Fall detection method, apparatus, computer device and storage medium

Abstract

The application relates to a falling detection method and device, computer equipment, a storage medium and a computer program product, and relates to the field of artificial intelligence. The method comprises the following steps: acquiring real-time images of a detection object at a current moment and a previous moment; determining current position information and previous moment position information of each joint of the detection object according to the real-time images at the current moment and the previous moment; determining displacement information of a first type of joint of the detection object according to the current position information and the previous moment position information; the first type of joint is a characteristic joint of the detection object that can detect falling; and determining that the detection object is in a falling state at the current moment in the case that the displacement information of the first type of joint meets a preset falling condition. The method can accurately and timely detect falling behavior.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular to a fall detection method, apparatus, computer equipment, storage medium, and computer program product. Background Technology

[0002] With the development of artificial intelligence, technologies related to human fall detection have emerged. The common method used is to wear a smart bracelet.

[0003] However, the wrist is the part of the body with the largest range of motion and the most frequent movement. When using a wristband to determine whether someone has fallen, it may mistakenly identify other actions such as squatting or lying down as a fall, leading to incorrect judgments. Summary of the Invention

[0004] Therefore, it is necessary to provide a fall detection method, device, computer equipment, computer-readable storage medium, and computer program product that can accurately detect falls, addressing the aforementioned technical problems.

[0005] Firstly, this application provides a fall detection method. The method includes:

[0006] Acquire the real-time image of the detected object at the current moment and the real-time image at the previous moment;

[0007] Based on the real-time image at the current moment and the real-time image at the previous moment, determine the current position information and the previous position information of each joint of the detected object;

[0008] Based on the current position information and the previous position information, the displacement information of the first type of joints of the detection object is determined; the first type of joints are the characteristic joints on the detection object that can detect falls.

[0009] If the displacement information of the first type of joint meets the preset fall conditions, the detected object is determined to be in a fall state at the current moment.

[0010] In one embodiment, the displacement information includes displacement velocity and displacement distance; the first type of joint points includes a first joint point and a second joint point; the method further includes:

[0011] Obtain the speed change threshold and distance change threshold;

[0012] If the displacement velocity of the first joint exceeds the velocity change threshold and the displacement distance of the second joint exceeds the distance change threshold, the detected object is determined to be in a fallen state at the current moment.

[0013] In one embodiment, after determining that the detected object is in a fallen state at the current moment, the method further includes:

[0014] Based on the current position information and the previous position information, determine the displacement information of the second type of key points;

[0015] Based on the displacement information of the first type of joint and the displacement information of the second type of joint, the fall type corresponding to the detected object is determined.

[0016] In one embodiment, determining the fall type corresponding to the detected object based on the displacement information of the first type of joint points and the displacement information of the second type of joint points includes:

[0017] If the displacement information of the second type of joint exceeds the preset displacement change threshold, then the fall type corresponding to the detected object is determined based on the displacement information of the second type of joint.

[0018] If the displacement information of the second type of joint does not exceed the preset displacement change threshold, then the fall type corresponding to the detected object is determined based on the displacement information of the first type of joint.

[0019] In one embodiment, determining the fall type corresponding to the detected object based on the displacement information of the first type of joint points includes:

[0020] If the displacement information of the first type of joint meets the first condition, then the fall type corresponding to the detected object is determined to be a direct fall; the first condition is that the displacement information exceeds the first displacement threshold.

[0021] If the displacement information of the first type of joint meets the second condition, then the fall type corresponding to the detected object is determined to be a fall against the wall; the second condition is that the displacement information does not exceed the first displacement threshold.

[0022] In one embodiment, the second type of joint points includes a third joint point and a fourth joint point; the displacement change threshold includes a third displacement threshold and a fourth displacement threshold;

[0023] The step of determining the fall type corresponding to the detected object based on the displacement information of the second type of joint points includes:

[0024] If the displacement information of the third joint point meets the preset third condition, then the fall type corresponding to the detected object is determined to be a forward kneeling fall; the third condition is that the displacement information of the third joint point exceeds the third displacement threshold, and the displacement direction of the third joint point is towards the front of the detected object;

[0025] If the displacement information of the fourth joint point meets the preset fourth condition, then the fall type corresponding to the detected object is determined to be a backward sitting fall; the fourth condition is that the displacement information of the fourth joint point exceeds the fourth displacement threshold, and the displacement direction of the fourth joint point is towards the rear of the detected object.

[0026] In one embodiment, the method further includes:

[0027] If the displacement information of the first type of joint does not meet the preset fall conditions, the non-fall behavior type of the detected object at the current moment is determined based on the displacement information of the first type of joint.

[0028] Secondly, this application also provides a fall detection device. The device includes:

[0029] The image acquisition module is used to acquire the real-time image of the detected object at the current moment and the real-time image at the previous moment;

[0030] The position determination module is used to determine the current position information and the previous position information of each joint of the detected object based on the real-time image at the current moment and the real-time image at the previous moment.

[0031] The displacement calculation module is used to determine the displacement information of the first type of joints of the detection object based on the current position information and the previous position information; the first type of joints are the feature joints on the detection object that can detect falls.

[0032] The behavior determination module is used to determine that the detected object is in a falling state at the current moment if the displacement information of the first type of joint meets the preset falling conditions.

[0033] Thirdly, this application also provides a computer device. The computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to perform the following steps:

[0034] Acquire the real-time image of the detected object at the current moment and the real-time image at the previous moment;

[0035] Based on the real-time image at the current moment and the real-time image at the previous moment, determine the current position information and the previous position information of each joint of the detected object;

[0036] Based on the current position information and the previous position information, the displacement information of the first type of joints of the detection object is determined; the first type of joints are the characteristic joints on the detection object that can detect falls.

[0037] If the displacement information of the first type of joint meets the preset fall conditions, the detected object is determined to be in a fall state at the current moment.

[0038] Fourthly, this application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program thereon, which, when executed by a processor, performs the following steps:

[0039] Acquire the real-time image of the detected object at the current moment and the real-time image at the previous moment;

[0040] Based on the real-time image at the current moment and the real-time image at the previous moment, determine the current position information and the previous position information of each joint of the detected object;

[0041] Based on the current position information and the previous position information, the displacement information of the first type of joints of the detection object is determined; the first type of joints are the characteristic joints on the detection object that can detect falls.

[0042] If the displacement information of the first type of joint meets the preset fall conditions, the detected object is determined to be in a fall state at the current moment.

[0043] Fifthly, this application also provides a computer program product. The computer program product includes a computer program that, when executed by a processor, performs the following steps:

[0044] Acquire the real-time image of the detected object at the current moment and the real-time image at the previous moment;

[0045] Based on the real-time image at the current moment and the real-time image at the previous moment, determine the current position information and the previous position information of each joint of the detected object;

[0046] Based on the current position information and the previous position information, the displacement information of the first type of joints of the detection object is determined; the first type of joints are the characteristic joints on the detection object that can detect falls.

[0047] If the displacement information of the first type of joint meets the preset fall conditions, the detected object is determined to be in a fall state at the current moment.

[0048] The aforementioned fall detection methods, devices, computer equipment, storage media, and computer program products obtain displacement information of target joints by comparing real-time images of the detected object, and then determine whether the detected object has fallen based on the displacement information. They can accurately identify the fall behavior of the detected object, thereby better reminding users to take appropriate measures. Attached Figure Description

[0049] Figure 1 This is a diagram illustrating the application environment of a fall detection method in one embodiment.

[0050] Figure 2 This is a flowchart illustrating a fall detection method in one embodiment;

[0051] Figure 3 This is a flowchart illustrating the process of determining the fall type corresponding to a detection object in one embodiment;

[0052] Figure 4 This is a flowchart illustrating a fall detection method in another embodiment;

[0053] Figure 5 This is a schematic diagram of joints in one embodiment;

[0054] Figure 6 This is a flowchart illustrating a fall detection method in one embodiment;

[0055] Figure 7 This is a structural block diagram of a fall detection device in one embodiment;

[0056] Figure 8 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0057] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0058] The fall detection method provided in this application embodiment can be applied to, for example, Figure 1 In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be integrated onto server 104, or it can be located in the cloud or on another network server. Server 104 receives real-time images from terminal 102 and, based on information such as fall thresholds in the data storage system, analyzes the displacement information of the target joints of the detected object in the real-time image to determine whether the detected object has fallen, and displays this conclusion on terminal 102. Terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, and tablets. Server 104 can be implemented using a standalone server or a server cluster consisting of multiple servers.

[0059] In one embodiment, such as Figure 2As shown, a fall detection method is provided, which can be applied to... Figure 1 Taking server 104 as an example, the following steps are included:

[0060] Step 201: Obtain the real-time image of the detected object at the current moment and the real-time image at the previous moment.

[0061] In this context, the real-time image can be a video image, and the real-time image at each moment is the image of each frame of the video image.

[0062] For example, server 104 receives video images sent by terminal 102, then extracts the frame of the video image at the current moment as the real-time image at the current moment, and extracts the frame of the previous moment as the real-time image at the previous moment.

[0063] Step 202: Based on the real-time image at the current moment and the real-time image at the previous moment, determine the current position information and the previous position information of each joint of the detected object.

[0064] The location information can be the coordinate information of the joints; furthermore, the coordinate information is determined by a pre-established coordinate system.

[0065] Each joint point can include the shoulder joint, knee joint, and neck joint of the monitored object.

[0066] In this process, each key point of the object being detected is determined from real-time images using publicly available image recognition technology.

[0067] For example, server 104 uses publicly available image recognition technology to detect each joint of the detected object from real-time images at two different times, and determines the coordinate information of each joint at the two times based on a pre-established coordinate system, which are then used as the position information of each joint at the two times.

[0068] Step 203: Based on the current position information and the position information at the previous moment, determine the displacement information of the first type of joints of the detection object; the first type of joints are the feature joints on the detection object that can detect falls.

[0069] The displacement information includes displacement velocity information and displacement distance information.

[0070] For example, the server 104 determines the current position information and the previous position information of the first type of joint from the position information of each joint, calculates the displacement distance information of the first type of joint, and calculates the displacement velocity information of the first type of joint through the time interval between the two times.

[0071] Step 204: If the displacement information of the first type of joints meets the preset fall conditions, determine that the detected object is in a fall state at the current moment.

[0072] The pre-set fall conditions are determined through prior training.

[0073] For example, server 104 obtains preset fall conditions pre-trained and determined in the data storage system, and determines whether the displacement information of the first type of joint meets the preset fall conditions, that is, whether the displacement speed or displacement distance of the first type of joint exceeds the speed threshold or distance threshold in the preset fall conditions. When the displacement information of the first type of joint meets the preset fall conditions, it is determined that the detected object is in a fall state at the current moment.

[0074] In the aforementioned fall detection method, by acquiring real-time images of the object being detected and comparing their positional information, the displacement information of the target joint is obtained. Based on this displacement information, it is then determined whether the object has fallen. This method can accurately identify the object's fall behavior and thus better remind users to take appropriate measures.

[0075] In one embodiment, the displacement information includes displacement velocity and displacement distance; the first type of joint points includes a first joint point and a second joint point. Step 204, if the displacement information of the first type of joint points meets the preset fall conditions, determines that the detected object is in a fall state at the current moment. This can be further implemented through the following steps:

[0076] Step 1: Obtain the speed change threshold and distance change threshold;

[0077] Step 2: If the displacement velocity at the first joint exceeds the velocity change threshold and the displacement distance at the second joint exceeds the distance change threshold, then the detected object is determined to be in a fallen state at the current moment.

[0078] The velocity change threshold and distance change threshold can be obtained through training. Furthermore, the velocity change threshold can be the maximum downward displacement velocity of the first joint point of the detected object between two moments during the squatting process, and the distance change threshold can be the maximum displacement distance of the second joint point of the detected object between two moments during the squatting process.

[0079] The first joint point can be the nose, and the second joint point can include the cervical joint, shoulder joint, and lumbar joint, etc.

[0080] For example, the speed change threshold and distance change threshold obtained from pre-training are obtained. When the descent speed of the first joint (i.e. the nose) exceeds the speed change threshold and the displacement distance of each second joint (i.e. the neck joint, shoulder joint, and waist joint, etc.) exceeds the distance change threshold (i.e. the detected object has undergone a more intense downward displacement behavior than squatting), it is determined that the detected object is in a falling state at the current moment.

[0081] In this embodiment, by comparing the displacement information of the detected object at the current moment to see if it exceeds the displacement change threshold determined based on the squatting process, it can accurately identify whether the detected object has fallen, rather than detecting other behaviors such as squatting and lying down as falling.

[0082] In one embodiment, after step 204 determines that the detected object is in a fallen state at the current moment, the following steps are also included:

[0083] Step 1: Determine the displacement information of the second type of key points based on the current position information and the position information at the previous moment;

[0084] Step 2: Based on the displacement information of the first type of joints and the displacement information of the second type of joints, determine the fall type corresponding to the detected object.

[0085] The types of falls can include falling directly, falling forward to kneel, falling backward to sit, and falling against a wall.

[0086] Furthermore, falling directly means falling directly to the ground in any direction, such as forward, backward, left, or right; falling forward while kneeling means falling forward, with the knees landing first, and then continuing to fall; falling backward while sitting means falling backward, with the buttocks landing first, and then continuing to fall; and falling against a wall means falling against a wall in any direction, such as forward, backward, left, or right.

[0087] The second type of joint can include the knee joint and hip joint, among others.

[0088] For example, to better detect falls and distinguish different fall types, it is necessary to introduce displacement information from more joints (i.e., second-type joints). For a detection object determined to be in a fall state, the displacement information of the first-type and second-type joints is combined with the displacement change conditions corresponding to different fall types to determine the fall type of the detection object. It should be noted that the displacement change conditions corresponding to different fall types are obtained in advance through training.

[0089] In one embodiment, such as Figure 3 As shown, the above steps, based on the displacement information of the first type of joints and the displacement information of the second type of joints, determine the fall type corresponding to the detected object. This can also be achieved through the following steps:

[0090] Step 301: If the displacement information of the second type of joint exceeds the preset displacement change threshold, then the fall type corresponding to the detection object is determined based on the displacement information of the second type of joint.

[0091] Step 302: If the displacement information of the second type of joint does not exceed the preset displacement change threshold, then the fall type corresponding to the detection object is determined based on the displacement information of the first type of joint.

[0092] For example, it is determined whether there is a significant displacement change in the second type of joint (i.e., the knee joint or hip joint). That is, it is detected whether the displacement information of the second type of joint exceeds a preset displacement change threshold. If it exceeds the preset displacement change threshold, it can be determined that the fall type of the detected object is a kneeling fall or a sitting fall. Then, the fall type is further determined based on the displacement information of the second type of joint. If it does not exceed the preset displacement change threshold, it can be determined that the fall type of the detected object is a direct fall or a fall against a wall. Then, the fall type is further determined based on the displacement information of the first type of joint.

[0093] In the above embodiments, by analyzing the displacement information of the two types of joints, the specific fall type of the detected object can be detected, resulting in more accurate fall detection results and providing better reference for users.

[0094] In one embodiment, step 302 above, which determines the fall type corresponding to the detected object based on the displacement information of the first type of joint points, can also be achieved through the following steps:

[0095] Step 1: If the displacement information of the first type of joint meets the first condition, then the fall type corresponding to the detected object is determined to be a direct fall; the first condition is that the displacement information exceeds the first displacement threshold.

[0096] Step 2: If the displacement information of the first type of joint meets the second condition, then the fall type corresponding to the detected object is determined to be a fall against the wall; the second condition is that the displacement information does not exceed the first displacement threshold.

[0097] For example, since the speed and distance changes of a fall against a wall fall fall between those of a direct fall and a squatting fall, a preset first displacement threshold is used to distinguish between these two types of falls. When the displacement information of the first type of joint of the detected object exceeds the first displacement threshold, i.e., the displacement change is large, the fall type is determined to be a direct fall; when the displacement information of the first type of joint of the detected object does not exceed the first displacement threshold, i.e., the displacement change is larger than that of a squatting fall but not as large as that of a direct fall, the fall type is determined to be a fall against a wall.

[0098] In this embodiment, by determining whether the displacement information of the first type of joint reaches the threshold for a direct fall, a distinction is made between a direct fall and a fall against a wall, resulting in a more accurate fall detection result and providing better reference for the user.

[0099] In one embodiment, the second type of joint points includes a third joint point and a fourth joint point; the displacement change threshold includes a third displacement threshold and a fourth displacement threshold. Step 301 above determines the fall type corresponding to the detected object based on the displacement information of the second type of joint points, which can also be achieved through the following steps:

[0100] Step 1: If the displacement information of the third joint point meets the preset third condition, then the fall type corresponding to the detected object is determined to be a forward kneeling fall; the third condition is that the displacement information of the third joint point exceeds the third displacement threshold, and the displacement direction of the third joint point is towards the front of the detected object.

[0101] Step 2: If the displacement information of the fourth joint point meets the preset fourth condition, then the fall type corresponding to the detected object is determined to be a backward fall. The fourth condition is that the displacement information of the fourth joint point exceeds the fourth displacement threshold, and the displacement direction of the fourth joint point is towards the rear of the detected object.

[0102] The third joint point can be the knee joint.

[0103] The fourth joint can be referred to as the hip joint.

[0104] For example, falling forward on a kneeling position causes a significant change in the knee joint, which differs from the first type of joint point, while falling backward on a sitting position causes a significant change in the hip joint, which differs from the first type of joint point. Therefore, if the displacement information of the third joint point exceeds the third displacement threshold, and the displacement direction of the third joint point is towards the front of the detected object, that is, the change in the knee joint's amplitude and direction meet the conditions for falling forward on a kneeling position, then the fall type is determined to be falling forward on a kneeling position; if the displacement information of the fourth joint point exceeds the fourth displacement threshold, and the displacement direction of the fourth joint point is towards the rear of the detected object, that is, the change in the hip joint's amplitude and direction meet the conditions for falling backward on a sitting position, then the fall type is determined to be falling backward on a sitting position.

[0105] In this embodiment, based on the displacement information of the second type of joints meeting the conditions for falling forward on a kneeling position or falling backward on a sitting position, forward kneeling falls and backward sitting falls are detected to obtain more accurate fall detection results and provide users with better reference.

[0106] In one embodiment, if the displacement information of the first type of joint does not meet the preset fall conditions, the non-fall behavior type of the detected object at the current moment is determined based on the displacement information of the first type of joint.

[0107] Non-falling behaviors can include lying down, walking quickly, bending over, squatting, lying down, sitting, running, and standing.

[0108] For example, when it is detected that the detected object has not fallen, the non-fall behavior type of the detected object is determined based on the displacement information of the first type of joint and the behavior threshold obtained by pre-training.

[0109] In this embodiment, the non-fall behavior type of the detected object can be determined based on the behavior threshold obtained through pre-training, providing users with clearer detection results.

[0110] In one embodiment, such as Figure 4 As shown, a fall detection method is provided. In this embodiment, it includes the following steps:

[0111] Step 401: Establish an absolute coordinate system;

[0112] Step 402, collect data;

[0113] Step 403, Data Training;

[0114] Step 404, fall detection.

[0115] For example,

[0116] The absolute coordinate system established in step 401 has its origin at one meter above the ground, the direction towards the camera as the positive x-axis, the direction to the right of the camera as the positive y-axis, and the direction perpendicular to the ground and upward as the positive z-axis.

[0117] The data collected in step 402 includes: various behavioral displacement data collected by sensors worn by the experimenter, various behavioral displacement data from video images obtained from cameras based on image processing technology, and publicly available human pose recognition datasets MobiAct and SisFall. A schematic diagram of the collected joint points is shown below. Figure 5 As shown, Table 1 is... Figure 5 A table of reference points for joints.

[0118] Table 1 Joint Comparison Table

[0119] 0 nose 1 Neck joint 2 Right shoulder joint 3 right elbow joint 4 right wrist 5 left shoulder joint 6 left elbow joint 7 left wrist 8 Lumbar joint 9 Right hip 10 Right knee 11 right ankle 12 Left hip 13 Left knee 14 left ankle 15 right eye 16 Left eye 17 right ear 18 left ear 19 left big toe 20 left little toe 21 left heel 22 right big toe 23 right little toe 24 right heel

[0120] Furthermore, the displacement data acquired in step 402 by the sensors and camera is displacement data on the x, y, and z axes, therefore it still needs to be integrated and processed. This is calculated using the following formula:

[0121]

[0122]

[0123] Sn is the integrated displacement, Vn is the integrated velocity, S represents displacement, V represents velocity, xi represents the i-th moment on the x-axis coordinate, and Δt is the time interval between two moments, which is set to 0.02 seconds in this embodiment.

[0124] In step 403, for data training, firstly, the displacement data of displacement distance, displacement linear velocity, and displacement angular velocity of various behaviors in the data are obtained. Through training, the weights of the three types of data are obtained, and the corresponding displacement data for each behavior is determined by weighted calculation, thereby determining the displacement change threshold for each behavior.

[0125] like Figure 6 As shown, step 404, fall detection, is implemented through the following steps:

[0126] Step 601: Determine whether the descent speed of the nose is greater than the set threshold.

[0127] Step 602: Determine whether the displacement distance of the neck joint relative to the previous frame is greater than the maximum displacement threshold during a normal squatting process.

[0128] Step 603: Determine whether the displacement distance of the shoulder joint relative to the previous frame is greater than the maximum displacement threshold during a normal squatting process.

[0129] Step 604: Determine whether the displacement distance of the waist joint relative to the previous frame is greater than the maximum displacement threshold during a normal squatting process.

[0130] Step 605: Determine the specific type of fall based on the displacement speed and distance of the nose, neck joint, shoulder joint, waist joint, knee joint, and hip joint, as well as different thresholds for fall types.

[0131] If the judgment results in steps 601, 602, 603 and 604 are negative, it is determined that there was no fall and no further steps are required; if the judgment results are all positive, it is determined that there was a fall and step 605 is executed.

[0132] In this embodiment, displacement change thresholds for each behavior are obtained through data collection and training. A multi-threaded approach is then employed: on one hand, displacement information is output for each human joint in each frame of the image; on the other hand, threshold judgment is performed, enabling more accurate and real-time updates of human behavior. Furthermore, different thresholds can be used to identify different behaviors, thus providing a more comprehensive reference.

[0133] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0134] Based on the same inventive concept, this application also provides a fall detection device for implementing the fall detection method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more fall detection device embodiments provided below can be found in the limitations of the fall detection method described above, and will not be repeated here.

[0135] In one embodiment, such as Figure 7 As shown, a fall detection device is provided, comprising: an image acquisition module, a position determination module, a displacement calculation module, and a behavior judgment module, wherein:

[0136] Image acquisition module 701 is used to acquire the real-time image of the detected object at the current moment and the real-time image at the previous moment;

[0137] The position determination module 702 is used to determine the current position information and the previous position information of each joint of the detected object based on the real-time image at the current moment and the real-time image at the previous moment.

[0138] The displacement calculation module 703 is used to determine the displacement information of the first type of joints of the detection object based on the current position information and the position information of the previous moment; the first type of joints are the feature joints on the detection object that can detect falls.

[0139] The behavior judgment module 704 is used to determine that the detected object is in a falling state at the current moment when the displacement information of the first type of joint meets the preset falling conditions.

[0140] In one embodiment, the displacement information includes displacement velocity and displacement distance; the first type of joint points includes a first joint point and a second joint point; the behavior judgment module 704 is further configured to acquire a velocity change threshold and a distance change threshold; if the displacement velocity of the first joint point exceeds the velocity change threshold and the displacement distance of the second joint point exceeds the distance change threshold, it is determined that the detected object is in a fallen state at the current moment.

[0141] In one embodiment, the behavior judgment module 704 is further configured to determine the displacement information of the second type of joint based on the current position information and the previous position information; and determine the fall type corresponding to the detected object based on the displacement information of the first type of joint and the displacement information of the second type of joint.

[0142] In one embodiment, the behavior determination module 704 is further configured to, when the displacement information of the second type of joint exceeds a preset displacement change threshold, determine the fall type corresponding to the detected object based on the displacement information of the second type of joint; and when the displacement information of the second type of joint does not exceed the preset displacement change threshold, determine the fall type corresponding to the detected object based on the displacement information of the first type of joint.

[0143] In one embodiment, the behavior determination module 704 is further configured to, when the displacement information of the first type of joint meets the first condition, determine that the fall type corresponding to the detected object is a direct fall; the first condition is that the displacement information exceeds the first displacement threshold; when the displacement information of the first type of joint meets the second condition, determine that the fall type corresponding to the detected object is a fall against the wall; the second condition is that the displacement information does not exceed the first displacement threshold.

[0144] In one embodiment, the second type of joint points includes a third joint point and a fourth joint point; the displacement change threshold includes a third displacement threshold and a fourth displacement threshold. The behavior judgment module 704 is further configured to determine that the fall type corresponding to the detected object is a forward kneeling fall when the displacement information of the third joint point meets a preset third condition; the third condition is that the displacement information of the third joint point exceeds the third displacement threshold and the displacement direction of the third joint point is towards the front of the detected object; and to determine that the fall type corresponding to the detected object is a backward sitting fall when the displacement information of the fourth joint point meets a preset fourth condition; the fourth condition is that the displacement information of the fourth joint point exceeds the fourth displacement threshold and the displacement direction of the fourth joint point is towards the rear of the detected object.

[0145] In one embodiment, the behavior determination module 704 is further configured to, when the displacement information of the first type of joint does not meet the preset fall conditions, determine the non-fall behavior type of the detected object at the current moment based on the displacement information of the first type of joint.

[0146] Each module in the aforementioned fall detection device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0147] In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 8 As shown, the computer device includes a processor, memory, input / output interfaces, a communication interface, a display unit, and an input device. The processor, memory, and input / output interfaces are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interfaces. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The input / output interfaces are used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements a fall detection method. The display unit is used to form a visually visible image and can be a display screen, a projection device, or a virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0148] Those skilled in the art will understand that Figure 8 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0149] In one embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above method embodiments.

[0150] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the steps in the above method embodiments.

[0151] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0152] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0153] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0154] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0155] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A fall detection method, characterized in that, The method includes: Acquire the real-time image of the detected object at the current moment and the real-time image at the previous moment; Based on the real-time image at the current moment and the real-time image at the previous moment, determine the current position information and the previous position information of each joint of the detected object; Based on the current position information and the previous position information, the displacement information of the first type of joints among the joints is determined; the first type of joints are characteristic joints on the detection object that can detect falls; the first type of joints includes the nose, neck joint, shoulder joint, and waist joint; the displacement information includes displacement velocity and displacement distance. If the displacement velocity of the nose exceeds the velocity change threshold, and the displacement distance of the neck joint, shoulder joint, and waist joint all exceed the distance change threshold, it is determined that the detected object is undergoing a downward displacement action with a greater intensity than squatting, and the detected object is determined to be in a fallen state at the current moment. Based on the current position information and the previous position information, the displacement information of the second type of joints among the joints is determined; the second type of joints includes the knee joint and the hip joint. If the displacement information of the second type of joint does not exceed the preset displacement change threshold, and the displacement information of the first type of joint exceeds the first displacement threshold, then the fall type corresponding to the detected object is determined to be a direct fall. If the displacement information of the second type of joint does not exceed the displacement change threshold, and the displacement information of the first type of joint does not exceed the first displacement threshold, then the fall type corresponding to the detected object is determined to be a fall against the wall. If the displacement information of the second type of joint exceeds the displacement change threshold, the displacement information of the knee joint exceeds the third displacement threshold, and the displacement direction of the knee joint is in front of the detected object, then the fall type corresponding to the detected object is determined to be a forward kneeling fall. If the displacement information of the second type of joint exceeds the displacement change threshold, the displacement information of the hip joint exceeds the fourth displacement threshold, and the displacement direction of the hip joint is behind the detected object, then the fall type corresponding to the detected object is determined to be a backward sitting fall.

2. The method according to claim 1, characterized in that, If the displacement velocity of the nose at the first type of joint exceeds a velocity change threshold, and the displacement distances of the neck, shoulder, and waist joints at the first type of joint all exceed distance change thresholds, before determining that the detected object has undergone a downward displacement action with a intensity greater than squatting, the procedure further includes: Obtain the speed change threshold and distance change threshold.

3. The method according to claim 1, characterized in that, The method further includes: If the displacement information of the first type of joint does not meet the preset fall conditions, the non-fall behavior type of the detected object at the current moment is determined based on the displacement information of the first type of joint. The preset fall conditions are that the displacement speed of the nose in the first type of joint exceeds the speed change threshold, and the displacement distance of the neck joint, shoulder joint and waist joint in the first type of joint all exceed the distance change threshold.

4. The method according to claim 3, characterized in that, The step of determining the non-fall behavior type of the detected object at the current moment based on the displacement information of the first type of joints includes: Based on the displacement information of the first type of joints and the behavior threshold obtained through pre-training, the non-fall behavior type of the detected object is determined.

5. The method according to claim 3, characterized in that, The types of non-falling behaviors include lying down, walking quickly, bending over, squatting, lying down, sitting, running, and standing.

6. A fall detection device, characterized in that, The device includes: The image acquisition module is used to acquire the real-time image of the detected object at the current moment and the real-time image at the previous moment; The position determination module is used to determine the current position information and the previous position information of each joint of the detected object based on the real-time image at the current moment and the real-time image at the previous moment. The displacement calculation module is used to determine the displacement information of a first type of joint among the joints based on the current position information and the previous position information; the first type of joint is a feature joint on the detection object that can detect a fall; the first type of joint includes the nose, neck joint, shoulder joint and waist joint; the displacement information includes displacement velocity and displacement distance. The behavior judgment module is used to determine that the detected object is in a falling state at the current moment when the displacement speed of the nose exceeds the speed change threshold and the displacement distance of the neck joint, the shoulder joint and the waist joint all exceed the distance change threshold. The behavior judgment module is further configured to determine the displacement information of a second type of joint among the joints based on the current position information and the previous position information; the second type of joint includes the knee joint and the hip joint; if the displacement information of the second type of joint does not exceed a preset displacement change threshold, and the displacement information of the first type of joint exceeds a first displacement threshold, then the fall type corresponding to the detected object is determined to be a direct fall; if the displacement information of the second type of joint does not exceed the displacement change threshold, and the displacement information of the first type of joint does not exceed the first displacement threshold, then the fall type corresponding to the detected object is determined to be a fall against a wall; if the displacement information of the second type of joint exceeds the displacement change threshold, the displacement information of the knee joint exceeds a third displacement threshold, and the displacement direction of the knee joint is in front of the detected object, then the fall type corresponding to the detected object is determined to be a forward kneeling fall; if the displacement information of the second type of joint exceeds the displacement change threshold, the displacement information of the hip joint exceeds a fourth displacement threshold, and the displacement direction of the hip joint is behind the detected object, then the fall type corresponding to the detected object is determined to be a backward sitting fall.

7. The apparatus according to claim 6, characterized in that, The behavior judgment module is also used to obtain speed change threshold and distance change threshold.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 5.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

Images