Intelligent calling and autonomous tracking toilet system
By combining RFID human movement and posture recognition technology and deep learning and machine vision technology, the difficulty of using smart toilets in the elderly and semi-disabled patients is solved, and accurate detection of movement and posture and emotional recognition are achieved, improving the improvement of movement and posture The user experience is provided and the toilet is given the ability to autonomous navigation and path planning, meeting the needs of the elderly and semi-disabled patients.
Patent Information
- Application Number
- CN202411121278.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-15
- Publication Date
- 2024-11-26
AI Technical Summary
The existing smart toilet technology is difficult to popularize on a large scale, especially for the elderly and semi-disabled patients. There are problems such as high efficiency requirements for image processing algorithms, expensive equipment performance, and strict requirements for infringement of user privacy and lighting and angles, which are difficult to meet their needs. Toilet needs.
Using RFID-based human movement and posture recognition technology combined with deep learning and machine vision technology, the action and posture detection module, human emotion recognition module, communication module, motor control module, navigation and path planning module, human-computer interaction module, power supply module and The central processing module realizes the detection and recognition of the movement behavior and posture of the elderly and semi-disabled patients, triggers the corresponding functions, and assists in the identification in emergencies.
It realizes accurate detection and identification of toilet needs of the elderly and semi-disabled patients, improves the user experience of toilets, gives emotional intelligence, meets the usage needs of the elderly and semi-disabled patients, and through autonomous navigation and The path planning module has autonomous movement and environmental adaptability.
Smart Images

Figure CN119028017A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of smart toilets, and more specifically, to a smart summoning autonomous tracking toilet system. Background Art
[0002] The care of the elderly and semi-disabled patients has brought challenges to all aspects of society. For many elderly and semi-disabled patients, going to the toilet may become very difficult and they are prone to various safety problems such as falls, slips, difficulties and psychological pressure. This not only involves their health and safety, but also their dignity.
[0003] Smart toilet, also known as intelligent toilet, refers to the technical integration of traditional toilet and modern electronic technology, which can realize a series of cutting-edge functions such as wireless control, human buttocks washing, buttocks drying, seat heating, intelligent wake-up, voice control, fall recognition, intelligent adjustment of seat temperature and automatic hand washing, as well as autonomous cleaning and deodorization in addition to cleaning dirt. Among them, fall recognition is one of the key technologies for intelligent toilet. Research in this field mainly focuses on motion recognition and posture estimation. Its biggest advantage is that it does not require users to wear equipment during detection, and it has high accuracy and visual intuitiveness. However, when the existing technology is used for recognition, there are problems such as strict requirements on algorithm efficiency in image processing, high requirements on equipment performance, high price, and possible infringement of user privacy. In addition, the camera not only has high requirements on light intensity and shooting angle, but also requires no obstacles between the user and the camera. Therefore, it is difficult to popularize on a large scale and meet the needs of the elderly and semi-disabled patients. In order to solve the above problems, a technical solution is provided. Summary of the invention
[0004] In order to overcome the above-mentioned defects of the prior art, the present invention provides an intelligent summoning autonomous tracking toilet system, which detects and identifies the movement behaviors and postures of the elderly and semi-disabled patients based on RFID human motion and posture recognition technology, so as to judge their toilet needs and trigger corresponding functions. At the same time, deep learning and machine vision technology are used to assist in human motion and posture recognition in emergency and special situations to solve the problems raised in the above-mentioned background technology.
[0005] To achieve the above object, the present invention provides the following technical solutions:
[0006] A smart summoning autonomous tracking toilet system includes a motion and posture detection module, a human emotion recognition module, a communication module, a motor control module, a navigation and path planning module, a human-computer interaction module, a power supply module and a central processing module. The motion and posture detection module detects and recognizes the user's motion behavior and posture based on RFID human motion and posture recognition technology, and detects the user's fall behavior; the motion and posture detection module includes a first camera, a human infrared sensor, a buzzer, an abnormal behavior judgment unit, a signal data acquisition unit, a data processing unit and a fall detection unit; the signal data acquisition unit collects electromagnetic wave signals through an RFID reader, and the electromagnetic wave signal S(t) finally received is ′ It is expressed as:
[0007]
[0008] Where: S(t) ′ is the electromagnetic wave signal finally received, m is the number of paths obtained by non-tag reflection, S los (t) is the electromagnetic wave signal reflected by the tag, S mud (t) is the electromagnetic wave signal reflected by other indirect paths, θ los (t) is the phase of the electromagnetic wave signal reflected in the line of sight path at time t, θ i (t) is the phase of the electromagnetic wave signal reflected by other indirect paths at time t, S i (t) is the electromagnetic wave signal reflected by the i-th indirect path.
[0009] As a further solution of the present invention, the action and posture detection module detects and identifies the user's action behavior and posture based on RFID human action and posture recognition technology, and detects the user's falling behavior; the human emotion recognition module is used to integrate a camera and a voice sensor, and cooperate with a deep learning algorithm to judge the user's emotion; the communication module is used to realize cloud communication and data upload; the motor control module is used to control the speed and direction of the toilet, and output the location information of the toilet; the navigation and path planning module is used to enable the toilet to autonomously avoid obstacles and select the optimal path to reach the destination through intelligent navigation and path planning algorithms; the human-computer interaction module is used to display information and receive user instructions, and control the toilet according to the instructions; the power supply module is used to power the toilet; the central processing module is used to control the system through the NXP i.MX series processor and communicate with each module.
[0010] As a further solution of the present invention, the motion and posture detection module includes a first camera, a human infrared sensor, a buzzer, an abnormal behavior judgment unit, a signal data acquisition unit, a data processing unit and a fall detection unit; the first camera is used to collect an image data set of the user's toilet process; the human infrared sensor is used to detect the presence and distance of the user; the buzzer is used to buzz when abnormal behavior is detected in the user's toilet process; the abnormal behavior judgment unit is used to judge whether there is abnormal behavior in the user's toilet process; the signal data acquisition unit collects electromagnetic wave signals through an RFID reader; the data processing unit is used to process electromagnetic wave signals; the fall detection unit detects fall behavior through a deep learning network model.
[0011] As a further solution of the present invention, the abnormal behavior judgment unit is used to judge whether there is abnormal behavior in the user's toilet process, based on the image of the user's toilet process collected by the first camera, the step length parameter, step frequency parameter and posture change speed parameter of the user in the image are extracted, and the standard step length parameter, standard step frequency parameter and standard posture change speed parameter of the user when they do not fall are obtained, and the step length parameter, step frequency parameter and posture change speed parameter of the user and the standard step length parameter, standard step frequency parameter and standard posture change speed parameter of the user are imported into the behavior abnormality detection model to calculate the behavior abnormality value of the user when going to the toilet. The formula of the behavior abnormality detection model is:
[0012]
[0013] Where: rc y is the abnormal value of the user's behavior when using the toilet, M is the number of images during the user's toileting process, and d bm is the step length parameter, is the standard step length parameter, p bm is the step frequency parameter, is the standard step frequency parameter, v bm is the posture change speed parameter, is the standard posture change speed parameter.
[0014] As a further solution of the present invention, the signal data acquisition unit includes an RFID reader and a signal analysis unit. The RFID reader is used to collect signal data by emitting an electromagnetic wave signal to the tag antenna to form an induced current. After the tag chip is activated, it adjusts its own impedance according to the instructions of the RFID reader, and backscatters the electromagnetic wave signal from the RFID reader antenna, and transmits the electromagnetic wave signal to the target RFID reader in a backscattered manner; the signal analysis unit is used to obtain a received signal strength index (RSSI) in the electromagnetic wave signal received by the target RFID reader, wherein the received signal strength index (RSSI) includes the electromagnetic wave signal strength and the signal phase.
[0015] As a further solution of the present invention, the fall detection unit performs auxiliary detection through deep learning and machine vision technology. Based on the directional gradient histogram, key point detection, feature description and accelerated robust feature algorithm model, the open source OpenPose tool is used for real-time 2D posture estimation of the human body. By collecting image data sets of the toileting process of the elderly and semi-disabled patients, annotating and extracting key frames, the key frames are extracted and classified using a deep neural network. The classification results will be used to judge the user's movement behavior and posture.
[0016] As a further solution of the present invention, a navigation and path planning module is used to enable the toilet to autonomously avoid obstacles and select the optimal path to reach the destination through intelligent navigation and path planning algorithms. The environmental image is captured by a third camera, and obstacles and specific areas are identified using a computer vision algorithm. The ultrasonic sensor is used to detect obstacles in front such as walls and furniture in real time and obtain their position and distance information. The SLAM algorithm is used to build an environmental map in real time and locate the toilet. At the same time, the environmental map is updated in real time according to the data from the camera and the ultrasonic sensor, and the Dijkstra or A* algorithm is used to plan the initial cleaning path.
[0017] The technical effects and advantages of the intelligent summoning autonomous tracking toilet system of the present invention are as follows: the present invention detects and identifies the movement behaviors and postures of the elderly and semi-disabled patients based on RFID human movement and posture recognition technology, thereby judging their toilet needs and triggering corresponding functions; with the help of deep learning and machine vision technology, human movement and posture recognition are assisted in emergency and special situations, and emotional intelligence is given to the toilet, so that it can better adapt to the needs and preferences of users and improve user experience; based on intelligent navigation and path planning algorithms, the toilet has autonomous movement capabilities and environmental adaptability, which can better meet the use needs of the elderly and semi-disabled patients. BRIEF DESCRIPTION OF THE DRAWINGS
[0018] Figure 1A schematic diagram of the structure of an intelligent calling autonomous tracking toilet system provided in Example 1 of the present invention. DETAILED DESCRIPTION
[0019] The following will be combined with the accompanying drawings in the present invention to clearly and completely describe the technical solution in the present invention. Obviously, the described technical solution is only a part of the present invention, not all of it. Based on the technical solution in the present invention, all other technical solutions obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.
[0020] Embodiment 1
[0021] Figure 1 FIG. 1 is a flow chart of a smart calling autonomous tracking toilet system provided by Embodiment 1 of the present invention. Figure 1 As shown, in this embodiment, a smart summoning autonomous tracking toilet system, a smart summoning autonomous tracking toilet system, includes a motion and posture detection module, a human emotion recognition module, a communication module, a motor control module, a navigation and path planning module, a human-computer interaction module, a power supply module and a central processing module;
[0022] The motion and posture detection module detects and identifies the user's motion behavior and posture based on RFID human motion and posture recognition technology, and detects the user's falling behavior;
[0023] The human emotion recognition module is used to integrate cameras and voice sensors, and cooperate with deep learning algorithms to judge user emotions;
[0024] The communication module is used to realize cloud communication and data upload;
[0025] The motor control module is used to control the speed and direction of the toilet and output the position information of the toilet;
[0026] The navigation and path planning module is used to enable the toilet to autonomously avoid obstacles and choose the best path to reach the destination through intelligent navigation and path planning algorithms;
[0027] The human-computer interaction module is used to display information and receive user instructions, and control the toilet according to the instructions;
[0028] The power supply module is used to supply power to the toilet;
[0029] The central processing module is used to control the system operation through the NXP i.MX series processor and communicate with each module.
[0030] Among them, the motion and posture detection module includes a first camera, a human infrared sensor, a buzzer, an abnormal behavior judgment unit, a signal data acquisition unit, a data processing unit and a fall detection unit; the first camera is used to collect image data sets of the user's toilet process; the human infrared sensor is used to detect the user's presence and distance; the buzzer is used to emit a buzzer when abnormal behavior is detected in the user's toilet process; the signal data acquisition unit collects electromagnetic wave signals through an RFID reader.
[0031] The human emotion recognition module includes a second camera and a first voice unit; the second camera is used to collect user facial images; the first voice unit is used to collect user voice sequence data and provide feedback.
[0032] The communication module includes a Wi-Fi unit and a 4G / 5G module; the Wi-Fi unit is used to connect to the wireless network and communicate with other devices and the cloud; the 4G / 5G module is used for remote control and data upload.
[0033] The motor control module includes a motor drive unit and an encoder; the motor drive unit is used to control the movement and action of the toilet; the encoder is used to output the position information of the toilet.
[0034] The navigation and path planning module includes a third camera and an ultrasonic sensor; the third camera is used to collect images of the user's surrounding environment; the ultrasonic sensor is used to detect obstacles.
[0035] The human-computer interaction module includes a display screen and a second voice unit; the display screen is used to display information and receive user instructions; the second voice unit is used to realize voice recognition and voice output, so that the user can control the toilet by voice.
[0036] The power supply module includes a battery, a charging unit, and a DC / DC module; the battery is used to power all devices and modules; the charging unit is used to charge the battery while ensuring safe power supply of the battery; the DC / DC module converts the direct current (DC) provided by the battery into direct current of the required voltage or current.
[0037] The motion and posture detection module includes a first camera, a human infrared sensor, a buzzer, an abnormal behavior judgment unit, a signal data acquisition unit, a data processing unit and a fall detection unit; the first camera is used to collect images of the user's toilet process; the human infrared sensor is used to detect the user's presence and distance; the abnormal behavior judgment unit is used to judge whether there is abnormal behavior during the user's toilet process; the signal data acquisition unit collects electromagnetic wave signals through an RFID reader; the data processing unit is used to process electromagnetic wave signals; the fall detection unit detects fall behavior through a deep learning network model.
[0038] The abnormal behavior judgment unit is used to judge whether there is abnormal behavior in the user's toilet process, based on the image of the user's toilet process collected by the first camera, extract the user's step length parameter, step frequency parameter and posture change speed parameter in the image, and obtain the standard step length parameter, standard step frequency parameter and standard posture change speed parameter when the user does not fall, and import the user's step length parameter, step frequency parameter and posture change speed parameter and standard step length parameter, standard step frequency parameter and standard posture change speed parameter into the behavior abnormality detection model to calculate the behavior abnormality value of the user when using the toilet. The formula of the behavior abnormality detection model is:
[0039]
[0040] Where: rc y is the abnormal value of the user's behavior when using the toilet, M is the number of images during the user's toileting process, and d bm is the step length parameter, is the standard step length parameter, p bm is the step frequency parameter, is the standard step frequency parameter, v bm is the posture change speed parameter, is the standard posture change speed parameter.
[0041] The abnormal behavior value of the user when using the toilet is compared with the preset abnormal behavior threshold. If the abnormal behavior value of the user when using the toilet is greater than or equal to the preset abnormal behavior threshold, the user has abnormal behavior when using the toilet and the buzzer will be triggered; if the abnormal behavior value of the user when using the toilet is less than the preset abnormal behavior threshold, the user has no abnormal behavior when using the toilet.
[0042] The signal data acquisition unit includes an RFID reader and a signal analysis unit. The RFID reader is used to collect signal data by emitting electromagnetic wave signals to the tag antenna to form an induced current. After the tag chip is activated, it adjusts its own impedance according to the instructions of the RFID reader, and backscatters the electromagnetic wave signals from the RFID reader antenna, and transmits the electromagnetic wave signals to the target RFID reader in a backscattered manner; the signal analysis unit is used to obtain the received signal strength index (RSS I) in the electromagnetic wave signal received by the target RFID reader, wherein the received signal strength index (RSS I) includes the electromagnetic wave signal strength and the signal phase.
[0043] The steps for RFID reader to collect signal data are as follows:
[0044] The electromagnetic wave signal backscattered by the tag chip received by the RFID reader is S(t), which is expressed as:
[0045] S(t)=|S(t)e -jθ(t) |;
[0046] Where: S(t) is the electromagnetic wave signal backscattered by the tag chip received by the RFID reader, e is an exponential function with e as the base, j is an imaginary unit, and θ(t) is the phase of the electromagnetic wave signal at time t;
[0047] According to the Friis equation and the signal reflection process, the electromagnetic wave signal intensity is quantified. The quantified electromagnetic wave signal intensity is expressed as:
[0048]
[0049] Where: R(t) is the quantized electromagnetic wave signal strength, G t is the label gain, P T is the transmission power of the reader, T b represents the backscatter transmission loss, G r is the gain of the reader antenna, d is the distance between the tag chip and the RFID reader antenna, m is the unit conversion factor, W is the unit conversion factor, usually 1, f is the frequency of the electromagnetic wave signal, and c is the propagation speed of the electromagnetic wave in a vacuum;
[0050] The phase reading of the electromagnetic pen signal scattered by the tag chip is quantitatively expressed as:
[0051]
[0052] Where: θ(t) is the quantized phase reading of the electromagnetic pen signal scattered by the tag chip, d is the distance between the tag chip and the RFID reader antenna, λ is the wavelength of the electromagnetic pen signal scattered by the tag chip, θ T (t) is the phase shift produced by the transmitting circuit of the RFID reader, θ TAG (t) is the phase shift caused by the reflection characteristics of the tag chip, θ R (t) is the phase offset generated by the receiving circuit of the RFID reader, and mod is the modulus operator, which means taking the remainder.
[0053] Among them, the electromagnetic wave signal emitted by the RFID reader will be affected by the floor, walls, human body, etc. during indoor propagation, resulting in reflection, refraction, diffraction and scattering. In addition to the direct path, it will also be reflected by different reflection surfaces, so that the receiving end receives signals from different paths at the same time, thus forming a multipath effect.
[0054] The electromagnetic wave signal emitted by the RFID reader will cause different signal strength attenuation and phase shift on different propagation paths, and the final received tag chip reflection signal S(t) ′The received electromagnetic wave signal S(t) is mainly a superposition of signals from the line-of-sight path (reflected only by the tag) and other indirect paths (reflected multiple times by nearby obstacles). ′ It is expressed as:
[0055]
[0056] Where: S(t) ′ is the electromagnetic wave signal finally received, m is the number of paths obtained by non-tag reflection, S los (t) is the electromagnetic wave signal reflected by the tag, S mud (t) is the electromagnetic wave signal reflected by other indirect paths, θ los (t) is the phase of the electromagnetic wave signal reflected in the line of sight path at time t, θ i (t) is the phase of the electromagnetic wave signal reflected by other indirect paths at time t, S i (t) is the electromagnetic wave signal reflected by the i-th indirect path.
[0057] The motion and posture detection module includes a first camera, a human infrared sensor, a buzzer, a signal data acquisition unit, a data processing unit, and a fall detection unit. First, the RFID reader reads the original phase and received signal strength index (RSSI) data stream of the passive tag in real time, and performs phase calibration and smoothing denoising through the data processing unit to minimize the environmental sensitivity problems caused by the noise in the original data and the multipath effect of the signal. The motion data is extracted by segmentation and a sample data set is established. After that, a neural network model for fall detection is constructed, and key features are extracted through a two-layer Mixer, and training parameters are adjusted and experimental comparisons are performed. Finally, the rapid online detection of the elderly's fall behavior and daily behavior is completed.
[0058] In terms of the detection and recognition of human motion behaviors and postures, deep learning and machine vision technologies are used for auxiliary detection. Based on the traditional computer vision methods: Histogram of OrientedGradients, key point detection, feature description (Scale-Invariant Feature Transform) and accelerated robust features (Speeded-Up Robust Features) algorithm model, the open source OpenPose tool is used for real-time 2D posture estimation of the human body. First, the image dataset of the toilet process of the elderly and semi-disabled patients is collected, annotated and key frames are extracted. Then, the key frames are extracted and classified using deep neural networks (such as convolutional neural networks CNN). The classification results will be used to judge the user's motion behaviors and postures, and then trigger the corresponding functions.
[0059] The human emotion recognition module is used to integrate the second camera and the first voice unit, and cooperate with the deep learning algorithm to judge the user's emotions. The user's expression feature data and voice feature data are obtained through the second camera and the first voice unit, and the expression feature data and voice feature data are used to train the deep neural network model. The trained deep neural network will be able to recognize the user's expression and voice emotions, thereby providing emotional support and personalized services to the user. For facial expression recognition: Convolutional Neural Networks technology is used for the collected user's facial images. By extracting expression feature data from the user's facial images and classifying them, the convolutional neural network can recognize people's emotions. Speech and intonation analysis: Recurrent Neural Networks technology is used to process user speech sequence data, and pre-trained deep learning models (Bidirectional Encoder Representations from Transformers) and their derivative models are used to provide text emotion analysis. In order to improve the recognition accuracy and speed, multimodal emotion recognition methods are used, such as the fusion technology (Fusion Techniques) model to fuse data features from different sources (such as face, voice and text) to improve the accuracy of emotion recognition.
[0060] When it detects that the user is in a bad mood, the human emotion recognition module can intervene in a variety of ways, such as playing relaxing piano music or playing an encouraging voice: "Everything will be fine, take a deep breath, and relax." For users who often feel depressed or lonely, some relaxing meditation exercises are recommended or users are advised to communicate with professional psychologists. In addition, the toilet is also equipped with an adaptive adjustment function. Over time, it can continuously learn and understand the user's habits and preferences more accurately, thereby providing more personalized services. For example, remember the type of music the user likes to listen to, or play a certain type of sound for the user at a specific time.
[0061] Based on deep learning and machine vision algorithms, the research on detection and recognition of human motion behaviors and postures can accurately judge the needs of the elderly and semi-disabled patients and trigger corresponding functions, providing core technical support for the intelligentization of toilets. The toilet can provide emotional support and corresponding services based on the user's facial expressions and voice emotions, giving the toilet emotional intelligence, enabling it to better adapt to the needs and preferences of users and improving the user experience.
[0062] The navigation and path planning module is used to enable the toilet to avoid obstacles autonomously and choose the best path to reach the destination through intelligent navigation and path planning algorithms. The third camera captures the environmental image and uses the computer vision algorithm to identify obstacles and specific areas. The ultrasonic sensor is responsible for real-time detection of obstacles in front such as walls and furniture, and obtains their position and distance information to assist the toilet in making obstacle avoidance decisions. The SLAM algorithm is used to build the environmental map in real time and locate the toilet position. At the same time, the environmental map is updated in real time according to the data of the camera and ultrasonic sensor. The initial path planning depends on the environmental map built by SLAM. The initial cleaning path is planned using algorithms such as Di jkstra or A*. The real-time path optimization ensures that the toilet can adjust the path according to the real-time environmental data to avoid new obstacles. The performance of the third camera, ultrasonic sensor and SLAM algorithm is tested separately to ensure that each module works properly. Then all hardware and software modules are integrated into a complete navigation system, and the system performance is tested and optimized in the actual environment, in order to achieve the goal of improving the navigation performance of the toilet and user satisfaction, and provide users with efficient and convenient cleaning services.
[0063] Based on intelligent navigation and path planning algorithms, the toilet is equipped with autonomous navigation technology, which can automatically avoid obstacles and choose the optimal path to reach the destination. The toilet has autonomous movement capabilities and environmental adaptability, which can better meet the needs of the elderly and semi-disabled patients.
[0064] In terms of toilet control and human-computer interaction, the toilet motion control technology based on model predictive control and the human-computer interaction technology based on natural language processing and computer vision use the model predictive control algorithm to achieve precise motion control of the toilet, and use the natural language processing algorithm to achieve natural language interaction with the user, and use computer vision technology to achieve visual interaction with the user, so that the user can control the operation of the toilet through simple gestures or visual commands.
[0065] Based on RFID human motion and posture recognition technology, the action behaviors and postures of the elderly and semi-disabled patients are detected and identified, so as to judge their toilet needs and trigger the corresponding functions. At the same time, deep learning and machine vision technology are used to assist in human motion and posture recognition in emergency and special situations. For example, when an elderly or semi-disabled patient sits on the toilet, the toilet can automatically detect and start the corresponding service program using RFID human motion and posture recognition technology; when the RFID human motion and posture recognition technology detects an emergency situation such as an accidental fall of the elderly or semi-disabled patient, the specific emergency situation is further judged with the help of deep learning and machine vision technology, and corresponding processing is performed, such as calling a rescue phone, sound and light alarm, or information push.
[0066] The embodiment of the present invention integrates cameras, infrared sensors and RFID technology to monitor and accurately identify the user's movements and postures in real time to ensure safe use and reduce the risk of accidents. The toilet is equipped with advanced SLAM technology and machine vision system to achieve intelligent navigation and autonomous path planning, so that it can automatically and efficiently move to the user's location or designated location. It deeply integrates emotion recognition technology to perceive and respond to the user's emotional needs through their voice, expression and biometric data to provide more humane services. In order to ensure that every user can enjoy intimate and personalized services, we have introduced machine learning algorithms to enable the system to gradually learn and adjust service strategies according to the specific needs and habits of each user. The combination of human emotion recognition and human-computer interaction technology helps to achieve a more natural and humane interactive experience and promote the depth of research on human-computer collaboration.
[0067] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.
[0068] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.
Claims
1. An intelligent summoning autonomous tracking toilet system, comprising a motion and posture detection module, a human emotion recognition module, a communication module, a motor control module, a navigation and path planning module, a human-computer interaction module, a power supply module and a central processing module, characterized in that: The motion and posture detection module detects and identifies the user's motion behavior and posture based on RFID human motion and posture recognition technology, and detects the user's falling behavior; The motion and posture detection module includes a first camera, a human infrared sensor, a buzzer, an abnormal behavior judgment unit, a signal data acquisition unit, a data processing unit, and a fall detection unit; the signal data acquisition unit collects electromagnetic wave signals through an RFID reader, and the electromagnetic wave signal S(t)′ finally received is expressed as: Where: S(t)′ is the electromagnetic wave signal finally received, m is the number of paths obtained by non-tag reflection, S los (t) is the electromagnetic wave signal reflected by the tag, S mud (t) is the electromagnetic wave signal reflected by other indirect paths, θ los (t) is the phase of the electromagnetic wave signal reflected in the line of sight path at time t, θ i (t) is the phase of the electromagnetic wave signal reflected by other indirect paths at time t, S i (t) is the electromagnetic wave signal reflected by the i-th indirect path.
2. According to claim 1, the intelligent calling autonomous tracking toilet system is characterized in that: The motion and posture detection module detects and identifies the user's motion behavior and posture based on RFID human motion and posture recognition technology, and detects the user's falling behavior; the human emotion recognition module is used to integrate cameras and voice sensors, and cooperate with deep learning algorithms to judge user emotions; the communication module is used to realize cloud communication and data upload; The motor control module is used to control the speed and direction of the toilet and output the position information of the toilet; The navigation and path planning module is used to enable the toilet to autonomously avoid obstacles and select the optimal path to reach the destination through intelligent navigation and path planning algorithms; the human-computer interaction module is used to display information and receive user instructions, and control the toilet according to the instructions; the power supply module is used to power the toilet; the central processing module is used to control the system through the NXP i.MX series processor and communicate with each module.
3. The intelligent calling autonomous tracking toilet system according to claim 1 is characterized in that: The motion and posture detection module includes a first camera, a human infrared sensor, a buzzer, an abnormal behavior judgment unit, a signal data acquisition unit, a data processing unit and a fall detection unit; the first camera is used to collect an image data set of the user's toilet process; the human infrared sensor is used to detect the presence and distance of the user; the buzzer is used to make a buzzer when abnormal behavior is detected in the user's toilet process; the abnormal behavior judgment unit is used to judge whether there is abnormal behavior in the user's toilet process; the signal data acquisition unit collects electromagnetic wave signals through an RFID reader; the data processing unit is used to process electromagnetic wave signals; the fall detection unit detects fall behavior through a deep learning network model.
4. The intelligent summoning autonomous tracking toilet system according to claim 3 is characterized in that: The abnormal behavior judgment unit is used to judge whether there is abnormal behavior in the user's toilet process, based on the image of the user's toilet process collected by the first camera, extract the user's step length parameter, step frequency parameter and posture change speed parameter in the image, and obtain the standard step length parameter, standard step frequency parameter and standard posture change speed parameter when the user does not fall, and import the user's step length parameter, step frequency parameter and posture change speed parameter and standard step length parameter, standard step frequency parameter and standard posture change speed parameter into the behavior abnormality detection model to calculate the behavior abnormality value of the user when using the toilet. The formula of the behavior abnormality detection model is: Where: rc y is the abnormal value of the user's behavior when using the toilet, M is the number of images during the user's toileting process, and d bm is the step length parameter, is the standard step length parameter, p bm is the step frequency parameter, is the standard step frequency parameter, v bm is the posture change speed parameter, is the standard posture change speed parameter.
5. The intelligent calling autonomous tracking toilet system according to claim 3 is characterized in that: The signal data acquisition unit includes an RFID reader and a signal analysis unit. The RFID reader is used to collect signal data by emitting electromagnetic wave signals to the tag antenna to form an induced current. After the tag chip is activated, it adjusts its own impedance according to the instructions of the RFID reader, and backscatters the electromagnetic wave signals from the RFID reader antenna, and transmits the electromagnetic wave signals to the target RFID reader in a backscattered manner; the signal analysis unit is used to obtain the received signal strength index (RSSI) in the electromagnetic wave signal received by the target RFID reader, wherein the received signal strength index (RSSI) includes the electromagnetic wave signal strength and the signal phase.
6. The intelligent calling autonomous tracking toilet system according to claim 1, characterized in that: The fall detection unit uses deep learning and machine vision technology to assist in detection. It uses the open source OpenPose tool for real-time 2D posture estimation of the human body based on directional gradient histograms, key point detection, feature description, and accelerated robust feature algorithm models. It collects image datasets of the toileting process of the elderly and semi-disabled patients, annotates them, extracts key frames, and uses deep neural networks to extract features and classify key frames. The classification results will be used to judge the user's movements and postures.
7. The intelligent calling autonomous tracking toilet system according to claim 1 is characterized in that: The navigation and path planning module is used to enable the toilet to autonomously avoid obstacles and select the optimal path to reach the destination through intelligent navigation and path planning algorithms. The third camera captures the environmental image and uses the computer vision algorithm to identify obstacles and specific areas. The ultrasonic sensor is used to detect obstacles in front such as walls and furniture in real time and obtain their position and distance information. The SLAM algorithm is used to build an environmental map in real time and locate the toilet. At the same time, the environmental map is updated in real time according to the data from the camera and ultrasonic sensor, and the Dijkstra or A* algorithm is used to plan the initial cleaning path.
Citation Information
Patent Citations
Robot system based on automatic tracking
CN110989689A
Human body posture radio frequency identification method for online multi-task learning
CN112183586A
Mobile pedestal pan capable of being controlled by sound and remotely and using method thereof
CN113089785A
RFID-based activity sequence identification method and system, medium and terminal
CN113361450A
Intelligent elderly accompanying system based on computer vision technology
CN116945156A
