Intelligent interconnection system of health service robot

Abstract

The invention discloses an intelligent interconnection system of a health service robot, and relates to the field of medical health. Comprising a sensor module used for acquiring vital sign information of a user; the data processing module is used for analyzing and processing the data acquired by the sensor module; the communication module is used for interconnecting the system with a mobile terminal of a guardian or a user; the user interaction module is used for interaction between the system and a guardian or a user; the online inquiry module is connected with the AI and an online doctor, online inquiry and remote medical treatment of the user are achieved, and meanwhile the doctor can remotely control the user to conduct diagnosis and treatment actions during remote medical treatment; and the large model API is used for establishing a health data model of the user through the data processing module, the inquiry module and the interaction module. According to the invention, interconnection and intercommunication with equipment such as a smart phone and a smart watch are realized through the communication module, and comprehensive health services are provided.

Description

Technical Field

[0001] The present invention relates to the field of medical health, and in particular to an intelligent interconnection system of a health service robot. Background Art

[0002] At present, the existing health service robots on the market have single main functions, lack intelligence and interconnection capabilities, and cannot meet the diverse health service needs of users.

[0003] For example, after searching, the patent of the health service system for the elderly based on the health service robot discloses a health service system for the elderly based on the health service robot, including a health service robot, an intelligent terminal and a cloud server, the health service robot includes a robot body, a main control unit, a human-computer interaction unit and a medical detection unit; the human-computer interaction unit is connected to the main control unit, which includes a tablet computer, and the tablet computer is placed on the chest of the robot body; the medical detection unit is connected to the main control unit, which includes an electrocardiogram detection device and a blood oxygen saturation detection device, the electrocardiogram detection device and the blood oxygen saturation detection device are integrated on the robot body, and are respectively connected to the intelligent terminal and the tablet computer through Bluetooth signals; the tablet computer and the intelligent terminal are connected to the cloud server through mobile Internet.

[0004] It has the following shortcomings: single function, lack of intelligence and interconnection capabilities, and inability to meet users' diverse health service needs.

[0005] To this end, the present invention proposes a smart interconnection system for a health service robot Summary of the invention

[0006] The purpose of the present invention is to solve the shortcomings of the prior art and to propose an intelligent interconnection system for a health service robot.

[0007] In order to achieve the above object, the present invention adopts the following technical solutions:

[0008] An intelligent interconnection system of a health service robot, comprising:

[0009] A sensor module, which is used to obtain the user's vital signs information;

[0010] A data processing module, which is used to analyze and process the data collected by the sensor module;

[0011] A communication module, which is used to interconnect the system with the mobile terminal of the guardian or user;

[0012] A user interaction module, which is used for the system to interact with the guardian or user;

[0013] The online consultation module connects with AI and online doctors to enable users to have online consultations and telemedicine. During telemedicine, doctors can also remotely control the diagnosis and treatment of patients.

[0014] The big model API builds the user's health data model through the data processing module, the consultation module and the interaction module.

[0015] Preferably, the sensor module includes any one or more combinations of a heart rate sensor, a blood pressure sensor, a body temperature sensor, a respiration sensor, and a blood oxygen sensor.

[0016] Preferably: the communication module supports Wi-Fi, Bluetooth or operator communication mode.

[0017] Preferably, the data processing module adopts a threshold-based processing method for vital sign data monitoring.

[0018] Preferably, the method for processing vital sign data by the data processing module comprises the following steps:

[0019] A1: Set the normal threshold range for each vital sign and the measurement error of each sensor Δ i ;

[0020] A2: Get the data collected by the sensor and process the data to get M;

[0021] A3: If or If it is abnormal, it is considered to be over-limit, otherwise it is normal.

[0022] Preferably: in step A2, the data accuracy processing includes the following steps:

[0023] A21: According to the correlation between the vital signs, a multivariate function relationship is established with one vital sign as the dependent variable and the other vital signs as independent variables, a=f(b, c, d, ...), where a, b, c, d, ... are the vital signs and f is the function relationship;

[0024] A22: Set the offset proportional coefficient k;

[0025] A23: Calculate the theoretical value M1 of the current vital sign and the actual value M2 of the current vital sign obtained by the sensor through the parameters collected by other vital signs and the multivariate function relationship;

[0026] A24: If M2∈((1-k)M1, (1+k)M1), then M=M2, otherwise M=M1.

[0027] In the step A22, k∈(0.05, 0.15).

[0028] Preferably, the intelligent interconnection system of the health service robot further includes an obstacle avoidance module, and its working logic includes the following steps:

[0029] S1: Environmental data collection, using a camera to capture environmental images and obtain real-time data of nearby environmental images;

[0030] S2: Data processing, the data processing module processes and analyzes the collected images to determine the relative position of the surrounding environment and the robot;

[0031] S3: Positioning: According to the robot model, determine the maximum three-dimensional space occupied by the robot when it moves;

[0032] S4: Execution control, controlling the robot to move to avoid obstacles according to the relative position of the robot and the surrounding environment and the maximum three-dimensional space occupied by the robot when moving;

[0033] S5: During movement, the laser radar and ultrasonic radar continuously scan the surrounding area to ensure successful obstacle avoidance.

[0034] Preferably: in the step S2, the data processing method is:

[0035] S21: Image acquisition, obtaining image information of the external environment captured by the camera;

[0036] S22: image preprocessing, preprocessing the collected image;

[0037] S23: feature extraction, performing feature extraction on the processed image;

[0038] S24: feature matching, matching the extracted features with existing features to identify objects in the image;

[0039] S25: relative position calculation, determining the positions of the robot and the object according to the shooting angle of the camera and the position of the object relative to the environment, thereby obtaining the relative position.

[0040] Preferably, in S24, the feature matching algorithm adopts a matching algorithm based on Gaussian difference pyramid and gradient direction histogram, which specifically includes the following steps:

[0041] A1: Scale space extreme value detection: The scale space of the image is obtained through the Gaussian difference pyramid, each scale space is Gaussian blurred, and then the Gaussian difference is used to detect the extreme points in the image;

[0042] A2: Key point location, by detecting extreme points in the scale space, using interpolation to accurately locate the position and scale of key points;

[0043] A3: Direction assignment: In the neighborhood around the key point, the histogram of gradient magnitude and direction is calculated to obtain the main direction, and the gradient in the neighborhood is rotated to make its direction consistent with the main direction;

[0044] A4: Key point description, in the neighborhood around the key point, calculate the histogram of gradient magnitude and direction to obtain the feature vector, and normalize the feature vector to obtain the descriptor.

[0045] The beneficial effects of the present invention are:

[0046] 1. Multi-device interconnection: Through the communication module, it can realize interconnection with smart phones, smart watches and other devices to provide comprehensive health services.

[0047] 2. Intelligent services: Through data processing modules and control modules, personalized health monitoring and services can be realized to improve user experience.

[0048] 3. Real-time monitoring: The sensor module can monitor the user's health data in real time, detect health problems in time and provide early warnings.

[0049] 4. Telemedicine: Through the online consultation module, users can conveniently conduct telemedicine consultations and obtain professional health advice.

[0050] 5. Intelligent analysis: By accessing the big model API, the robot can conduct in-depth analysis of the collected health data and provide more accurate health assessments and recommendations. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 This is an architectural diagram of the intelligent interconnection system of a health service robot proposed in the present invention. DETAILED DESCRIPTION

[0052] The technical solution of the present invention is further described in detail below in conjunction with specific implementation methods.

[0053] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", and "set" should be understood in a broad sense, for example, they can be fixedly connected or set, or detachably connected or set, or integrally connected or set. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0054] Embodiment 1:

[0055] A smart interconnection system for a health service robot, comprising:

[0056] A sensor module, which is used to obtain the user's vital signs information;

[0057] A data processing module, which is used to analyze and process the data collected by the sensor module;

[0058] A communication module, which is used to interconnect the system with the mobile terminal of the guardian or user;

[0059] A user interaction module, which is used for the system to interact with the guardian or user;

[0060] The online consultation module connects with AI and online doctors to enable users to have online consultations and telemedicine. During telemedicine, doctors can also remotely control the diagnosis and treatment of patients.

[0061] The big model API builds the user's health data model through the data processing module, the consultation module and the interaction module.

[0062] The sensor module includes any one or more combinations of a heart rate sensor, a blood pressure sensor, a body temperature sensor, a breathing sensor, and a blood oxygen sensor.

[0063] The communication module supports Wi-Fi, Bluetooth or operator communication mode.

[0064] The data processing module adopts a threshold-based processing method for monitoring vital sign data.

[0065] The method for processing vital sign data by the data processing module comprises the following steps:

[0066] A1: Set the normal threshold range for each vital sign and the measurement error of each sensor Δ i ;

[0067] A2: Get the data collected by the sensor and process the data to get M;

[0068] A3: If or If it is abnormal, it is considered to be over-limit, otherwise it is normal.

[0069] In step A2, the data accuracy processing includes the following steps:

[0070] A21: According to the correlation between the vital signs, a multivariate function relationship is established with one vital sign as the dependent variable and the other vital signs as independent variables, a=f(b, c, d, ...), where a, b, c, d, ... are the vital signs and f is the function relationship;

[0071] A22: Set the offset proportional coefficient k;

[0072] A23: Calculate the theoretical value M1 of the current vital sign and the actual value M2 of the current vital sign obtained by the sensor through the parameters collected by other vital signs and the multivariate function relationship;

[0073] A24: If M2∈((1-k)M1, (1+k)M1), then M=M2, otherwise M=M1.

[0074] In the step A22, k=0.05.

[0075] Embodiment 2:

[0076] A smart interconnection system for a health service robot, comprising:

[0077] A sensor module, which is used to obtain the user's vital signs information;

[0078] A data processing module, which is used to analyze and process the data collected by the sensor module;

[0079] A communication module, which is used to interconnect the system with the mobile terminal of the guardian or user;

[0080] A user interaction module, which is used for the system to interact with the guardian or user;

[0081] The online consultation module connects with AI and online doctors to enable users to have online consultations and telemedicine. During telemedicine, doctors can also remotely control the diagnosis and treatment of patients.

[0082] The big model API builds the user's health data model through the data processing module, the consultation module and the interaction module.

[0083] The sensor module includes any one or more combinations of a heart rate sensor, a blood pressure sensor, a body temperature sensor, a breathing sensor, and a blood oxygen sensor.

[0084] The communication module supports Wi-Fi, Bluetooth or operator communication mode.

[0085] The data processing module adopts a threshold-based processing method for monitoring vital sign data.

[0086] The method for processing vital sign data by the data processing module comprises the following steps:

[0087] A1: Set the normal threshold range for each vital sign and the measurement error of each sensor Δ i ;

[0088] A2: Get the data collected by the sensor and process the data to get M;

[0089] A3: If or If it is abnormal, it is considered to be over-limit, otherwise it is normal.

[0090] In step A2, the data accuracy processing includes the following steps:

[0091] A21: According to the correlation between the vital signs, a multivariate function relationship is established with one vital sign as the dependent variable and the other vital signs as independent variables, a=f(b, c, d, ...), where a, b, c, d, ... are the vital signs and f is the function relationship;

[0092] A22: Set the offset proportional coefficient k;

[0093] A23: Calculate the theoretical value M1 of the current vital sign and the actual value M2 of the current vital sign obtained by the sensor through the parameters collected by other vital signs and the multivariate function relationship;

[0094] A24: If M2∈((1-k)M1, (1+k)M1), then M=M2, otherwise M=M1.

[0095] In the step A22, k=0.1.

[0096] Embodiment 3:

[0097] A smart interconnection system for a health service robot, comprising:

[0098] A sensor module, which is used to obtain the user's vital signs information;

[0099] A data processing module, which is used to analyze and process the data collected by the sensor module;

[0100] A communication module, which is used to interconnect the system with the mobile terminal of the guardian or user;

[0101] A user interaction module, which is used for the system to interact with the guardian or user;

[0102] The online consultation module connects with AI and online doctors to enable users to have online consultations and telemedicine. During telemedicine, doctors can also remotely control the diagnosis and treatment of patients.

[0103] The big model API builds the user's health data model through the data processing module, the consultation module and the interaction module.

[0104] The sensor module includes any one or more combinations of a heart rate sensor, a blood pressure sensor, a body temperature sensor, a breathing sensor, and a blood oxygen sensor.

[0105] The communication module supports Wi-Fi, Bluetooth or operator communication mode.

[0106] The data processing module adopts a threshold-based processing method for monitoring vital sign data.

[0107] The method for processing vital sign data by the data processing module comprises the following steps:

[0108] A1: Set the normal threshold range for each vital sign and the measurement error of each sensor Δ i ;

[0109] A2: Get the data collected by the sensor and process the data to get M;

[0110] A3: If or If it is abnormal, it is considered to be over-limit, otherwise it is normal.

[0111] In step A2, the data accuracy processing includes the following steps:

[0112] A21: According to the correlation between the vital signs, a multivariate function relationship is established with one vital sign as the dependent variable and the other vital signs as independent variables, a=f(b, c, d, ...), where a, b, c, d, ... are the vital signs and f is the function relationship;

[0113] A22: Set the offset proportional coefficient k;

[0114] A23: Calculate the theoretical value M1 of the current vital sign and the actual value M2 of the current vital sign obtained by the sensor through the parameters collected by other vital signs and the multivariate function relationship;

[0115] A24: If M2∈((1-k)M1, (1+k)M1), then M=M2, otherwise M=M1.

[0116] In the step A22, k=0.15.

[0117] Embodiment 4:

[0118] A smart interconnection system for a health service robot, comprising:

[0119] A sensor module, which is used to obtain the user's vital signs information;

[0120] A data processing module, which is used to analyze and process the data collected by the sensor module;

[0121] A communication module, which is used to interconnect the system with the mobile terminal of the guardian or user;

[0122] A user interaction module, which is used for the system to interact with the guardian or user;

[0123] The online consultation module connects with AI and online doctors to enable users to have online consultations and telemedicine. During telemedicine, doctors can also remotely control the diagnosis and treatment of patients.

[0124] The big model API builds the user's health data model through the data processing module, the consultation module and the interaction module.

[0125] The sensor module includes any one or more combinations of a heart rate sensor, a blood pressure sensor, a body temperature sensor, a breathing sensor, and a blood oxygen sensor.

[0126] The communication module supports Wi-Fi, Bluetooth or operator communication mode.

[0127] The data processing module adopts a threshold-based processing method for monitoring vital sign data.

[0128] The method for processing vital sign data by the data processing module comprises the following steps:

[0129] A1: Set the normal threshold range for each vital sign and the measurement error of each sensor Δ i ;

[0130] A2: Get the data collected by the sensor and process the data to get M;

[0131] A3: If or If it is abnormal, it is considered to be over-limit, otherwise it is normal.

[0132] In step A2, the data accuracy processing includes the following steps:

[0133] A21: According to the correlation between the vital signs, a multivariate function relationship is established with one vital sign as the dependent variable and the other vital signs as independent variables, a=f(b, c, d, ...), where a, b, c, d, ... are the vital signs and f is the function relationship;

[0134] A22: Set the offset proportional coefficient k;

[0135] A23: Calculate the theoretical value M1 of the current vital sign and the actual value M2 of the current vital sign obtained by the sensor through the parameters collected by other vital signs and the multivariate function relationship;

[0136] A24: If M2∈((1-k)M1, (1+k)M1), then M=M2, otherwise M=M1.

[0137] In the step A22, k∈(0.05, 0.15).

[0138] The intelligent interconnection system of the health service robot also includes an obstacle avoidance module, and its working logic includes the following steps:

[0139] S1: Environmental data collection, using a camera to capture environmental images and obtain real-time data of nearby environmental images;

[0140] S2: Data processing, the data processing module processes and analyzes the collected images to determine the relative position of the surrounding environment and the robot;

[0141] S3: Positioning: According to the robot model, determine the maximum three-dimensional space occupied by the robot when it moves;

[0142] S4: Execution control, controlling the robot to move to avoid obstacles according to the relative position of the robot and the surrounding environment and the maximum three-dimensional space occupied by the robot when moving;

[0143] S5: During movement, the laser radar and ultrasonic radar continuously scan the surrounding area to ensure successful obstacle avoidance.

[0144] In the step S2, the data processing method is:

[0145] S21: Image acquisition, obtaining image information of the external environment captured by the camera;

[0146] S22: image preprocessing, preprocessing the collected image;

[0147] S23: feature extraction, performing feature extraction on the processed image;

[0148] S24: feature matching, matching the extracted features with existing features to identify objects in the image;

[0149] S25: relative position calculation, determining the positions of the robot and the object according to the shooting angle of the camera and the position of the object relative to the environment, thereby obtaining the relative position.

[0150] Embodiment 5:

[0151] A smart interconnection system for a health service robot, comprising:

[0152] A sensor module, which is used to obtain the user's vital signs information;

[0153] A data processing module, which is used to analyze and process the data collected by the sensor module;

[0154] A communication module, which is used to interconnect the system with the mobile terminal of the guardian or user;

[0155] A user interaction module, which is used for the system to interact with the guardian or user;

[0156] The online consultation module connects with AI and online doctors to enable users to have online consultations and telemedicine. During telemedicine, doctors can also remotely control the diagnosis and treatment of patients.

[0157] The big model API builds the user's health data model through the data processing module, the consultation module and the interaction module.

[0158] The sensor module includes any one or more combinations of a heart rate sensor, a blood pressure sensor, a body temperature sensor, a breathing sensor, and a blood oxygen sensor.

[0159] The communication module supports Wi-Fi, Bluetooth or operator communication mode.

[0160] The data processing module adopts a threshold-based processing method for monitoring vital sign data.

[0161] The method for processing vital sign data by the data processing module comprises the following steps:

[0162] A1: Set the normal threshold range for each vital sign and the measurement error of each sensor Δ i ;

[0163] A2: Get the data collected by the sensor and process the data to get M;

[0164] A3: If or If it is abnormal, it is considered to be over-limit, otherwise it is normal.

[0165] In step A2, the data accuracy processing includes the following steps:

[0166] A21: According to the correlation between the vital signs, a multivariate function relationship is established with one vital sign as the dependent variable and the other vital signs as independent variables, a=f(b, c, d, ...), where a, b, c, d, ... are the vital signs and f is the function relationship;

[0167] A22: Set the offset proportional coefficient k;

[0168] A23: Calculate the theoretical value M1 of the current vital sign and the actual value M2 of the current vital sign obtained by the sensor through the parameters collected by other vital signs and the multivariate function relationship;

[0169] A24: If M2∈((1-k)M1, (1+k)M1), then M=M2, otherwise M=M1.

[0170] In the step A22, k∈(0.05, 0.15).

[0171] The intelligent interconnection system of the health service robot also includes an obstacle avoidance module, and its working logic includes the following steps:

[0172] S1: Environmental data collection, using a camera to capture environmental images and obtain real-time data of nearby environmental images;

[0173] S2: Data processing, the data processing module processes and analyzes the collected images to determine the relative position of the surrounding environment and the robot;

[0174] S3: Positioning: According to the robot model, determine the maximum three-dimensional space occupied by the robot when it moves;

[0175] S4: Execution control, controlling the robot to move to avoid obstacles according to the relative position of the robot and the surrounding environment and the maximum three-dimensional space occupied by the robot when moving;

[0176] S5: During movement, the laser radar and ultrasonic radar continuously scan the surrounding area to ensure successful obstacle avoidance.

[0177] In the step S2, the data processing method is:

[0178] S21: Image acquisition, obtaining image information of the external environment captured by the camera;

[0179] S22: image preprocessing, preprocessing the collected image;

[0180] S23: feature extraction, performing feature extraction on the processed image;

[0181] S24: feature matching, matching the extracted features with existing features to identify objects in the image;

[0182] S25: relative position calculation, determining the positions of the robot and the object according to the shooting angle of the camera and the position of the object relative to the environment, thereby obtaining the relative position.

[0183] In S24, the feature matching algorithm adopts a matching algorithm based on Gaussian difference pyramid and gradient direction histogram, which specifically includes the following steps:

[0184] A1: Scale space extreme value detection: The scale space of the image is obtained through the Gaussian difference pyramid, each scale space is Gaussian blurred, and then the Gaussian difference is used to detect the extreme points in the image;

[0185] A2: Key point location, by detecting extreme points in the scale space, using interpolation to accurately locate the position and scale of key points;

[0186] A3: Direction assignment: In the neighborhood around the key point, the histogram of gradient magnitude and direction is calculated to obtain the main direction, and the gradient in the neighborhood is rotated to make its direction consistent with the main direction;

[0187] A4: Key point description, in the neighborhood around the key point, calculate the histogram of gradient magnitude and direction to obtain the feature vector, and normalize the feature vector to obtain the descriptor.

[0188] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. An intelligent interconnection system for a health service robot, characterized in that: include: A sensor module, which is used to obtain vital sign information of the user; A data processing module, which is used to analyze and process the data collected by the sensor module; A communication module, which is used to interconnect the system with the mobile terminal of the guardian or user; A user interaction module, which is used for the system to interact with the guardian or user; The online consultation module connects with AI and online doctors to enable users to have online consultations and telemedicine. During telemedicine, doctors can also remotely control the diagnosis and treatment of patients. The big model API builds the user's health data model through the data processing module, the consultation module and the interaction module.

2. The intelligent interconnection system of a health service robot according to claim 1, characterized in that: The sensor module includes any one or more combinations of a heart rate sensor, a blood pressure sensor, a body temperature sensor, a breathing sensor, and a blood oxygen sensor.

3. The intelligent interconnection system of a health service robot according to claim 1, characterized in that: The communication module supports Wi-Fi, Bluetooth or operator communication mode.

4. The intelligent interconnection system of a health service robot according to claim 1, characterized in that: The data processing module adopts a threshold-based processing method for monitoring vital sign data.

5. The intelligent interconnection system of a health service robot according to claim 4, characterized in that: The method for processing vital sign data by the data processing module comprises the following steps: A1: Set the normal threshold range for each vital sign and the measurement error of each sensor Δ i ; A2: Get the data collected by the sensor and process the data to get M; A3: If or If it is abnormal, it is considered to be over-limit, otherwise it is normal.

6. The intelligent interconnection system of a health service robot according to claim 5, characterized in that: In step A2, the data accuracy processing includes the following steps: A21: According to the correlation between the vital signs, a multivariate function relationship is established with one vital sign as the dependent variable and the other vital signs as independent variables, a=f(b, c, d, ...), where a, b, c, d, ... are the vital signs and f is the function relationship; A22: Set the offset proportional coefficient k; A23: Calculate the theoretical value M1 of the current vital sign and the actual value M2 of the current vital sign obtained by the sensor through the parameters collected by other vital signs and the multivariate function relationship; A24: If M2∈((1-k)M1, (1+k)M1), then M=M2, otherwise M=M1. In the step A22, k∈(0.05, 0.15).

7. The intelligent interconnection system of a health service robot according to claim 1, characterized in that: The intelligent interconnection system of the health service robot also includes an obstacle avoidance module, and its working logic includes the following steps: S1: Environmental data collection, using a camera to capture environmental images and obtain real-time data of nearby environmental images; S2: Data processing, the data processing module processes and analyzes the collected images to determine the relative position of the surrounding environment and the robot; S3: Positioning: According to the robot model, determine the maximum three-dimensional space occupied by the robot when it moves; S4: Execution control, controlling the robot to move to avoid obstacles according to the relative position of the robot and the surrounding environment and the maximum three-dimensional space occupied by the robot when moving; S5: During movement, the laser radar and ultrasonic radar continuously scan the surrounding area to ensure successful obstacle avoidance.

8. The intelligent interconnection system of a health service robot according to claim 7, characterized in that: In the step S2, the data processing method is: S21: Image acquisition, obtaining image information of the external environment captured by the camera; S22: image preprocessing, preprocessing the collected image; S23: feature extraction, performing feature extraction on the processed image; S24: feature matching, matching the extracted features with existing features to identify objects in the image; S25: relative position calculation, determining the positions of the robot and the object according to the shooting angle of the camera and the position of the object relative to the environment, thereby obtaining the relative position.

9. The intelligent interconnection system of a health service robot according to claim 8, characterized in that: In S24, the feature matching algorithm adopts a matching algorithm based on Gaussian difference pyramid and gradient direction histogram, which specifically includes the following steps: A1: Scale space extreme value detection: The scale space of the image is obtained through the Gaussian difference pyramid, each scale space is Gaussian blurred, and then the Gaussian difference is used to detect the extreme points in the image; A2: Key point location, by detecting extreme points in the scale space, using interpolation to accurately locate the position and scale of key points; A3: Direction assignment: In the neighborhood around the key point, the histogram of gradient magnitude and direction is calculated to obtain the main direction, and the gradient in the neighborhood is rotated to make its direction consistent with the main direction; A4: Key point description, in the neighborhood around the key point, calculate the histogram of gradient magnitude and direction to obtain the feature vector, and normalize the feature vector to obtain the descriptor.

Images