Self-service bone age detection method and device, computer equipment and storage medium
Through the self-service bone age detection method, the preset sensing area and real-time bone age analysis engine are used to achieve efficient, accurate and convenient bone age detection on campus, solving the problems of low detection efficiency and low registration efficiency in the existing technology, and improving the detection Real-time and accuracy.
Patent Information
- Application Number
- CN202410952325.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-16
- Publication Date
- 2024-11-08
AI Technical Summary
The existing campus bone age detection methods are inefficient, resulting in a long testing time period and low registration efficiency, which affects the teaching order and testing scale.
Using self-service bone age detection method, the user is sensed through preset sensing areas, identity proofreading and image acquisition are performed, bone age detection images are generated using image optimization technology, and real-time detection is carried out through real-time bone age analysis engine to achieve contactless identity verification and accurate bone age judgment .
It improves the convenience and accuracy of bone age detection, shortens detection time, enhances the real-time and efficiency of the system, and reduces manual intervention and misdetection.
Smart Images

Figure CN118902484A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of bone age detection, and in particular to a self-service bone age detection method, device, computer equipment and storage medium. Background Art
[0002] At present, with the development of science and technology, more and more smart devices are entering campuses to provide students with more convenient services. Among them, the self-service detection campus bone age meter, as a new type of smart device, has received widespread attention. With the increasing attention to the growth and development of adolescents, bone age detection has become a vital health check-up item. Bone age detection can not only reflect the growth and development of adolescents, but also provide a basis for scientific assessment of their health status.
[0003] Currently, bone age testing for teenagers on campus is done by schools cooperating with relevant institutions, organizing students to take unified bone images, and then bringing the images back and analyzing them uniformly after the shooting. Finally, the results are transmitted back to the school. However, the testing institutions have a large number of testing subjects. When dealing with a specific group, the corresponding testing efficiency is not improved due to the reduction in the range of bone age data types, resulting in a longer time period for bone age testing. When faced with a large number of students to be tested, these methods not only affect the teaching order and limit the scale and frequency of bone age testing, but also lead to low registration efficiency when manually registering a large number of students.
[0004] The above-mentioned existing technical solutions have the following defects: the poor convenience of bone age detection leads to a long bone age detection process and low efficiency, so there is room for improvement. Summary of the invention
[0005] In order to improve the convenience of bone age detection, the present application provides a self-service bone age detection method, device, computer equipment and storage medium.
[0006] The above-mentioned invention objective of the present application is achieved through the following technical solutions: A self-service bone age detection method, the self-service bone age detection method comprising: Sensing the current user through a number of preset sensing areas, obtaining a person identity verification result, and generating a bone image shooting instruction when the user verification result matches the person identity; According to the image shooting instruction, an initial bone image is acquired, and image optimization is performed on the initial bone image to generate a bone age detection image; The age data of the person is obtained from the result of the person identity verification. When the age data of the person is within the age threshold preset by the system, a real-time bone age analysis engine is used to perform real-time detection on a number of the bone age detection images to obtain a bone age detection result.
[0007] By adopting the above technical scheme, since the bone age detection of teenagers on campus is carried out after the school cooperates with relevant institutions, the students are organized to take unified bone images, and after the shooting, the relevant institutions bring back the images and analyze them uniformly, and finally transmit the results back to the school. However, the detection agency has a large number of detection object groups. When dealing with a specific group, the corresponding detection efficiency will not be improved due to the reduction in the range of bone age data types, resulting in a longer time period for bone age detection. Therefore, the present application senses the current user through several preset sensing areas, obtains the personnel identity proofreading results, can perform corresponding identity detection on multiple users in the sensing area, thereby performing contactless identity authentication, and improving the convenience of access control; when the user proofreading result matches the personnel identity, a bone image shooting instruction is generated, the user's identity information can be automatically verified, and the bone initialization can be performed after the verification is successful. By collecting images, it can reduce the situations of proxy inspection and false inspection, and improve the accuracy of detection; by optimizing the initial bone image and generating a bone age detection image, the bone structure in the image can be made clearer and easier to distinguish through image optimization, thereby improving the efficiency of the model in image analysis, and at the same time, improving the accuracy of the system in judging the bone age of the person; by obtaining the age data of the person from the identity verification result of the person, when the age data of the person is within the age threshold preset by the system, a real-time bone age analysis engine is used to perform real-time detection on several bone age detection images to obtain the bone age detection result, and the age of the person undergoing bone age detection can be screened first. When the age meets the requirements of the system, the initial bone image will be analyzed more accurately and quickly according to the real-time bone age analysis engine built by the system, thereby further improving the efficiency of bone age judgment and improving the real-time performance of the system in bone age detection.
[0008] In a preferred example, the present application can be further configured as follows: the current user is sensed through a number of preset sensing areas to obtain a person identity verification result, and when the user verification result matches the person identity, a bone image shooting instruction is generated, specifically including: When the sensing area senses the presence of a user, a facial image of the person is acquired, and at the same time, facial image features are extracted according to computer vision technology; The person identity information is queried according to the facial image features, and when the person identity information is obtained, a person identity matching message is generated.
[0009] By adopting the above technical solution, when the sensing area senses the presence of a user, a facial image of the person is obtained. At the same time, facial image features are extracted based on computer vision technology, and facial features of the person in the inspected area can be extracted based on facial recognition technology, so as to facilitate subsequent accurate and rapid identification of the person's identity; by querying the person's identity information based on facial image features, when the person's identity information is obtained, a person's identity matching message is generated, and a comparison and search can be performed from the person's identity information database based on the person's facial information, and identity matching can be performed when the corresponding person's identity information can be found, thereby reducing false detection, proxy inspection, etc., and reducing potential errors in manual identification and comparison, thereby improving data accuracy.
[0010] In a preferred example, the present application may be further configured as follows: acquiring the initial bone image according to the image shooting instruction specifically includes: Acquire a current action image according to the personnel action prompt information, compare the current action image with the personnel action prompt information, and generate a comparison result; When the personnel action image is different from the personnel action prompt information, a personnel action abnormality message is generated, and at the same time, a personnel re-prompt instruction is generated; When the personnel action image is the same as the action in the personnel action prompt information, the skeleton initial image is obtained.
[0011] By adopting the above technical scheme, by obtaining a personnel action image according to the personnel action prompt information, comparing the personnel action image with the personnel action prompt information action, and generating a comparison result, when the personnel action image is different from the personnel action prompt information action, a personnel action abnormality message can be generated to remind the person under test that the current action does not meet the system requirements and that appropriate measures need to be taken in time. At the same time, a personnel re-prompt instruction is generated to remind the person under test to modify the current action again, thereby improving the accuracy of the captured image; when the personnel action image is the same as the personnel action prompt information action, an initial bone image is obtained, and an accurate bone image can be obtained when the person under test's action meets the system requirements, thereby improving the accuracy of data collection.
[0012] In a preferred example, the present application may be further configured as follows: the image optimization of the initial bone image to generate a bone age detection image specifically includes: Converting the initial bone image into a grayscale original image, and adjusting the grayscale original image using a local contrast enhancement method to obtain a first adjusted image; The first adjusted image is denoised to obtain a second adjusted image, and the second adjusted image is processed according to an edge enhancement method to obtain the bone age detection image.
[0013] By adopting the above technical scheme, the initial bone image is converted into a grayscale original image, and the grayscale original image is adjusted by a local contrast enhancement method to obtain a first adjusted image, thereby highlighting the details and structures in the image and improving the contrast of the bone part; the first adjusted image is denoised to obtain a second adjusted image, which can eliminate image noise introduced by the shooting equipment or other factors and improve the clarity and quality of the image; the second adjusted image is processed according to the edge enhancement method to obtain a bone age detection image, thereby more clearly displaying the edges of the bone structure and making the bone age detection more accurate.
[0014] In a preferred example, the present application can be further configured as follows: the construction of the real-time bone age analysis engine specifically includes: The real-time bone age analysis engine is structured according to a high-concurrency architecture, and at the same time, a message queue is used to manage the bone age detection image input into the real-time bone age analysis engine; The bone age detection image in the message queue is analyzed in real time through a self-trained bone age detection model, and the bone age detection result is output after the analysis is completed.
[0015] By adopting the above technical solution and structuring the real-time bone age analysis engine according to a high-concurrency architecture, the real-time and concurrency of the system in processing input data can be improved. The bone age detection images input into the real-time bone age analysis engine are managed by a message queue, and the asynchronous stream of image data can be further processed, thereby enhancing the scalability and stability of the system. The bone age detection images in the message queue are analyzed in real time through a self-trained bone age detection model, and the bone age detection results can be obtained from the bone age detection model after the analysis is completed, thereby using the bone age detection model to accurately analyze the initial bone image to obtain more accurate detection results, reduce manual intervention, and improve detection efficiency.
[0016] In a preferred example, the present application can be further configured as follows: the construction of the bone age detection model specifically includes: Acquire historical bone age detection images, and filter the historical bone age detection images according to an age threshold set by the system to obtain a subject bone data set; Extracting features from the object skeleton data set according to a convolutional neural network to obtain an object skeleton feature set; A model training data set is generated according to the object bone data set and the object bone feature set, and a deep learning architecture is iteratively trained according to the model training data set to obtain the bone age detection model.
[0017] By adopting the above technical scheme, by acquiring historical bone age detection images and screening the historical bone age detection images according to the age threshold set by the system, an object bone data set is obtained, and corresponding screening conditions can be selected according to different personnel needs, thereby reducing the dimension of training data and improving the efficiency of subsequent model training; according to the convolutional neural network, the data in the object bone data set is feature extracted to obtain the object bone feature set, and the feature data in the image can be extracted, thereby amplifying the feature points of the bone age detection image to highlight the feature expression of the bones and improve the model's image processing speed; a model training data set is generated according to the object bone data set and the object bone feature set, and the deep learning architecture is iteratively trained according to the model training data set, so as to obtain a bone age detection model, thereby improving the accuracy of the system's analysis of bone age detection images.
[0018] The second object of the invention is achieved by the following technical solutions: A self-service bone age detection device, the self-service bone age detection device comprising: An image acquisition module is used to sense the current user through a number of preset sensing areas, obtain a person identity verification result, and generate a bone image shooting instruction when the user verification result matches the person identity; An image processing module, used to obtain an initial bone image according to the bone image shooting instruction, and perform image optimization on the initial bone image to generate a bone age detection image; The image analysis module is used to obtain the age data of the personnel from the identity verification results of the personnel. When the age data of the personnel is within the age threshold preset by the system, the real-time bone age analysis engine is used to perform real-time detection on several bone age detection images to obtain the bone age detection results.
[0019] By adopting the above-mentioned technical scheme, since the bone age detection of teenagers on campus is carried out after the school cooperates with relevant institutions, the students are organized to take unified bone images, and after the shooting, the images are brought back by the relevant institutions for unified analysis, and finally the results are transmitted back to the school. However, the detection agency has a large group of detection objects. When dealing with a specific group, the corresponding detection efficiency will not be improved due to the reduction in the range of bone age data types, resulting in a longer time period for bone age detection. Therefore, the present application senses the current user through several preset sensing areas, obtains the personnel identity proofreading results, and can perform corresponding identity detection on multiple users in the sensing area, thereby performing contactless identity verification, thereby improving the convenience of access control; by generating a bone image shooting instruction when the user proofreading result matches the personnel identity, the user's identity information can be automatically verified The system can obtain the personnel's age data from the personnel's identity verification results, and collect the initial bone image after successful verification, thereby reducing the situations of proxy inspection and false detection, and improving the accuracy of detection; by optimizing the initial bone image and generating a bone age detection image, the bone structure in the image can be made clearer and easier to distinguish through image optimization, thereby improving the efficiency of the model's analysis of the image, and at the same time, improving the accuracy of the system's judgment of the personnel's bone age; by obtaining the personnel's age data from the personnel's identity verification results, when the personnel's age data is within the age threshold preset by the system, a number of bone age detection images are detected in real time through the real-time bone age analysis engine that has been built to obtain the bone age detection results, and the real-time bone age analysis engine that has been built by the system can be used to analyze the initial bone image more accurately and quickly, thereby further improving the efficiency of bone age judgment and improving the real-time performance of the system for bone age detection.
[0020] The third objective of the present application is achieved through the following technical solutions: A computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned self-service bone age detection method when executing the computer program.
[0021] The fourth objective of the present application is achieved through the following technical solutions: A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the above-mentioned self-service bone age detection method.
[0022] In summary, the present application includes at least one of the following beneficial technical effects: 1. By converting the initial bone image into a grayscale original image, and adjusting the grayscale original image by a local contrast enhancement method, a first adjusted image is obtained, thereby highlighting the details and structures in the image and improving the contrast of the bone part; the first adjusted image is subjected to denoising to obtain a second adjusted image, which can eliminate image noise introduced by the shooting equipment or other factors, improve the clarity and quality of the image, and the second adjusted image is processed according to an edge enhancement method to obtain a bone age detection image, thereby more clearly showing the edge of the bone structure and making the bone age detection more accurate; 2. By constructing the real-time bone age analysis engine according to the high-concurrency architecture, the real-time and concurrency of the system in processing input data can be improved. The message queue is used to manage the bone age detection images input into the real-time bone age analysis engine, which can further process the asynchronous stream of image data and enhance the scalability and stability of the system. The self-trained bone age detection model is used to analyze the bone age detection images in the message queue in real time. The bone age detection results can be obtained from the bone age detection model after the analysis is completed, so that the bone age detection model can be used to accurately analyze the initial bone image to obtain more accurate detection results, reduce manual intervention, and improve detection efficiency; 3. By acquiring historical bone age detection images and screening the historical bone age detection images according to the age threshold set by the system, an object bone data set is obtained. The corresponding screening conditions can be selected according to different personnel needs, thereby reducing the dimension of the training data and improving the efficiency of subsequent model training; according to the convolutional neural network, the data in the object bone data set is feature extracted to obtain the object bone feature set, which can extract the feature data in the image, thereby amplifying the feature points of the bone age detection image to highlight the feature expression of the bones and improve the model's image processing speed; a model training data set is generated according to the object bone data set and the object bone feature set, and the deep learning architecture is iteratively trained according to the model training data set, so as to obtain a bone age detection model, thereby improving the accuracy of the system's analysis of bone age detection images. BRIEF DESCRIPTION OF THE DRAWINGS
[0023] Figure 1 is a flow chart of a self-service bone age detection method in one embodiment of the present application; Figure 2 This is a flowchart for implementing step S10 in the self-service bone age detection method in one embodiment of the present application; Figure 3 is a flowchart for implementing step S20 in the self-service bone age detection method in one embodiment of the present application; Figure 4 is another implementation flow chart of step S20 in the self-service bone age detection method in one embodiment of the present application; Figure 5It is a flowchart for implementing the construction of a real-time bone age analysis engine in a self-service bone age detection method in one embodiment of the present application; Figure 6 It is a flowchart of the process of building a bone age detection model in a self-service bone age detection method in one embodiment of the present application; Figure 7 This is a principle block diagram of a self-service bone age detection device in one embodiment of the present application; Figure 8 It is a schematic diagram of a device in an embodiment of the present application. DETAILED DESCRIPTION
[0024] The present application is further described in detail below in conjunction with the accompanying drawings.
[0025] In one embodiment, if Figure 1 As shown, the present application discloses a self-service bone age detection method, which specifically includes the following steps: S10: The current user is sensed through a number of preset sensing areas to obtain a person identity verification result, and when the user verification result matches the person identity, a bone image shooting instruction is generated.
[0026] In this embodiment, the person identity verification result refers to the result obtained after the person identity is verified. The skeleton shooting instruction refers to the instruction for controlling the device to shoot the skeleton of the person.
[0027] Specifically, several preset sensing areas are set around the device, such as four sensing areas are set in the directions corresponding to the four vertical surfaces of the device. When a person stands in the preset sensing area, the system will use the installed image sensor to identify and proofread the person in the sensing area, and trigger the bone image shooting instruction when the person identity proofreading result is a match for the person identity.
[0028] S20: According to the bone image shooting instruction, an initial bone image is obtained, and the initial bone image is optimized to generate a bone age detection image.
[0029] In this embodiment, the initial skeleton image refers to an image obtained after photographing the skeleton, and the bone age detection image refers to an image used in bone age detection.
[0030] Specifically, after the initial bone image is acquired, the initial bone image is processed by image optimization methods such as contrast enhancement, denoising, edge enhancement or filtering, and the processed image is used as a bone age detection image.
[0031] S30: Obtain the age data of the person from the result of the person identity verification. When the age data of the person is within the age threshold preset by the system, a real-time bone age analysis engine is used to perform real-time detection on several bone age detection images to obtain a bone age detection result.
[0032] In this embodiment, the real-time bone age analysis engine refers to an engine for analyzing the skeleton of a person and obtaining the corresponding bone age. The bone age detection result refers to the result obtained after the bone age detection.
[0033] Specifically, after obtaining the bone detection image, the system will first screen the age of the person. When it meets the age threshold set by this system, it means that the real-time bone age analysis engine installed in this system can monitor the bone age detection image more accurately. At this time, the bone age detection image is analyzed by the real-time bone age analysis engine, and the corresponding bone age detection result is obtained after the analysis is completed.
[0034] In this embodiment, since the bone age detection of teenagers on campus is carried out after the school cooperates with relevant institutions, the students are organized to take unified bone images, and after the shooting, the relevant institutions bring back the images and analyze them uniformly, and finally transmit the results back to the school. However, the detection agency has a large number of detection object groups. When dealing with a specific group, the corresponding detection efficiency will not be improved due to the reduction in the range of bone age data types, resulting in a longer time period for bone age detection. Therefore, this application senses the current user through several preset sensing areas, obtains the personnel identity proofreading results, and can perform corresponding identity detection on multiple users in the sensing area, thereby performing contactless identity verification, which improves the convenience of access control; when the user proofreading result matches the personnel identity, a bone image shooting instruction is generated, and the user's identity information can be automatically verified. After successful verification, the initial bone image is collected to reduce the situations of proxy inspection and false detection, and improve the accuracy of detection; by optimizing the initial bone image to generate a bone age detection image, the bone structure in the image can be made clearer and easier to distinguish through image optimization, thereby improving the efficiency of the model in image analysis, and at the same time, improving the accuracy of the system in judging the bone age of the person; by obtaining the person's age data from the person's identity proofreading result, when the person's age data is within the age threshold preset by the system, a real-time bone age analysis engine is used to perform real-time detection of several bone age detection images to obtain the bone age detection result, and the real-time bone age analysis engine built by the system can be used to perform a more accurate and rapid analysis of the initial bone image, thereby further improving the efficiency of bone age judgment and improving the real-time performance of the system in bone age detection.
[0035] In one embodiment, if Figure 2 As shown, in step S10, the current user is sensed through a number of preset sensing areas to obtain a user identity verification result. When the user verification result is a user identity match, a bone image shooting instruction is generated, which specifically includes: S11: When the sensing area senses the presence of a user, a facial image of the person is acquired, and at the same time, facial image features are extracted based on computer vision technology.
[0036] Specifically, the system will use facial recognition technology to capture facial information within the sensing area through cameras or other image sensors, and use computer vision technology to extract facial features of people.
[0037] S12: querying the person's identity information according to the facial image features, and generating a user identity matching message when the person's identity information is obtained.
[0038] Specifically, the identity information of the person is queried based on the facial image features. When the identity information of the person can be obtained from the personnel information database, it means that the person is the inspected person of the institution to which he belongs, and an identity matching message is generated at this time.
[0039] In one embodiment, if Figure 3 As shown, in step S20, namely, according to the bone image shooting instruction, the initial bone image is obtained, which specifically includes: S21: acquiring a current action image according to the personnel action prompt information, comparing the current action image with the personnel action prompt information, and generating a comparison result.
[0040] In this embodiment, the personnel action prompt information refers to information displayed on the device display for prompting and limiting the personnel's actions. The personnel action image refers to the current action state image of the personnel.
[0041] Specifically, a personnel action image is obtained according to the personnel action prompt information displayed on the display end, and the personnel action image is compared with the established action in the personnel action prompt information to generate a corresponding comparison result.
[0042] S22: When the personnel action image is different from the personnel action prompt information, a personnel action abnormality message is generated, and at the same time, a personnel re-prompt instruction is generated.
[0043] In this embodiment, the personnel action abnormality message refers to a message triggered when the personnel action is incorrect.
[0044] Specifically, when the comparison result is that the personnel action image is different from the personnel action prompt information, it means that the person being tested did not operate according to the specified prompt information. At this time, a personnel action abnormality message is generated to prompt the person being tested to make corresponding adjustments according to the actions in the personnel action prompt information.
[0045] S23: When the person action image is the same as the person action prompt information action, an initial skeleton image is obtained.
[0046] Specifically, when the comparison result is that the personnel action image and the personnel action prompt information action are the same, at this time the personnel action is the same as the specified action. Through X-ray technology, combined with technologies such as ultrasound and MRI, the specified positions of the person being tested, such as the palm or wrist, are photographed to obtain the initial bone image. After the acquisition is completed, a person leaving message is generated to remind the person being tested to leave in time, so as to improve the event cycle of the test and improve the test efficiency.
[0047] In one embodiment, if Figure 4 As shown, in step S30, the initial bone image is optimized to generate a bone age detection image, which specifically includes: S31: Convert the initial bone image into a grayscale original image, and adjust the grayscale original image using a local contrast enhancement method to obtain a first adjusted image.
[0048] In this embodiment, the grayscale original image refers to an image obtained by graying the bone image. The first adjusted image refers to an image obtained by performing local contrast enhancement on the image.
[0049] Specifically, the initial image of the skeleton is firstly grayed to obtain a gray original image, and local contrast enhancement is performed on the basis of the gray original image to highlight the characteristics of the human skeleton, and after the adjustment is completed, a first adjusted image is obtained.
[0050] S32: performing denoising processing on the first adjusted image to obtain a second adjusted image, and processing the second adjusted image according to an edge enhancement method to obtain a bone age detection image.
[0051] In this embodiment, the second adjusted image refers to an image obtained by denoising the first adjusted image.
[0052] Specifically, after obtaining the first adjusted image, the first adjusted image is denoised according to a denoising model preset in the system, such as a Gaussian denoising model, to obtain a second adjusted image, and the second adjusted image is edge enhanced using edge enhancement methods such as the Sobel operator and Canny edge detection to obtain a bone age detection image.
[0053] In one embodiment, if Figure 5 As shown, in step S40, the construction of the real-time bone age analysis engine specifically includes: S41: The real-time bone age analysis engine is structured according to a high-concurrency architecture. At the same time, a message queue is used to manage the bone age detection images input into the real-time bone age analysis engine.
[0054] Specifically, high-concurrency frameworks such as microservice architecture, server-less architecture or distributed system technologies are used to build the infrastructure of the real-time bone age analysis engine, and message queues are introduced to manage the image data input into the engine. When several image data enter the engine, concurrent tasks can be allocated to multiple threads.
[0055] S42: The bone age detection images in the message queue are analyzed in real time through the self-trained bone age detection model, and the bone age detection results are output after the analysis is completed.
[0056] Specifically, the initial bone image is analyzed by a bone age detection model, and the initial bone image is analyzed by methods such as image segmentation or key point detection to obtain a bone age detection result.
[0057] In one embodiment, if Figure 6 As shown in the figure, the construction of the bone age detection model includes: S401: Acquire historical bone age detection images, and filter the bone age detection images according to the age threshold set by the system to obtain the object bone data set.
[0058] In this embodiment, the historical bone age detection image refers to the bone age detection image data input during the historical test. The object bone data set refers to the set of bone age detection images after screening.
[0059] Specifically, the bone age detection images imported by the staff in the system are obtained as historical bone age detection images, and the historical bone age detection images are screened according to the age threshold set by the system. For example, when targeting campus students, the age threshold is set to the age range of teenagers, and the object bone data set is obtained.
[0060] S402: Extract features from the object skeleton data set according to a convolutional neural network to obtain an object skeleton feature set.
[0061] In this embodiment, the object bone feature set refers to a set of bone feature data corresponding to the historical bone age detection image.
[0062] Specifically, convolutional neural network technology is used to extract local features of image data in the object skeleton dataset according to the convolution layer, and the extracted local features are downsampled according to the pooling layer to reduce the size of the feature map to reduce the amount of calculation. The extracted features are then integrated in the fully connected layer, and different features are classified to improve the subsequent comparison and analysis of the feature areas.
[0063] S403: Generate a model training data set according to the object bone data set and the object bone feature set, and iteratively train the deep learning architecture according to the model training data set to obtain a bone age detection model.
[0064] Specifically, the object bone data set and the object bone feature set are matched one by one to obtain a model training data set, and the deep learning architecture is iteratively trained according to the model training data set to continuously train the deep learning architecture to learn and understand the bone age detection image and its corresponding bone age. When the error value between the input bone age detection image and the output bone age does not exceed 3%, the iteration is stopped and the bone age detection model is obtained.
[0065] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.
[0066] In one embodiment, a self-service bone age detection device is provided, which corresponds one-to-one to the self-service bone age detection method in the above embodiment. Figure 7 As shown, the self-service bone age detection device includes an image acquisition module, an image processing module, and an image analysis module. The detailed description of each functional module is as follows: An image acquisition module is used to sense the current user through a number of preset sensing areas, obtain the person's identity verification result, and generate a bone image shooting instruction when the user's verification result matches the person's identity; An image processing module is used to obtain an initial bone image according to an image shooting instruction, and optimize the initial bone image to generate a bone age detection image; The image analysis module is used to perform real-time inference on a number of bone age detection images by deploying a self-trained bone age detection model to obtain bone age detection results. At the same time, it generates a bone age analysis end message and sends the bone age detection results to the storage terminal.
[0067] Optionally, the image acquisition module specifically includes: The identity comparison submodule is used to obtain the facial image of the person when the sensing area senses the presence of the user, and at the same time, extract the facial image features based on computer vision technology; The identity matching submodule is used to obtain the identity information of the person according to the facial image features, and generate a user identity matching message when the identity information of the person is obtained.
[0068] Optional image processing modules include: The personnel action prompt submodule is used to obtain the current action image according to the personnel action prompt information, compare the current action image with the personnel action prompt information, and generate a comparison result; The abnormal action submodule is used to generate an abnormal action message when the action image of the personnel is different from the action prompt information of the personnel, and at the same time, generate a personnel re-prompt instruction; The image acquisition submodule is used to acquire the initial skeleton image when the personnel action image is the same as the personnel action prompt information action.
[0069] Optional image processing modules include: The image primary enhancement submodule is used to convert the initial bone image into a grayscale original image, and adjust the grayscale original image using a local contrast enhancement method to obtain a first adjusted image; The image secondary enhancement submodule is used to perform denoising on the first adjusted image to obtain a second adjusted image, and process the second adjusted image according to an edge enhancement method to obtain a bone age detection image.
[0070] Optional, construction of a real-time bone age analysis engine, including: The architecture building submodule is used to build the structure of the real-time bone age analysis engine according to the high-concurrency architecture. At the same time, the message queue is used to manage the bone age detection images input into the real-time bone age analysis engine. The model building submodule is used to perform real-time analysis of the bone age detection images in the message queue through a self-trained bone age detection model, and output the bone age detection results after the analysis is completed.
[0071] Optional, construction of a bone age detection model, specifically including: The image data sorting submodule is used to obtain historical personnel skeleton images and filter the personnel skeleton images according to the age threshold set by the system to obtain the object skeleton data set; The feature data sorting submodule is used to extract features from the object skeleton data set according to the convolutional neural network to obtain the object skeleton feature set; The model training submodule is used to generate a model training data set based on the object bone data set and the object bone feature set, and iteratively train the deep learning architecture based on the model training data set to obtain a bone age detection model.
[0072] For the specific limitations of the self-service bone age detection device, please refer to the limitations of the self-service bone age detection method above, which will not be repeated here. The various modules in the above-mentioned self-service bone age detection device can be implemented in whole or in part through software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.
[0073] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 8As shown. The computer device includes a processor, a memory, a network interface and a database connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store historical initial bone images. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a self-service bone age detection method is implemented.
[0074] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program: The current user is sensed through a number of preset sensing areas to obtain the person's identity verification result, and when the user's verification result matches the person's identity, a bone image shooting instruction is generated; According to the bone image shooting instruction, the initial bone image is obtained, and the initial bone image is optimized to generate a bone age detection image; The person's age data is obtained from the person's identity verification results. When the person's age data is within the age threshold preset by the system, a real-time bone age analysis engine is used to perform real-time detection on several bone age detection images to obtain the bone age detection results.
[0075] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented: The current user is sensed through a number of preset sensing areas to obtain the person's identity verification result, and when the user's verification result matches the person's identity, a bone image shooting instruction is generated; According to the bone image shooting instruction, the initial bone image is obtained, and the initial bone image is optimized to generate a bone age detection image; The person's age data is obtained from the person's identity verification results. When the person's age data is within the age threshold preset by the system, a real-time bone age analysis engine is used to perform real-time detection on several bone age detection images to obtain the bone age detection results.
[0076] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.
[0077] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.
[0078] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.
Claims
1. A self-service bone age detection method, It is characterized in that The self-service bone age detection method comprises: Sensing the current user through a number of preset sensing areas, obtaining a person identity verification result, and generating a bone image shooting instruction when the user verification result matches the person identity; According to the bone image shooting instruction, an initial bone image is acquired, and image optimization is performed on the initial bone image to generate a bone age detection image; The age data of the person is obtained from the result of the person identity verification. When the age data of the person is within the age threshold preset by the system, a real-time bone age analysis engine is used to perform real-time detection on a number of the bone age detection images to obtain a bone age detection result.
2. The self-service bone age detection method according to claim 1, It is characterized in that The method of sensing the current user through a plurality of preset sensing areas to obtain a personnel identity verification result, and generating a bone image shooting instruction when the user verification result matches the personnel identity, specifically includes: When the sensing area senses the presence of a user, a facial image of the person is acquired, and at the same time, facial image features are extracted according to computer vision technology; The person identity information is queried according to the facial image features, and when the person identity information is obtained, a person identity matching message is generated.
3. The self-service bone age detection method according to claim 1, It is characterized in that The step of acquiring the initial bone image according to the bone image shooting instruction specifically includes: Acquire a current action image according to the personnel action prompt information, compare the current action image with the personnel action prompt information, and generate a comparison result; When the personnel action image is different from the personnel action prompt information, a personnel action abnormality message is generated, and at the same time, a personnel re-prompt instruction is generated; When the personnel action image is the same as the action in the personnel action prompt information, the skeleton initial image is obtained.
4. The self-service bone age detection method according to claim 3, It is characterized in that The image optimization of the initial bone image to generate a bone age detection image specifically includes: Converting the initial bone image into a grayscale original image, and adjusting the grayscale original image using a local contrast enhancement method to obtain a first adjusted image; The first adjusted image is denoised to obtain a second adjusted image, and the second adjusted image is processed according to an edge enhancement method to obtain the bone age detection image.
5. The self-service bone age detection method according to claim 1, It is characterized in that The construction of the real-time bone age analysis engine specifically includes: The real-time bone age analysis engine is structured according to a high-concurrency architecture, and at the same time, a message queue is used to manage the bone age detection image input into the real-time bone age analysis engine; The bone age detection image in the message queue is analyzed in real time through a self-trained bone age detection model, and the bone age detection result is output after the analysis is completed.
6. The self-service bone age detection method according to claim 5, It is characterized in that The construction of the bone age detection model specifically includes: Acquire historical bone age detection images, and filter the bone age detection images according to an age threshold set by the system to obtain a subject bone data set; Extracting features from the object skeleton data set according to a convolutional neural network to obtain an object skeleton feature set; A model training data set is generated according to the object bone data set and the object bone feature set, and a deep learning architecture is iteratively trained according to the model training data set to obtain the bone age detection model.
7. A self-service bone age detection device, It is characterized in that The self-service bone age detection device comprises: An image acquisition module is used to sense the current user through a number of preset sensing areas, obtain a person identity verification result, and generate a bone image shooting instruction when the user verification result matches the person identity; An image processing module, used to obtain an initial bone image according to the image shooting instruction, and perform image optimization on the initial bone image to generate a bone age detection image; The image analysis module is used to perform real-time reasoning on a number of bone age detection images by deploying a self-trained bone age detection model to obtain bone age detection results. At the same time, a bone age analysis end message is generated and the bone age detection results are sent to a storage terminal.
8. The self-service bone age detection device according to claim 7, It is characterized in that The image acquisition module specifically includes: The identity comparison submodule is used to obtain the facial image of the person when the sensing area senses the presence of the user, and at the same time, extract the facial image features based on computer vision technology; The identity matching submodule is used to obtain the identity information of the person according to the facial image features, and generate a user identity matching message when the identity information of the person is obtained.
9. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, It is characterized in that When the processor executes the computer program, the steps of the self-service bone age detection method as described in any one of claims 1 to 6 are implemented.
10. A computer-readable storage medium storing a computer program. It is characterized in that When the computer program is executed by a processor, the steps of the self-service bone age detection method as described in any one of claims 1 to 6 are implemented.
