22 results about "Right knee" patented technology

To provide a body support device for a toilet capable of reducing splashing and splashing of urine. [Solution] A toilet body support device 100 attached to a Western-style toilet 500 comprises a first base 111 located on the floor surface FL on the right side of the Western-style toilet, a second base located on the floor surface on the right side of the Western-style toilet, a first support pillar 121 extending upward from the first base, a second support pillar extending upward from the second base, a first knee pad 131 attached to the upper end of the first support pillar to support the right knee 810 of a user who is squatting towards the Western-style toilet to urinate, and a second knee pad attached to the upper end of the second support pillar to support the left knee of a user who is squatting towards the Western-style toilet to urinate.

The invention discloses a human body posture recognition method, which comprises the following steps of: 1, acquiring the number of people to be recognized and file data of human body posture key points corresponding to each person by utilizing a posture key point classification model; 2, screening key points of the human body posture, and screening six main key nodes of a left elbow, a right elbow, a left hip, a right hip, a left knee and a right knee; 3, judging whether the coordinates of the three auxiliary key points of each main key node are complete or not; 4, recognizing a human body posture; and 5, determining a human body shielding condition. According to the method, the human body posture is judged by determining the angle of each key point, learning the position of each joint of the human body, the angle of limbs and other complex features, calculating the angles of six body key nodes of left and right elbows, left and right hips and left and right knees and refining different conditions, and the problems that a traditional human body posture recognition technology is few in recognizable posture type and high in recognition efficiency are solved. And the precision and the speed of human body posture recognition are improved.

The invention relates to a sitting body forward bending action standard detection method, and belongs to the technical field of action standard detection, and the method comprises the steps: detecting left and right ankle coordinates, left and right knee coordinates, left and right hipbone coordinates, left and right shoulder coordinates, left and right elbow coordinates, left and right wrist coordinates, and hand tip coordinates of left and right hands of a person; when the 14 coordinates are detected, judging that the personnel has made preparation work, and starting to detect the forward bending action standard of the sitting body; 14 coordinates of each frame are recorded in the detection process; determining a completion score of each frame according to the hand tip coordinates of the left and right hands; judging whether the person completes the action according to the completion score of each frame, and obtaining a valid score or an invalid score; determining an upper limb termination score and a lower limb termination score of each frame, and determining a termination score of each frame according to the upper limb termination score and the lower limb termination score of each frame; and judging whether the action is illegal or not according to the termination score of each frame, and if the judgment result is that the action is illegal, terminating the assessment, and judging that the score is invalid.

The invention relates to the field of intelligent identification, in particular to a complex maintenance scene-oriented virtual human intelligent motion control method, which comprises the following steps of: performing virtual human physical modeling, and establishing a virtual human model based on a GB10000 human body size standard and a human-machine ergonomics standard, wherein the kinematics architecture divides a virtual human into an upper body and a lower body, the upper body comprises eight joints including a spine, a neck, left / right shoulders, left / right elbows and left / right wrists, the lower body comprises six joints including left / right thigh roots, left / right knees and left / right ankles, and all the joints are abstracted into spherical hinge, hinge or universal joint kinematics mechanisms respectively. Joint angle constraints are set; the technical problem that a mode of designing animations in advance and combining prefabricated animations is adopted on the basis of action methods such as a state machine and a behavior tree, calculation and analysis generally face conventional maintainability tasks, and special and complex maintenance scenes are difficult to meet is solved.

A device is disclosed for facilitating a therapeutic exercise of a user. The device includes a stabilizing bracket. The stabilizing bracket includes a horizontal member, a fixed position left side vertical member mechanically coupled at a left end of the horizontal member, a fixed position right side vertical member mechanically coupled at a right end of the horizontal member, and a user adjustable vertical member configured to be mechanically coupled in a plurality of horizontal member locations between the fixed position left side vertical member and the fixed position right side vertical member. The device further includes an upper rigid member configured to be mechanically coupled with the user adjustable vertical member and further configured for contacting a right knee of the user; and a lower rigid member configured to be mechanically coupled with the user adjustable vertical member and further configured for contacting a right instep of the user.

The systems, methods, and apparatus disclosed herein can incorporate one or more real-world performers performing in a venue into interactive content. These systems, methods, and apparatus can identify one or more joints or ligaments of one or more real-world performers, for example, the left shoulder, the right knee, etc., from an image or series of images of one or more real-world performers. These joints or ligaments can be represented as one or more performer markers. These systems, methods, and apparatus can generate one or more three-dimensional models of one or more real-world performers in three-dimensional space from one or more performer markers.

The application belongs to the technical field of exoskeleton robots, and particularly relates to a single-motor flexible walking aid exoskeleton, which comprises a binding assembly, a driving device and a linkage device. In the application, the forward and reverse rotation of a single motor and the cooperation between a left knee clamp and a right knee clamp are controlled by a control unit according to the joint motion state of a human body, so as to drive a forward winding wheel set and a reverse winding wheel set to rotate respectively, rotate and recover a left leg driving line and a right leg driving line corresponding to a left leg and a right leg respectively, and then respectively pull the left knee clamp and the right knee clamp, so that the double legs of the driving exoskeleton complete the whole gait walking cycle. In this way, only a single motor can be used to control two legs at the same time, the mechanical structure of the exoskeleton is simplified, and the manufacturing and maintenance costs are reduced.

Systems, methods, and apparatuses disclosed herein can incorporate one or more real-world performers that are performing within a venue into interactive content. These systems, methods, and apparatuses can identify one or more joints or ligaments, for example, left shoulders, right knees, among others, of the one or more real-world performers from the image, or the series of images, of the one or more real-world performers. These joints or ligaments can be represented as one or more performer markers. These systems, methods, and apparatuses can generate one or more three-dimensional models of the one or more real-world performers in a three-dimensional space from the one or more performer markers. These systems, methods, and apparatuses can apply the one or more three-dimensional models of the one or more real-world performers to the one or more virtual characters in the three-dimensional space. These systems, methods, and apparatuses can render the one or more virtual characters from the three-dimensional space into a two-dimensional space of the interactive content to integrate the one or more real-world performers into the interactive content.

A chair for facilitating cross-legged seating is disclosed. The chair includes a seat portion defining an upper surface positioned towards rear end of the chair, and a leg support portion positioned towards a front end of the chair. The leg support portion defines a base, a left sidewall, and a right sidewall. The left sidewall defines a left inclined surface configured to support a left knee region of the user, and a left chamber configured to support a right foot of the user. The right sidewall defines a right inclined surface configured to support a right knee region of the user, and a right chamber configured to support a right foot of the user.

To provide a technology capable of calculating a physical specificity attention amount regarding at least either a left food part or a right foot part of a subject.SOLUTION: An attention amount calculation device includes: a light source control part configured to light a left light source part and a right light source part; a detection part configured to be capable of detecting that a pressing part is pressed by the subject; and a calculation part configured to calculate a physical specificity attention amount on the basis of a reaction time. The left light source part may be disposed at a position overlapping the subject's left foot part in a left knee extension state. The right light source part may be disposed at a position overlapping the subject's right foot part in a right knee extension state. The calculation part is configured at least to calculate a physical specificity attention amount with respect to the subject's left foot part on the basis of expression (1) or to calculate a physical specificity attention amount of the subject's right foot part on the basis of expression (2): Ta(LL)-Ta(EL) (1), Ta(RR)-Ta(ER) (2).SELECTED DRAWING: Figure 1

The invention provides a rehabilitation robot based on a non-circular gear train-five-bar mechanism, and relates to the technical field of rehabilitation medical instruments, the rehabilitation robot comprises a rack, a left rehabilitation mechanism and a right rehabilitation mechanism; the left rehabilitation mechanism and the right rehabilitation mechanism are symmetrical; the left rehabilitation mechanism comprises a left non-circular wheel train-five-rod mechanism, a left leg connecting rod and a left knee bandage; the right rehabilitation mechanism comprises a right non-circular wheel train-five-rod mechanism, a right leg connecting rod and a right knee bandage. The non-circular wheel train-five-bar mechanism is used for accurately reproducing the motion posture of the knee joint, the problems that the fitness between the motion track of an existing rehabilitation robot and the natural gait of the human body is insufficient, the control torque is large and the like can be effectively solved, and the non-circular wheel train-five-bar mechanism has the advantages of being simple in structure and low in manufacturing cost.

This invention discloses a biomimetic model of driver injury based on the physical characteristics of a Chinese male at the 50th percentile. The model is based on a biomimetic model of a car occupant with the physical characteristics of a Chinese male at the 50th percentile. The occupant biomimetic model is modified and adjusted according to the driver's sitting posture while driving, and key angles of the human body structure and the physiological curvature of the spine are defined. The key angles include: head angle, trunk angle, upper arm angle, forearm angle, left and right hip angles, left and right knee angles, left and right leg angles, and ankle angle. The physiological curvature of the spine includes: cervical curvature, full thoracic curvature, upper thoracic curvature, lower thoracic curvature, and lumbar curvature. This invention also discloses a method for constructing the above-mentioned biomimetic model of injury, which has wide application scenarios.

An anti-glare method for a medical scanning bed and a computer program product thereof are disclosed, the method comprising: identifying a plurality of articulation points of a patient associated with implantation using an identification model, where the articulation points comprise articulation points of a face, hip, or lower limbs, where the face comprises eyes, mouth, or nose; the hip comprises a left hip and a right hip; the lower limbs comprise a left knee, a left ankle, a right knee and a right ankle; obtaining the confidence coefficient of each joint point; comparing each confidence coefficient with a preset confidence coefficient threshold value; and when the comparison accords with the preset condition, controlling the intensity of the light projected above the scanning bed. Through intelligent state recognition and automatic adjustment, the glare problem in the medical scanning process is effectively prevented, and the comfort level and the medical examination efficiency are improved.

The invention relates to a wearable amphibious assistance and control system which is characterized in that an exoskeleton right structure assembly is fixed on the upper and lower sides of a right leg knee joint of a user, an exoskeleton left structure assembly is fixed on the upper and lower sides of a left leg knee joint of the user, and a right propeller assembly is fixed on the upper part of the exoskeleton right structure assembly; the left propeller assembly is fixed to the upper portion of the exoskeleton left structure assembly. The multifunctional waist bag is fixed to the waist of a user, an integrated electric control system is installed in the multifunctional waist bag, the integrated electric control system is connected with the exoskeleton right structure assembly and the exoskeleton left structure assembly through watertight cables, and a land movement mode and / or an underwater movement mode are / is achieved so that the assisting function can be achieved for left and right knee joints of the user; the integrated electric control system is connected with the right propeller assembly and the left propeller assembly through watertight cables, and the diving propelling function on a user in the underwater movement mode is achieved. The structure portability of the power assisting device can be improved, and various power assisting propelling functions of the propeller on land and in water can be expanded.

The application belongs to the technical field of graph neural networks, and particularly relates to a bone joint health state evaluation method and system based on image graph representation and a graph neural network; the method comprises the following steps: acquiring bone joint images with labels; positioning the knee part in the bone joint images, and cutting each image into a left knee image and a right knee image; performing graph representation on each image to obtain node information, edge information and an adjacency matrix of the bone joint image; inputting a training set into a bone joint health state evaluation model based on a graph neural network for training, and evaluating the bone joint health grades of a test set to obtain the best evaluation result and the model; the model is used to evaluate the bone joint health grades, and a heat map of an image attention area in the evaluation process of the graph neural network is visualized; the application has high evaluation precision, good interpretability, high practicability and a good application prospect.

A humanoid robot, including an upper portion including a torso, a head, and left and right arm assemblies; a lower portion including left and right lower leg assemblies; a central portion coupled to the upper portion and including a spine, a pelvis, and left and right upper leg assemblies. Each upper leg assembly includes a hip flex actuator coupled to the pelvis and having a hip flex axis positioned at a non-zero downward angle relative to a transverse plane of the humanoid robot. The upper leg assembly also includes left and right knee actuators, wherein each knee actuator is an electric rotary actuator, has a knee axis that is co-planar with the hip flex axis, and has a momentary peak torque greater than a momentary peak torque of the torso twist actuator. The humanoid robot lacks a distinct torso pitch actuator and wherein the hip flex actuators are collectively arranged to provide for pitch movements of the torso.

Systems, methods, and apparatuses disclosed herein can incorporate one or more real-world performers that are performing within a venue into interactive content. These systems, methods, and apparatuses can identify one or more joints or ligaments, for example, left shoulders, right knees, among others, of the one or more real-world performers from the image, or the series of images, of the one or more real-world performers. These joints or ligaments can be represented as one or more performer markers. These systems, methods, and apparatuses can generate one or more three-dimensional models of the one or more real-world performers in a three-dimensional space from the one or more performer markers. These systems, methods, and apparatuses can apply the one or more three-dimensional models of the one or more real-world performers to the one or more virtual characters in the three-dimensional space. These systems, methods, and apparatuses can render the one or more virtual characters from the three-dimensional space into a two-dimensional space of the interactive content to integrate the one or more real-world performers into the interactive content.

The invention discloses a wheel-foot switchable humanoid robot leg configuration. The humanoid robot leg configuration comprises a left side leg configuration and a right side leg configuration which are symmetrical and have the same structure; wherein the left leg structure comprises a right thigh assembly, a robot right knee hub motor module, a right shank assembly, a right shank driven wheel module and a right foot assembly from top to bottom; the right leg structure comprises a left thigh assembly, a left knee hub motor module, a left shank assembly, a left shank driven wheel module and a left foot assembly from top to bottom. According to the design that the wheel state and the foot state can be switched, the maneuverability of the humanoid robot on a long straight road or a slow slope or other specific road conditions is improved.

Disclosed herein is a femoral component (100) of a knee arthroplasty system. The femoral component may be a multi-radius (MR) component having multiple centers of rotation within the functional knee range of motion. The femoral component may include at least two of a mid-flexion radius (141), an extension radius (142), or a deep flexion radius (140). The femoral component may include an articulate surface (112) that is handed to conform to a medial condyle, a lateral condyle, a right knee, a left knee, and / or the like. The femoral component may be configured based on, among other things, a relationship between a mid-flexion region (131) and an AP dwell (“MFR / AP relationship”). Accordingly, the femoral component may be in the form of a MR component with a spherical or substantially spherical mid-flexion region in a handed configuration. The femoral component may be configured as a unicompartmental femoral implant or a total knee arthroplasty (TKA) femoral implant.

The invention provides a cerebellar ataxia assessment method based on attitude estimation and a multi-layer perceptron, which comprises the following steps: firstly, acquiring a gait video of a subject, detecting a figure target, distributing a target frame, and distributing a unique track identity identification code; calculating the score of the change of the target frame corresponding to each track identity identification code, selecting the subject with the highest score to separate the subjects, and zooming the image sequence of the subjects into a uniform size; secondly, key point detection is conducted on a subject, coordinates of multiple key points of the human body in the whole process are obtained, a sequential sequence of conventional gait features is calculated through the coordinates of the key points, sequential sequence data is subjected to smooth processing, and the average value, the standard deviation, the extreme value and the range of all the features are calculated; calculating a spectrum entropy for the time sequence; calculating Manhattan distances of coordinate sequences of left and right knee joints and left and right ankle joints, generating multi-dimensional feature vectors, and inputting the multi-dimensional feature vectors into an MLP model to output a prediction score; the gait task video is analyzed through the computer, and early risk prediction and severity assessment of cerebellar ataxia are achieved.

A humanoid robot, including an upper portion including a torso, a head, and left and right arm assemblies; a lower portion including left and right lower leg assemblies; a central portion coupled to the upper portion and including a spine, a pelvis, and left and right upper leg assemblies. Each upper leg assembly includes a hip flex actuator coupled to the pelvis and having a hip flex axis positioned at a non-zero downward angle relative to a transverse plane of the humanoid robot. The upper leg assembly also includes left and right knee actuators, wherein each knee actuator is an electric rotary actuator, has a knee axis that is co-planar with the hip flex axis, and has a momentary peak torque greater than a momentary peak torque of the torso twist actuator. The humanoid robot lacks a distinct torso pitch actuator and wherein the hip flex actuators are collectively arranged to provide for pitch movements of the torso.

A non-standing body measurement method and device based on real-time key point detection, the method comprising: (1) collecting non-standing data set images; (2) marking joint key points: left ear, right ear, left shoulder, right shoulder, left elbow, right elbow, left wrist, right wrist, left hip, right hip, left knee, right knee, left ankle, right ankle; marking detection frames: upper arm, forearm, thigh, calf, body trunk, head; marking the longest edge key points of the local cross section: upper arm, forearm, thigh, calf; (3) designing a composite network, merging all key points and detection frames into network training; (4) the composite network supports three methods of video real-time detection, video upload, and picture upload to obtain body data results. The present invention has the characteristics of high precision, low latency, and strong robustness, and can be used for body measurement of people with physical disabilities, assisting healthy people in rapid body data measurement, and intelligent body measurement of clothing merchants, etc., and has good commercial prospects.