embodiments, the determining the first sitting posture recognition result of the user based on the deflection corrected human body keypoints comprises: determining corresponding key points o