Forensic medicine joint motion range detection method and electronic equipment
By using image recognition technology and deep learning algorithms to detect joint mobility in real time, the problem of long detection time and large error in traditional detection methods has been solved, achieving efficient and accurate joint mobility detection and report generation.
Patent Information
- Application Number
- CN202511267151.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-09-05
- Publication Date
- 2025-11-07
AI Technical Summary
Traditional human activity level detection requires multiple participants, is time-consuming, has large errors, and is cumbersome and inefficient, making it difficult to achieve convenient and efficient detection results.
By combining image recognition technology with deep learning algorithms, the system detects joint mobility in real time using a camera, identifies key joint points using the MediaPipe model, calculates joint mobility in accordance with forensic clinical standards, and generates a detection report.
It enables real-time and accurate joint range of motion detection, reduces the subjectivity of inspectors, improves detection efficiency and the reliability of results, and supports cross-platform data management and report generation.
Smart Images
Figure CN120899235A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of artificial intelligence, in particular to a forensic joint activity detection method and an electronic device. BACKGROUND
[0002] Traditional human activity detection requires at least two testers to manually detect the joints of the user to be detected, and the detection result has the problems of long detection time, large tester subjectivity, and large result error. Moreover, one detection needs to go through manual measurement, and the loss value of joint activity is obtained by looking up a table or manual calculation according to the measurement result, which is a relatively cumbersome process and may have human errors, resulting in the problems of inconvenience and low efficiency of manual measurement of joint activity or identification of artificial injury.
[0003] Therefore, there is an urgent need for a convenient detection process, which can detect the joint activity of the user to be detected by the tester entering information and the camera in real time, and directly generate a detection report after the detection is completed. SUMMARY
[0004] The present application aims to overcome the deficiencies of the prior art, such as long detection time, large tester subjectivity, large result error, inconvenience and low efficiency of manual measurement of joint activity or identification of artificial injury, and provides a forensic joint activity detection method and an electronic device.
[0005] In order to solve the above technical problems, the present application provides the following technical solutions: On the one hand, a forensic joint activity detection method is disclosed, comprising the following steps: S1: inputting the information of the user to be detected and selecting the joint to be detected, and judging whether it is unilateral involvement, if yes, executing step S2, otherwise executing step S3; S2: detecting the activity of the healthy side, and judging whether it meets the normal reference value range of forensic clinical identification, if yes, using the loss value table lookup method to detect the affected side, otherwise using the method of equal division to detect the affected side; S3: using the normal reference value of forensic clinical identification as a reference, and using the loss value table lookup method to detect the affected side; S4: acquiring the dimension coordinates of the key points of the joint to be detected in real time; S5: drawing a detection graph and displaying it in real time, using a posture evaluation model to obtain a drawing feature value, and evaluating the posture information of the user to be detected; S6: if the posture evaluation of the user to be detected is passed, acquiring the activity of the joint to be detected in real time until the detection is completed, and obtaining the final joint activity data; S7: using the loss value table lookup method or the direction equal division method to calculate the activity loss value of the joint to be detected; S8: The detection information is archived and backed up.
[0006] As a preferred scheme of the present application, step S4 comprises: using a MediaPipe model to obtain the position of the key point of the joint to be detected in each frame of image in real time, and the key point is in the x and y dimension coordinates of the image.
[0007] As a preferred scheme of the present application, each frame of image obtained by the camera is processed using the mediapipe model, which identifies the x and y dimension coordinate information of the key point of the human joint in the image, and calculates the target joint activity degree according to the dimension coordinate information.
[0008] As a preferred scheme of the present application, the dimension coordinates of the key points of 17 joints of the human body on the image are obtained through the inverse normalization processing. The normalized coordinates output by MediaPipe are (x_norm, y_norm). The formula for obtaining the actual pixel coordinates (x_pixel, y_pixel) of the image through inverse normalization is: x_pixel = x_norm * image_width y_pixel = y_norm * image_height Wherein, x_norm is the normalized x coordinate of the key point of the human body output by mediapipe; y_norm is the normalized y coordinate of the key point of the human body output by mediapipe; image_width is the image width; image_height is the image height.
[0009] As a preferred scheme of the present application, step S5 comprises: using OpenCV library to draw and display the detection image in real time.
[0010] As a preferred scheme of the present application, the loss value lookup table method of step S7 comprises: matching the joint activity degree data of the healthy side and the injured side obtained by detection with the function measurement table of the large joint of the limb to obtain the joint loss value.
[0011] As a preferred scheme of the present application, the direction equal division method of step S7 comprises: substituting the joint activity degree of the healthy side and the injured side obtained by detection into the calculation formula in the evaluation method of forensic clinical examination specification to obtain the joint loss value.
[0012] On the other hand, an electronic device is disclosed, comprising at least one processor, and a memory connected with the at least one processor in communication; the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the forensic joint activity detection method of any one of the above.
[0013] Compared with the prior art, the application has the advantages that: Advanced image recognition technology is adopted, human-computer interaction and data management are combined to realize real-time detection data retention, human body motion images are captured frame by frame through a camera, and deep learning algorithm is used to analyze and process each frame of image and real-time echo processing data, and human body key point information is extracted. On this basis, according to the relevant standards of forensic clinical test specification, the human joint activity is calculated in real time, the original video of detection, the processing video and the detection result are saved to the cloud, data backup is realized, and according to the detection process and the detection result, the detection report can be directly generated, the detection time is shortened, the subjectivity of the detection personnel is avoided, the result error is reduced, the joint activity or artificial injury identification is convenient and efficient, the cross-platform property, the easy maintenance property and the expansibility are ensured, so that the application can be widely applied to the fields of judicial identification, medical rehabilitation, sports training, old-age care and the like. BRIEF DESCRIPTION OF DRAWINGS
[0014] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments with reference made to the accompanying drawings. The drawings are for purposes of illustration only and are not intended to limit the application thereto. The same reference numbers in different drawings identify the same components. In the drawings: Figure 1 A flow chart of a forensic joint activity detection method according to the application embodiment 1; Figure 2 A human joint key point schematic diagram of a forensic joint activity detection method according to the application embodiment 1; Figure 3 A structure block diagram of an electronic device according to the application embodiment 2. DETAILED DESCRIPTION
[0015] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments of the application. The components of the embodiments of the application described and shown in the drawings herein can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the application provided in the drawings is not intended to limit the scope of the claimed application, but only represents selected embodiments of the application. Based on the embodiments of the application, all other embodiments obtained by those skilled in the art without creative work belong to the scope of protection of the application.
[0016] It should be noted that similar reference numerals and letters refer to like items in the accompanying drawings, and once an item is defined in one drawing, it need not be further defined and explained in subsequent drawings. Also, in the description of the present application, the terms "first", "second", and the like are used only to distinguish different entities or operations, and do not imply or suggest relative importance or any such actual relationship or order between these entities or operations. In addition, the terms "connected", "coupled", and the like can mean direct connection between elements or indirect connection via other elements.
[0017] Embodiment 1 A forensic joint range of motion detection method, comprising the following steps: S1: inputting information of a person to be tested and selecting a joint to be tested, judging whether it is unilateral involvement, if yes, executing step S2, otherwise executing step S3; The information of the person to be tested in step S1 includes an image.
[0018] S2: detecting the range of motion of the healthy side, judging whether it meets the normal reference value range of forensic clinical identification, if yes, using the loss value table method to detect the affected side, otherwise using the method of equal division to detect the affected side; S3: using the normal reference value of forensic clinical identification as a reference, and using the loss value table method to detect the affected side; Specifically, S2 and S3 are two detection and evaluation methods, S2 is evaluated according to the evaluation method of 7.11.7.2.3 in the "Standard for Forensic Clinical Examination", and the reference for evaluation is that for a person with injury of one limb, when measuring the range of motion of the injured joint, the healthy side should be measured at the same time, and the normal reference value is used as the reference value to calculate the degree of loss of joint function of the injured side.
[0019] If both joints are injured or the healthy side cannot be used as a reference, the normal reference value range in Appendix A.6 (loss value table method) can be used as a reference according to the individual conditions such as the age of the person to be tested (i.e. using the S3 evaluation method).
[0020] S4: acquiring the dimension coordinates of the key points of the joint to be tested in real time; Specifically, step S4 includes: using MediaPipe to acquire the position of the human joint key points on each frame of image, i.e. the x, y dimension coordinates of the joint key points in the image, and MediaPipe identifies human joint key points as shown in the following table. Figure 2
[0021] Each frame of image obtained by the camera is processed using the mediapipe model, the mediapipe model will identify the human body key points in the image, and obtain the x, y coordinate information of the key points. According to the coordinate information, the target joint activity (for example, when calculating the joint activity of the left elbow joint, the coordinates of the left wrist joint, the left elbow joint and the left shoulder joint are needed Figure 2 15, 13, 11, by mapping the x, y coordinates of the three key points to a 2D plane, calculating the included angle of the 2D vector, and obtaining the joint activity).
[0022] After the inverse normalization processing, the dimension coordinates of the 17 joints of the human body on the image are obtained. The normalized coordinates output by MediaPipe are (x_norm, y_norm). The formula for obtaining the actual pixel coordinates (x_pixel, y_pixel) of the image through inverse normalization is: x_pixel = x_norm * image_width y_pixel = y_norm * image_height Where x_norm is the normalized x coordinate of the human body key point output by mediapipe; y_norm is the normalized y coordinate of the human body key point output by mediapipe; image_width is the image width; image_height is the image height.
[0023] S5: Draw the detection figure and display it in real time, use the posture evaluation model to obtain the drawing feature value, and evaluate the posture information of the person to be tested; Step S5 includes: using OpenCV library to draw the detection image and display it in real time.
[0024] S6: The posture evaluation of the person to be tested is passed, and the activity of the joint to be tested is obtained in real time until the detection is completed, and the final joint activity data is obtained; S7: Calculate the activity loss value of the joint to be tested using the loss value lookup table method or the direction equal division method. Specifically, the loss value lookup table method: match the joint activity obtained by detection with the Appendix B body joint function calculation table in "Limb Motor Function Evaluation", to obtain the joint loss value.
[0025] Direction equal division method: substitute the joint activity of the healthy side and the injured side obtained by detection into the calculation formula in the evaluation method in 7.11.7.2.3 of "Forensic Clinical Test Specification", to obtain the joint loss value.
[0026] S8: Archive and backup the detection information.
[0027] Embodiment 2 This embodiment is a preferred implementation of the forensic joint range of motion detection method described in Embodiment 1. The system adopts a modular design, mainly divided into Java and Python two core modules, each bearing different functional responsibilities, and together realizing the accurate detection and report output of joint range of motion.
[0028] Python side function description: (1) Provide real-time monitoring interface, intuitively show the real-time process of joint activity detection, ensure that the operator can observe and analyze in time.
[0029] (2) Implement video stream processing, including image recognition, feature extraction and motion analysis, to support accurate measurement of joint range of motion.
[0030] (3) Ensure data integrity, the system will automatically save the original video files in the detection process, so as to carry out subsequent data analysis and archive.
[0031] (4) Store and manage the processed video materials, which contain the visual markers and results of joint range of motion.
[0032] Java side function description: (1) Receive and process joint range of motion detection result data transmitted by Python side.
[0033] (2) According to the received data, execute complex algorithm calculation, and get joint range of motion loss value, which is used to quantify the degree of joint function limitation.
[0034] (3) Automatically generate detailed detection report, which covers the quantitative data of joint range of motion, analysis results and possible medical advice.
[0035] Joint range of motion calculation strategy: The system adopts two detection modes to meet the needs of different affected situations: according to whether the joint range of motion of human body meets the normal reference value range of forensic clinical identification, the detection personnel can choose the following detection mode.
[0036] (1) Direct detection mode of table lookup: directly detect the joint range of motion of the measured person, look up the activity degree obtained by detection according to the joint function loss table in "Limb Motor Function Assessment", and get the loss value.
[0037] (2) Compound detection mode: the healthy side and the injured side of the measured person are detected independently, and the activity degree of the healthy side and the injured side obtained by detection is substituted into the calculation formula in 7.11.7.2.3 evaluation method in "Forensic Clinical Test Specification", and the loss value is obtained. The calculation method formula for the limited range of motion of the shoulder, wrist, hip, ankle joint and forearm rotation dysfunction is: the limited range of motion (%) = [(A-A) / A+(B-B) / Bo+(C-C) / Co+....] / n, wherein: A, B,....--the motion range of the corresponding healthy joint of the subject or the normal reference value according to Appendix A.6: A, B, C...--the motion range of the injured joint of the subject; n--the number of directions of the joint motion.
[0038] The calculation method formula for the limited range of motion of the elbow, knee joint is: The limited range of motion (%) = [(A+B)-(A+B)] / (AB), wherein: A--the normal reference value of joint flexion, which should generally be the flexion motion range of the corresponding healthy joint of the subject, or can be selected from the normal reference value range in Appendix A.6 according to the situation: B--the normal reference value of joint hyperextension, which should generally be the hyperextension motion range of the corresponding healthy joint of the subject, or can be selected from the normal reference value range in Appendix A.6 according to the situation: A--the flexion motion range of the injured joint of the subject: B--the hyperextension motion range of the injured joint of the subject. If the joint cannot be hyperextended, B=0.
[0039] The hyperextension activity of the joint can be ignored.
[0040] The specific working principle is that the Python end uses the mediapipe pose and mediapipe hands models to identify human joint key points and human hand joint key points, obtains the key points, calculates the joint activity of each frame of image according to the calculation method in the detection specification, marks the activity of the target joint on each frame of image, and returns the processed image and joint activity to the front end in real time.
[0041] The system uses B / S (Browser / Server) architecture for human-computer interaction and data management. The back-end development framework selects SpringWebFlux to realize efficient and asynchronous data processing capability.
[0042] On the local side, a Python program is used, which is responsible for connecting the detection camera and receiving video streams in real time. Through image recognition technology, the program processes each frame of image in detail to complete the human activity detection task.
[0043] After the detection is completed, the local program uploads the processed detection data to the cloud server, realizing centralized storage and management of data, facilitating subsequent analysis, query and sharing.
[0044] Throughout the process, the B / S architecture ensures the flexibility and scalability of the system, the SpringWebFlux backend framework ensures the efficiency of data processing, and the local program written in Python fully utilizes image recognition technology to achieve accurate detection of human activity. Cloud uploading of data further improves data management efficiency and security.
[0045] Embodiment 3 As shown in the embodiment, an electronic device includes at least one processor, and a memory connected with the at least one processor in communication; the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the forensic joint activity detection method described in the foregoing embodiments. The input and output interface can include a display, a keyboard, a mouse, and a USB interface for inputting and outputting data; and the power supply provides power for the electronic device.
[0046] Those skilled in the art can understand that all or part of the steps of the above-mentioned method embodiments can be completed by program instruction related hardware, and the foregoing program can be stored in a computer readable storage medium, and the program is executed to execute the steps of the above-mentioned method embodiments; and the foregoing storage medium includes mobile storage equipment, read only memory (Read Only Memory, ROM), magnetic disc or optical disc and various storage program codes.
[0047] When the integrated unit of the present application is realized in the form of a software function unit and sold or used as an independent product, it can also be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the embodiments of the present application can be embodied in the form of a software product, which is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in the embodiments of the present application. The foregoing storage medium includes mobile storage equipment, ROM, magnetic disc or optical disc and various storage program codes.
[0048] In addition, it should be noted that various specific technical features described in the foregoing specific embodiments can be combined in any appropriate manner without contradiction, and to avoid unnecessary repetition, the present disclosure will not further describe various possible combinations.
[0049] Furthermore, the various embodiments of the present disclosure can be combined with each other as long as they do not violate the idea of the present disclosure, and they should be considered as disclosed in the present disclosure.
Claims
1. A forensic method for detecting joint mobility, characterized by, The method comprises the following steps: S1: the information of the person to be tested is inputted and the joint to be tested is selected, and it is judged whether it is unilateral involvement, if yes, step S2 is executed, and if not, step S3 is executed; S2: the range of motion of the healthy side is detected, and it is judged whether it conforms to the normal reference value range of forensic clinical identification, if yes, the loss value table lookup method is used for detection of the affected side, and if not, the method of equal division is used for detection of the affected side; S3: the normal reference value of forensic clinical identification is used as a reference basis, and the loss value table lookup method is used for detection of the affected side; S4: the dimension coordinates of the key points of the joint to be tested are obtained in real time; S5: a detection graph is drawn and real-time display is performed, a feature value of the drawing is obtained by using a posture evaluation model, and the posture information of the person to be tested is evaluated; S6: if the posture evaluation of the person to be tested is passed, the range of motion of the joint to be tested is obtained in real time until the detection is completed, and the final joint range of motion data is obtained; S7: the loss value of the range of motion of the joint to be tested is calculated by using the loss value table lookup method or the direction equal division method; S8: the detection information is archived and backed up.
2. The forensic joint range of motion detection method of claim 1, wherein, Step S4 comprises: using a MediaPipe model to obtain the positions of the key points of the joint to be tested in each frame of image in real time, and the key points are in x and y dimension coordinates of the image.
3. The forensic joint range of motion detection method of claim 2, wherein, Each frame of image obtained by the camera is processed by using the mediapipe model, the mediapipe model identifies the x and y dimension coordinate information of the key points of the human joints in the image, and the target joint range of motion is calculated according to the dimension coordinate information.
4. The forensic joint range of motion detection method of claim 3, wherein, Through the inverse normalization processing, the dimension coordinates of the key points of 17 joints of the human body on the image are obtained, the normalized coordinates (x_norm, y_norm) output by MediaPipe are inverse normalized to obtain the actual pixel coordinates (x_pixel, y_pixel) of the image, and the formula is: x_pixel = x_norm * image_width y_pixel = y_norm * image_height Wherein, x_norm is the normalized x coordinate of the key point of the human body output by mediapipe; y_norm is the normalized y coordinate of the key point of the human body output by mediapipe; image_width is the image width; image_height is the image height.
5. The forensic joint range of motion detection method of claim 1, wherein, Step S5 comprises: using an OpenCV library to draw a detection image and perform real-time display.
6. The forensic joint range of motion detection method of claim 1, wherein, The loss value table lookup method in step S7 comprises: matching the joint range of motion data of the healthy side and the injured side obtained by detection with a function table of major joints, to obtain the joint loss value.
7. The forensic joint range of motion detection method of claim 1, wherein, The direction equal division method in step S7 comprises: substituting the joint range of motion of the healthy side and the injured side obtained by detection into a calculation formula in the evaluation method of the forensic clinical test specification, to obtain the joint loss value.
8. An electronic device, comprising: The method comprises at least one processor and a memory connected with the at least one processor in communication; the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the forensic joint activity detection method in any one of claims 1 to 7.