Non-contact monitoring model construction method, device, equipment and readable storage medium

Abstract

This application discloses a non-contact monitoring model construction method, apparatus, device, and readable storage medium. The method includes the following steps: acquiring a set of image information of a human lower limb in a preset movement posture; analyzing the overall movement posture of the human lower limb relative to a stationary object in the image information, and obtaining analysis results; based on the analysis results, constructing a target model for monitoring the movement posture of the human lower limb; the target model is used to determine whether the movement posture of the human lower limb is normal. This application realizes the acquisition of image information of a human lower limb in a preset movement posture, analysis of the image information, analysis of the overall movement posture of the human lower limb relative to a stationary object in the image, that is, analysis of the dynamic changes of the human lower limb during movement, obtaining analysis results, and using the analysis results to construct a target model for determining whether the movement posture of the human lower limb is normal.

Description

Technical Field

[0001] This application relates to the field of model building, and in particular to a non-contact monitoring model building method, apparatus, device, and readable storage medium. Background Technology

[0002] Currently, some athletes have problems with abnormal lower limb movement posture. These athletes need to participate in training that helps restore their lower limb movement posture. In such training, athletes are equipped with corresponding detection hardware devices at the muscles and joints of their lower limbs to detect their lower limb movement posture.

[0003] However, the use of detection hardware can hinder athletes' normal movement, have a certain impact on athletes' lower limb movements, and cause problems such as displacement of the detection hardware when athletes' lower limb muscles make movements, making it difficult to detect athletes' lower limb movement postures. Summary of the Invention

[0004] In view of this, this application provides a non-contact monitoring model construction method, device, equipment and readable storage medium, which aims to reduce the complexity of monitoring the lower limb movement posture of athletes.

[0005] To achieve the above objectives, this application provides a method for constructing a non-contact monitoring model, which includes the following steps:

[0006] Acquire a set of image information of the human lower limbs in a preset motion posture;

[0007] The overall motion posture of the human lower limbs relative to a stationary object in the image information is analyzed to obtain the analysis results;

[0008] Based on the analysis results, a target model for monitoring the lower limb movement posture of the human body is constructed; the target model is used to determine whether the lower limb movement posture of the human body is in a normal state.

[0009] For example, the analysis of the overall motion posture of the human lower limb relative to a stationary object in the image information, to obtain the analysis result, includes:

[0010] Based on the image information, the knee joint of the human lower limb is marked as the primary monitoring point, and the quadriceps femoris muscle of the human lower limb is marked as the secondary monitoring point; the image information is an image with three-dimensional coordinate parameters captured by a depth camera.

[0011] Extract the overall motion posture of the main monitoring point and the auxiliary monitoring point relative to the stationary object in the image information, and extract the first position parameter of the main monitoring point and the second position parameter of the auxiliary monitoring point to obtain the extraction result;

[0012] Based on the extraction results, the motion posture of the main monitoring point and the auxiliary monitoring point is analyzed to obtain the analysis results.

[0013] For example, based on the extraction results, the analysis of the relative motion between the main monitoring point and the auxiliary monitoring point to obtain the analysis results includes:

[0014] Based on the extraction results, a three-dimensional coordinate system is established with the main monitoring points as the origin.

[0015] Based on the first position parameter and the second position parameter, the first motion trajectory of the main monitoring point relative to the stationary object in the image information is calculated, and the second motion trajectory of the auxiliary monitoring point in the three-dimensional coordinate system is calculated.

[0016] Based on the first motion trajectory and the second motion trajectory, the currently monitored motion posture characteristics are analyzed to obtain the analysis results.

[0017] For example, the posture features include a first posture feature and a second posture feature. The step of analyzing the currently monitored motion posture features based on the first motion trajectory and the second motion trajectory to obtain the analysis results includes:

[0018] Based on the first motion trajectory and the second motion trajectory, the first posture characteristics of the knee joint are determined, and the second posture characteristics of the quadriceps femoris are determined.

[0019] The dynamic correlation between the first posture feature and the second posture feature is analyzed, and the changes in the coordination of the joints and muscles of the lower limbs during the movement are determined to obtain the analysis results.

[0020] For example, before constructing the target model for monitoring human lower limb movement posture based on the analysis results, the following steps are included:

[0021] Acquire relevant data about the human knee joint; the relevant data includes the range of motion of the human knee joint and the normal movement posture of the human knee joint;

[0022] Based on the relevant data, a simulation model simulating normal human knee joint activity is constructed; the simulation model is a kinematic model for constructing a target model for monitoring the lower limb movement posture of the human body.

[0023] For example, constructing a target model for monitoring human lower limb movement posture based on the analysis results includes:

[0024] The analysis results are output as training samples;

[0025] Construct an initial model for monitoring the movement posture of the human lower limbs;

[0026] The training samples are output to the initial model to obtain the target model after training.

[0027] For example, after constructing the target model for monitoring the lower limb movement posture of the human body based on the analysis results, the process includes:

[0028] Acquire images of the lower limb movements of the monitored subject;

[0029] The lower limb motion images are output to the target model so that the target model can analyze and determine the lower limb motion posture of the monitored person.

[0030] If it is determined that the lower limb movement posture of the monitored person does not conform to the preset movement posture, then the lower limb movement posture of the monitored person is determined to be abnormal.

[0031] For example, to achieve the above objectives, this application also provides a non-contact monitoring model construction device, the device comprising:

[0032] Acquisition module: Used to acquire a set of image information of the human lower limbs in normal movement;

[0033] Analysis module: used to analyze the overall motion posture of the human lower limbs relative to a stationary object in the image information, and obtain analysis results;

[0034] Construction module: used to construct a target model for monitoring the movement posture of the human lower limbs based on the analysis results; the target model is used to determine whether the human lower limb movement posture is normal.

[0035] For example, to achieve the above objectives, this application also provides a non-contact monitoring model building device, the device comprising: a memory, a processor, and a non-contact monitoring model building program stored in the memory and executable on the processor, the non-contact monitoring model building program being configured to implement the steps of the non-contact monitoring model building method as described above.

[0036] For example, to achieve the above objectives, this application also provides a computer-readable storage medium storing a non-contact monitoring model building program, which, when executed by a processor, implements the steps of the non-contact monitoring model building method as described above.

[0037] Compared to existing technologies that involve installing hardware devices on the lower limbs of the subject to detect their movement posture, this application acquires a set of image information of the lower limbs in a preset movement posture and analyzes the overall movement posture of the stationary object in the image information. This involves analyzing the posture information or features of the lower limbs in the movement posture to obtain an analysis result. Based on this analysis result, a target model is constructed to determine whether the lower limb movement posture is normal. By using this target model, the complexity of detecting lower limb movement posture is reduced, eliminating the need for hardware devices. In short, this method uses a target model to determine whether the lower limb movement posture is normal without the need for hardware devices, thereby reducing the complexity of the detection process. Attached Figure Description

[0038] Figure 1 This is a flowchart illustrating the first embodiment of the non-contact monitoring model construction method of this application;

[0039] Figure 2 This is a flowchart illustrating the second embodiment of the non-contact monitoring model construction method of this application;

[0040] Figure 3 This is a schematic diagram of the hardware operating environment involved in the embodiments of this application.

[0041] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0042] It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit this application.

[0043] This application provides a method for constructing a non-contact monitoring model, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the non-contact monitoring model construction method of this application.

[0044] This application provides an embodiment of a non-contact monitoring model construction method. It should be noted that although the logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order. For ease of description, the following omits the description of the various steps of the non-contact monitoring model construction method, which includes:

[0045] Step S110: Acquire a set of image information of the human lower limbs in a preset motion posture;

[0046] Building a model requires collecting relevant data. For example, when analyzing a dynamic target, it is necessary to collect the motion characteristic information of the target under different motion postures. Similarly, when analyzing a static target, it is necessary to collect the external appearance characteristic information of the target.

[0047] Acquiring a set of images of the human lower limbs in a preset movement posture is the relevant data required to build the model. Collecting images of continuous movements of the human lower limbs yields a set of image information. In this set of image information, there is a certain change in the movement posture of the human lower limbs in every two adjacent images.

[0048] The main functional points for generating movement postures in the human lower limbs are the knee joint and the most developed quadriceps femoris muscle in the thigh. Through the corresponding coordination and changes of the knee joint and quadriceps femoris muscle, the human lower limbs can produce different movement postures.

[0049] The preset state is the normal movement posture of the human lower limbs, such as the walking state in a natural state, the knee or thigh lifting action in a natural state, and the knee cushioning action when the leg falls in a natural state.

[0050] The natural state refers to the state in which the human lower limbs are not disturbed in any way and there is no functional abnormality in the human lower limbs.

[0051] For example, situations such as muscle atrophy or other functional impairments in the legs, or the use of exercise equipment such as sandbags on the legs, are not considered natural states.

[0052] Step S120: Analyze the overall motion posture of the human lower limbs relative to the stationary object in the image information, and obtain the analysis results;

[0053] When analyzing the movement posture of the human lower limbs, by analyzing and comparing a set of images in a continuous motion state, the changes in muscles and joints of the human lower limbs throughout the entire movement posture can be determined.

[0054] At the same time, there are stationary objects in the image information. For example, when acquiring image information of human lower limb movement, the image is not only of human lower limbs, but also includes objects that are stationary relative to the moving human lower limbs, such as the ground and background objects.

[0055] Using a stationary object as a reference, for example, the position of the same point is determined from all the images in the set of image information as a reference point. This reference point is a stationary point on the ground and exists in every image.

[0056] Using this benchmark as a reference point, the relative motion of the human lower limbs with respect to the stationary point is analyzed, and the overall motion posture of the human lower limbs is obtained.

[0057] Step S130: Based on the analysis results, a target model for monitoring the lower limb movement posture of the human body is constructed; the target model is used to determine whether the lower limb movement posture of the human body is in a normal state.

[0058] The analysis results are derived from the analysis of the human lower limbs in a preset movement posture. That is, the analysis results are derived from the analysis of normal walking, normal leg lifting and other movements of the human lower limbs. The analysis results meet the conditions for being used as training samples for a neural network model.

[0059] The target model is a neural network model, which is used to determine whether the lower limb movement posture of the human body is normal.

[0060] For example, the preset movement posture is the normal movement posture of the human lower limbs, that is, the preset movement posture is the normal state. The opposite movement posture is the abnormal movement posture, that is, the abnormal state. There are many kinds of abnormal states. The human lower limbs have different physiological functional disorders. For example, the human joints have physiological functional disorders: knee joint swelling, knee joint strain, etc., which prevent the human knee joint from moving normally. Another example is the human leg muscles have physiological functional disorders: leg muscle atrophy, leg muscle strain, functional loss of leg muscles caused by accidents (leg muscles cannot extend and contract normally, leg muscle connection is broken, etc.), which prevent the human leg muscles from moving normally.

[0061] For example, after constructing the target model for monitoring the lower limb movement posture of the human body based on the analysis results, the process includes:

[0062] Step a: Acquire images of the lower limb movements of the monitored subject;

[0063] After establishing a target model of the lower limb movement posture of the human body to be monitored, the target model is used to monitor the user and determine whether the lower limb movement posture of the monitored person is normal.

[0064] For example, the monitored individuals are those whose lower limb movement posture needs to be determined to be normal. These monitored individuals include: adolescent athletes who need to perform lower limb training, athletes who need to perform rehabilitation training after accidents or surgeries, etc.

[0065] Among them, adolescent athletes who need to undergo lower limb training need to be monitored to avoid situations caused by incorrect movement posture or failure to detect the problem themselves (for this type of athlete, the training samples of the target model need to be adjusted accordingly, and the training samples of the target model for this athlete should be the lower limb movement postures of the athlete at the current age).

[0066] For athletes who need rehabilitation training, the effectiveness of their rehabilitation is determined by monitoring their lower limb movement postures.

[0067] Lower limb motion images are image information of the monitored person’s lower limbs making movements, including the monitored person’s walking state, leg lifting state, leg lowering state and other related image information.

[0068] Step b: Output the lower limb motion image to the target model so that the target model can analyze and determine the lower limb motion posture of the monitored person;

[0069] After the lower limb motion image is output to the target model, the target model will perform feature analysis on the image and determine the motion state of the image.

[0070] Step c: If it is determined that the lower limb movement posture of the monitored person does not conform to the preset movement posture, then the lower limb movement posture of the monitored person is determined to be abnormal.

[0071] When judging the lower limb movement posture of the monitored person, the training samples of the target model are used as the comparison standard. The target model is trained by the analysis results, and the lower limb movement state in the analysis results is the preset movement posture. That is, the preset movement posture is the standard for the target model to judge the lower limb movement posture of the monitored person.

[0072] The preset movement posture is that the human lower limbs are in a normal movement state, i.e., the normal state.

[0073] If the lower limb movement posture of the monitored person meets the preset standard for movement posture, then the current lower limb movement posture of the monitored person is determined to be normal. Otherwise, the current lower limb movement posture of the monitored person is determined to be abnormal.

[0074] Compared to existing technologies that involve installing hardware devices on the lower limbs of the subject to detect their movement posture, this application acquires a set of image information of the lower limbs in a preset movement posture and analyzes the overall movement posture of the stationary object in the image information. This involves analyzing the posture information or features of the lower limbs in the movement posture to obtain an analysis result. Based on this analysis result, a target model is constructed to determine whether the lower limb movement posture is normal. By using this target model, the complexity of detecting lower limb movement posture is reduced, eliminating the need for hardware devices. In short, this method uses a target model to determine whether the lower limb movement posture is normal without the need for hardware devices, thereby reducing the complexity of the detection process.

[0075] For example, refer to Figure 2 , Figure 2 This is a flowchart illustrating the second embodiment of the non-contact monitoring model construction method of this application. Based on the first embodiment of the non-contact monitoring model construction method of this application described above, a second embodiment is proposed, wherein the method further includes:

[0076] Step S210: Based on the image information, mark the knee joint of the human lower limb as the main monitoring point and mark the quadriceps femoris muscle of the human lower limb as the auxiliary monitoring point; the image information is an image with three-dimensional coordinate parameters captured by a depth camera;

[0077] Image information includes overall structural information of the human lower limbs, such as the structure of the feet, calves, knees, and thighs.

[0078] The movement of the human lower limbs mainly involves the coordinated movements between the legs and knee joints, such as walking, running, and jumping.

[0079] The knee joint is one of the most important joints in the human lower limbs. All movements of the human lower limbs rely on the activity of the knee joint. Therefore, the most important key to detecting the movement of the human lower limbs is to detect the activity status of the human knee joint.

[0080] In addition, the quadriceps femoris muscle in the thigh is also a key point in the movement of the lower limbs. The quadriceps femoris is a muscle in the thigh located in front of the thigh muscles. To make the thigh strong, the first priority is to develop the quadriceps femoris, because the quadriceps femoris is one of the largest and most powerful muscles in the human body. Therefore, when testing the movement of the lower limbs, the knee joint and quadriceps femoris are used as monitoring points.

[0081] The knee joint was marked as the primary monitoring point, and the quadriceps femoris muscle as the secondary monitoring point.

[0082] The knee joint is the most complex joint in the human body, and its influence on the movement and posture of the lower limbs is obvious. Therefore, the knee joint is a major monitoring point for monitoring the movement of the lower limbs.

[0083] The quadriceps femoris is the most powerful muscle in the thigh. This type of muscle extends from the knee joint to the root of the thigh. As a muscle that generates muscle force, the quadriceps femoris has a crucial impact on the knee joint and is therefore used as an auxiliary monitoring point for monitoring human lower limb movement.

[0084] When monitoring the movement of the human lower limbs, the knee joint is used as the rotation point to make the thigh muscles and calf muscles work together to make the human lower limbs perform various movements.

[0085] The image information is obtained by using a depth camera. When taking a photo with a depth camera, not only is the image within the shooting range covered by the camera lens recorded, but the distance between each object in the image and the camera lens is also calculated through optical compensation, that is, the depth is calculated when the image is taken.

[0086] Based on the characteristics of the distance measurement generated when the depth camera takes pictures, the distance from each object in the image to the camera lens can be calculated. If a three-dimensional coordinate system is established with the camera lens as the origin, the position of each object in the image in the three-dimensional coordinate system can be deduced.

[0087] Step S220: Extract the overall motion posture of the main monitoring point and the auxiliary monitoring point relative to the stationary object in the image information, and extract the first position parameter of the main monitoring point and the second position parameter of the auxiliary monitoring point to obtain the extraction result;

[0088] The motion posture of the human lower limbs in the image is obtained by analyzing a set of image information. This posture is the relative motion posture with respect to stationary objects in the image, and the motion posture is analyzed by treating the human lower limbs as a whole.

[0089] Thus, the most intuitive overall movement posture of the human lower limbs can be extracted from the image information.

[0090] For example, when a human lower limb makes a leg-raising motion, relative to the stationary ground in the image information, the current position of the human leg in each of the acquired image information is extracted to form the posture change of the human leg gradually rising at each moment during the movement of the human lower limb.

[0091] Meanwhile, based on the characteristics of the images captured by the depth camera, the position parameters of the main monitoring point and the auxiliary monitoring point are calculated, which are the first position parameter and the second position parameter, respectively.

[0092] By extracting the first and second position parameters, the relative motion of the knee joint and quadriceps muscle in the overall movement of the human lower limb can be determined.

[0093] For example, during normal walking, in addition to the quadriceps muscle stretching and contracting when the leg is raised and lowered, the knee joint also rotates according to the movement of the leg muscles, moving the lower leg closer to or away from the thigh. At the same time, due to the structural characteristics of the knee joint, the knee joint will produce a certain angle of internal and external rotation, thereby achieving the effect of adjusting walking posture.

[0094] Step S230: Based on the extraction results, analyze the motion posture of the main monitoring point and the auxiliary monitoring point to obtain the analysis results.

[0095] Based on the extraction results, the motion posture of the main monitoring points and auxiliary monitoring points is analyzed. This analysis process mainly includes two aspects: first, the motion trend of the main monitoring points and auxiliary monitoring points as a whole. For example, when a person raises their leg, the knee joint and the thigh muscles rise together. Second, there is a relative motion posture between the main monitoring points and auxiliary monitoring points. For example, when a person walks forward, the thigh muscles are in a state of tightening and relaxing. When the thigh muscles produce corresponding dynamic changes, they will drive the knee joint to produce certain changes. At the same time, conversely, the current state of the knee joint will affect the state of the quadriceps.

[0096] For example, based on the extraction results, the analysis of the relative motion between the main monitoring point and the auxiliary monitoring point to obtain the analysis results includes:

[0097] Step d: Based on the extraction results, establish a three-dimensional coordinate system with the main monitoring points as the origin;

[0098] Based on the extraction results, the movement posture of the human lower limbs is monitored. The overall movement posture is calculated based on the relative motion of the human lower limbs and stationary objects in the image information. The relative motion of the main monitoring points and auxiliary monitoring points needs to be calculated by establishing a coordinate system.

[0099] The knee joint connects the thigh and calf of the human lower limb. A three-dimensional coordinate system is established with the knee joint as the origin. This system is used to calculate the contraction and extension of the quadriceps femoris muscle in the thigh when the human lower limbs are in motion.

[0100] Step e: Based on the first position parameter and the second position parameter, calculate the first motion trajectory of the main monitoring point relative to the stationary object in the image information, and calculate the second motion trajectory of the auxiliary monitoring point in the three-dimensional coordinate system;

[0101] The first motion trajectory is the motion trajectory generated by the knee joint relative to the stationary object. A static coordinate system is established with the stationary object as the origin, while the three-dimensional coordinate system established with the knee joint as the origin is the dynamic coordinate system, which is the motion point. Therefore, when calculating the motion trajectory, the calculation is based on the relative motion between the dynamic coordinate system and the static coordinate system.

[0102] The first motion trajectory is the result calculated relative to the static coordinate system, that is, the motion trajectory of the knee joint. The main monitoring point of the knee joint and the auxiliary monitoring point of the quadriceps femoris muscle are mainly the motion posture of the entire thigh. Therefore, the overall motion trajectory of the thigh part of the human lower limb is calculated directly from the motion trajectory of the knee joint of the main monitoring point in the static coordinate system.

[0103] For example, when a person walks, their thighs lift up and then fall down, repeating this motion to complete the walking action.

[0104] For example, when a person climbs stairs, the thighs will drive the knees and calves upwards together.

[0105] The second motion trajectory is the motion trajectory of the auxiliary monitoring point in the three-dimensional coordinate system. The knee joint will produce flexion, extension, internal rotation and external rotation movements, that is, during the movement of the human lower limbs, the knee joint will produce relative movement with the quadriceps femoris muscle of the thigh.

[0106] Using the knee joint as the stationary coordinate origin, the movement trajectory of the quadriceps muscle relative to the knee joint is calculated.

[0107] Step f: Based on the first motion trajectory and the second motion trajectory, analyze the currently monitored motion posture characteristics to obtain the analysis results.

[0108] Based on the first and second motion trajectories, the motion posture characteristics of the human lower limbs in the motion trajectory are determined. The motion posture characteristics are the posture characteristics of the human lower limbs during movement.

[0109] For example, when a person raises their thigh upwards and their lower leg is naturally perpendicular to the ground, the knee joint and the quadriceps femoris muscle of the thigh will produce corresponding movement postures. These movement postures determine the overall movement posture characteristics of the human lower limbs from the first movement trajectory, and determine the movement posture characteristics of the knee joint or quadriceps femoris muscle of the human lower limbs from the second movement trajectory.

[0110] The first movement trajectory includes the swing angle and lifting angle of the thigh.

[0111] The second trajectory includes knee flexion, extension, internal rotation, external rotation, and straightening.

[0112] The motion posture characteristics were analyzed to obtain the analysis results.

[0113] For example, the posture features include a first posture feature and a second posture feature. The step of analyzing the currently monitored motion posture features based on the first motion trajectory and the second motion trajectory to obtain the analysis results includes:

[0114] Step g: Based on the first motion trajectory and the second motion trajectory, determine the first posture characteristics of the knee joint and the second posture characteristics of the quadriceps femoris muscle;

[0115] The primary postural characteristics of the knee joint include the relative motion of the knee joint with respect to a stationary object.

[0116] The second postural characteristics of the quadriceps include the relative motion of the quadriceps relative to a stationary object, and the relative motion of the quadriceps relative to the knee joint.

[0117] The quadriceps, as a muscle that generates force inside the thigh, does not move relative to the thigh. It only stretches and contracts inside the thigh, and this movement is manifested through the stretching of the skin.

[0118] Therefore, the second posture characteristics of the quadriceps femoris, combined with changes in the skin of the thigh, can also be used to measure its movement posture.

[0119] Step h: Analyze the dynamic correlation between the first posture feature and the second posture feature, and determine the changes in the coordination of joints and muscles of the human lower limbs during movement to obtain the analysis results.

[0120] By combining the first and second posture characteristics, the coordination changes of the knee joint and muscles during human lower limb movement can be determined.

[0121] For example, when a person walks forward, the thighs rise and fall, and the knee joints flex and extend. At the same time, to maintain balance while walking, the knee joints will also rotate internally or externally (internal or external rotation of the knee joint affects the swing angle of the foot relative to the torso).

[0122] In this embodiment, the process of analyzing the lower limb movement posture of the human body based on image information is carried out. The knee joint and quadriceps muscle of the human lower limb are used as the main monitoring point and auxiliary monitoring point, respectively. At the same time, a three-dimensional coordinate system is established with the main monitoring point as the origin. The relative motion between the main monitoring point and the auxiliary monitoring point is analyzed. Combined with the relative motion of the human lower limb relative to a stationary object, the motion trajectory and movement posture of the human lower limb during the movement process are determined, thereby accurately analyzing the movement posture of the human lower limb.

[0123] For example, based on the first and second embodiments of the non-contact monitoring model construction method of this application described above, a third embodiment is proposed, wherein the method further includes:

[0124] Step i: Output the analysis results as training samples;

[0125] The analysis results include information such as the motion trajectory and motion posture characteristics of the main monitoring points and auxiliary monitoring points. This information encompasses relevant data on the motion posture of the human lower limbs, meaning that the analysis results meet the conditions for being used as training samples.

[0126] Step j: Construct an initial model for monitoring the movement posture of the human lower limbs;

[0127] The initial model was an untrained neural network model, which is currently unable to monitor the movement posture of the human lower limbs.

[0128] Meanwhile, the establishment of the initial model includes two aspects: one is to establish the connection and judgment logic between neurons, and the other is to establish a model that meets the requirements of kinematic feature analysis. This kinematic feature analysis mainly targets the kinematic features of the joints, muscles and overall leg movements of the human lower limbs.

[0129] Step k: Output the training samples to the initial model to obtain the trained target model.

[0130] After the initial model is established, the training samples composed of the analysis results are output to the initial model, so as to train the initial model with the training samples and obtain a target model that has been trained and can be used to judge the lower limb movement posture of the human body.

[0131] For example, before constructing the target model for monitoring human lower limb movement posture based on the analysis results, the following steps are included:

[0132] Step 1: Obtain relevant data about the human knee joint; the relevant data includes the range of motion of the human knee joint and the normal movement posture of the human knee joint;

[0133] As the most complex joint in the human body, the knee joint has various types of movements and range of motion, which are all relevant data.

[0134] The knee joint is in a neutral position when fully extended. The normal range of motion of the knee joint is 135 degrees when flexed and 0 degrees when extended. The knee joint can also hyperextend, with a hyperextension angle of approximately 10 degrees. Additionally, when the knee joint is flexed to 90 degrees, it can perform axial rotation: internal rotation (towards the toes, with an internal rotation angle of 0-30 degrees) and external rotation (towards the toes, with an external rotation angle of 0-40 degrees). If the knee joint is flexed beyond 90 degrees, the range of rotational motion of the lower limb will be significantly reduced.

[0135] Step m: Based on the relevant data, construct a simulation model that simulates the normal activity of the human knee joint; the simulation model is a kinematic model for constructing a target model for monitoring the lower limb movement posture of the human body.

[0136] Based on relevant data, the simulation model is constructed by relying on the kinematic data related to the knee joint, such as the flexion-extension angle and rotation angle of the knee joint, in order to simulate the effect of normal free movement of the human knee joint.

[0137] In this embodiment, a simulation model is established to demonstrate the state of human lower limb movement. At the same time, a neural network model is established on the basis of the simulation model to generate a neural network model based on kinematic analysis. The image information of the human lower limb movement posture is used to perform in-depth analysis on its movement posture. On the one hand, the movement posture of the knee joint and quadriceps femoris muscle is analyzed, and on the other hand, the overall movement posture of the lower limb is analyzed to determine whether the human lower limb movement posture is normal.

[0138] Furthermore, this application also provides a non-contact monitoring model construction device, which includes:

[0139] Acquisition module: Used to acquire a set of image information of the human lower limbs in normal movement;

[0140] Analysis module: used to analyze the overall motion posture of the human lower limbs relative to a stationary object in the image information, and obtain analysis results;

[0141] Construction module: used to construct a target model for monitoring the movement posture of the human lower limbs based on the analysis results; the target model is used to determine whether the human lower limb movement posture is normal.

[0142] For example, the analysis module includes:

[0143] The labeling submodule is used to label the knee joint of the human lower limb as the primary monitoring point and the quadriceps femoris muscle of the human lower limb as the auxiliary monitoring point based on the image information; the image information is an image with three-dimensional coordinate parameters captured by a depth camera.

[0144] Extraction submodule: used to extract the overall motion posture of the main monitoring point and the auxiliary monitoring point relative to the stationary object in the image information, and to extract the first position parameter of the main monitoring point and the second position parameter of the auxiliary monitoring point to obtain the extraction result;

[0145] Analysis submodule: Based on the extraction results, it analyzes the motion posture of the main monitoring point and the auxiliary monitoring point to obtain analysis results.

[0146] For example, the analysis submodule includes:

[0147] Establishment Unit: Used to establish a three-dimensional coordinate system based on the extraction results, with the main monitoring points as the origin;

[0148] Calculation unit: used to calculate the first motion trajectory of the main monitoring point relative to the stationary object in the image information based on the first position parameter and the second position parameter, and to calculate the second motion trajectory of the auxiliary monitoring point in the three-dimensional coordinate system;

[0149] Analysis unit: used to analyze the currently monitored motion posture characteristics based on the first motion trajectory and the second motion trajectory, and obtain analysis results.

[0150] For example, the analysis unit includes:

[0151] Determining subunit: used to determine the first posture features of the knee joint and the second posture features of the quadriceps muscle based on the first motion trajectory and the second motion trajectory;

[0152] Analysis subunit: used to analyze the dynamic correlation between the first posture feature and the second posture feature, and to determine the changes in the coordination of the joints and muscles of the human lower limbs during movement, and to obtain the analysis results.

[0153] For example, the building module includes:

[0154] The first acquisition submodule is used to acquire relevant data about the human knee joint; the relevant data includes the range of motion of the human knee joint and the normal movement posture of the human knee joint.

[0155] The first construction submodule is used to construct a simulation model that simulates the normal activity of the human knee joint based on the relevant data; the simulation model is a kinematic model for constructing a target model for monitoring the lower limb movement posture of the human body.

[0156] First output submodule: Used to output the analysis results as training samples;

[0157] The second construction submodule is used to build an initial model for monitoring the movement posture of the human lower limbs;

[0158] The second output submodule is used to output the training samples to the initial model to obtain the trained target model.

[0159] The second acquisition submodule is used to acquire lower limb movement images of the monitored subject.

[0160] The third output submodule is used to output the lower limb motion image to the target model so that the target model can analyze and determine the lower limb motion posture of the monitored person.

[0161] Determine submodule: If it is determined that the lower limb movement posture of the monitored person does not conform to the preset movement posture, then the lower limb movement posture of the monitored person is determined to be abnormal.

[0162] The specific implementation of the non-contact monitoring model construction device of this application is basically the same as the embodiments of the non-contact monitoring model construction method described above, and will not be repeated here.

[0163] In addition, this application also provides a non-contact monitoring model building device. For example... Figure 3 As shown, Figure 3 This is a schematic diagram of the hardware operating environment involved in the embodiments of this application.

[0164] For example, Figure 3 This is a schematic diagram of the hardware operating environment for building a non-contact monitoring model.

[0165] like Figure 3 As shown, the non-contact monitoring model building device may include a processor 301, a communication interface 302, a memory 303, and a communication bus 304. The processor 301, the communication interface 302, and the memory 303 communicate with each other through the communication bus 304. The memory 303 is used to store computer programs. When the processor 301 executes the program stored in the memory 303, it implements the steps of the non-contact monitoring model building method.

[0166] The communication bus 304 mentioned in the aforementioned non-contact monitoring model construction device can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus 304 can be divided into an address bus, a data bus, and a control bus, etc. For ease of illustration, it is represented by only one thick line in the figure, but this does not indicate that there is only one bus or one type of bus.

[0167] The communication interface 302 is used for communication between the aforementioned non-contact monitoring model building device and other devices.

[0168] The memory 303 may include random access memory (RMD) or non-volatile memory (NM), such as at least one disk storage device. Optionally, the memory 303 may also be at least one storage device located remotely from the aforementioned processor 301.

[0169] The processor 301 mentioned above can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0170] The specific implementation method of the non-contact monitoring model construction device in this application is basically the same as the embodiments of the above-mentioned non-contact monitoring model construction method, and will not be repeated here.

[0171] Furthermore, embodiments of this application also propose a computer-readable storage medium storing a non-contact monitoring model building program, which, when executed by a processor, implements the steps of the non-contact monitoring model building method described above.

[0172] The specific implementation of the computer-readable storage medium in this application is basically the same as the embodiments of the non-contact monitoring model construction method described above, and will not be repeated here.

[0173] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0174] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0175] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a readable storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0176] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A non-contact monitoring model construction method characterized by comprising: The non-contact monitoring model construction method comprises the following steps: Obtain a set of image information of a human lower limb in a preset motion posture; Analyze the overall motion posture of the human lower limb relative to the stationary object in the image information to obtain an analysis result; Based on the analysis result, a target model for monitoring the motion posture of the human lower limb is constructed; the target model is used to determine whether the motion posture of the human lower limb is normal; The analysis of the overall motion posture of the human lower limb relative to the stationary object in the image information to obtain an analysis result comprises: Based on the image information, mark the knee joint of the human lower limb as a main monitoring point and mark the quadriceps femoris of the human lower limb as an auxiliary monitoring point; the image information is an image with three-dimensional coordinate parameters captured by a depth camera; Extract the overall motion posture of the main monitoring point and the auxiliary monitoring point relative to the stationary object in the image information, and extract the first position parameter of the main monitoring point and the second position parameter of the auxiliary monitoring point to obtain an extraction result; Based on the extraction result, analyze the motion posture of the main monitoring point and the auxiliary monitoring point to obtain an analysis result.

2. The non-contact monitoring model building method of claim 1, wherein, The analysis of the relative motion between the main monitoring point and the auxiliary monitoring point based on the extraction result to obtain an analysis result comprises: Based on the extraction result, a three-dimensional coordinate system is established with the main monitoring point as the coordinate origin; Based on the first position parameter and the second position parameter, the first motion trajectory of the main monitoring point relative to the stationary object in the image information is calculated, and the second motion trajectory of the auxiliary monitoring point in the three-dimensional coordinate system is calculated; Based on the first motion trajectory and the second motion trajectory, analyze the currently monitored motion posture features to obtain an analysis result.

3. The non-contact monitoring model building method of claim 2, wherein, The posture features include first posture features and second posture features, and the analysis of the currently monitored motion posture features based on the first motion trajectory and the second motion trajectory to obtain an analysis result comprises: Based on the first motion trajectory and the second motion trajectory, determine the first posture features of the knee joint and the second posture features of the quadriceps femoris; Analyze the dynamic correlation of the first posture features and the second posture features, and determine the joint and muscle coordination changes of the human lower limb during the motion process to obtain an analysis result.

4. The non-contact monitoring model building method of claim 1, wherein, Before the construction of the target model for monitoring the motion posture of the human lower limb based on the analysis result, the following steps are included: Obtain relevant data of the human knee joint; the relevant data includes the range of motion of the human knee joint and the normal motion posture of the human knee joint; Based on the relevant data, a simulation model simulating the normal activity of the human knee joint is constructed; the simulation model is a kinematic model for constructing the target model for monitoring the motion posture of the human lower limb.

5. The non-contact monitoring model building method of claim 1, wherein, The construction of the target model for monitoring the motion posture of the human lower limb based on the analysis result comprises: Output the analysis result as a training sample; Construct an initial model for monitoring the motion posture of the human lower limb; Output the training sample to the initial model to obtain a trained target model.

6. The non-contact monitoring model building method of claim 1, wherein, After the target model for monitoring the motion posture of the lower limbs of the human body is constructed based on the analysis result, the method comprises the following steps: acquiring a lower limb motion image of a monitored person; outputting the lower limb motion image to the target model, so that the target model analyzes and determines the motion posture of the lower limbs of the monitored person; if it is determined that the motion posture of the lower limbs of the monitored person does not conform to the preset motion posture, determining that the motion posture of the lower limbs of the monitored person is an abnormal state.

7. A non-contact monitoring model construction apparatus characterized by comprising: The non-contact monitoring model construction device comprises: an acquisition module configured to acquire a set of image information of the lower limbs of the human body in a normal motion state; an analysis module configured to analyze the overall motion posture of the lower limbs of the human body relative to a stationary object in the image information, to obtain an analysis result; a construction module configured to construct a target model for monitoring the motion posture of the lower limbs of the human body based on the analysis result; the target model is used to determine whether the motion posture of the lower limbs of the human body is in a normal state; The non-contact monitoring model construction device is used to achieve the following steps: based on the image information, marking the knee joint of the lower limbs of the human body as a main monitoring point, and marking the quadriceps femoris of the lower limbs of the human body as an auxiliary monitoring point; the image information is an image with three-dimensional coordinate parameters captured by a depth camera; extracting the overall motion posture of the main monitoring point and the auxiliary monitoring point relative to the stationary object in the image information, and extracting a first position parameter of the main monitoring point and a second position parameter of the auxiliary monitoring point, to obtain an extraction result; based on the extraction result, analyzing the motion postures of the main monitoring point and the auxiliary monitoring point, to obtain an analysis result.

8. A non-contact monitoring model construction device, characterized by, The device comprises a memory, a processor, and a non-contact monitoring model construction program stored on the memory and executable on the processor, and the non-contact monitoring model construction program is configured to implement the steps of the non-contact monitoring model construction method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a non-contact monitoring model construction program, and the non-contact monitoring model construction program is executed by the processor to implement the steps of the non-contact monitoring model construction method according to any one of claims 1 to 6.

Images