Intelligent nasal cavity cleaning pressure pump regulation and control method and system based on image processing

Abstract

The invention relates to the technical field of control, in particular to an intelligent nasal cavity cleaning pressure pump regulation and control method and system based on image processing. The method comprises the following steps: creating a three-dimensional model based on a first image at different positions of a plurality of angles in a target object; pre-cleaning the target object according to the pre-cleaning pressure and the cleaning position corresponding to each grid; clustering according to the pixel value of each pixel on the basis of the three-dimensional image, inputting a sub-image corresponding to each target cluster into a classification model, and obtaining a state type and cleaning information of a sub-region corresponding to each sub-image; dividing each sub-region into a plurality of cleaning units, simulating regulation and control information when the cleaning nozzle cleans each cleaning unit and a flow path corresponding to the first cleaning agent, and cleaning the cleaning units based on the regulation and control information; and repeating the cleaning method of the sub-regions to complete cleaning of the target object. The problem that when the nasal cavity is cleaned, the comfort of a user is poor is solved, and the comfort of the user when the nasal cavity is cleaned is improved.

Description

Technical Field

[0001] The present application relates to the field of control technology, and in particular to an intelligent nasal cleaning pressure pump control method and system based on image processing. Background Art

[0002] Nasal cleaning is a common nasal care method used to remove pollutants, allergens or obstructions in the nasal cavity. It plays an important therapeutic and care role, especially for patients with rhinitis, especially allergic rhinitis. With the advancement of technology, traditional cleaning methods have been gradually replaced by various automated and intelligent equipment. Similar prior art includes a Chinese patent with publication number CN110955166B, which proposes an intelligent washing device and control method based on image recognition, wherein the device includes a cleaning device and a control device, and the control device includes: an image acquisition module, a background module and a control module; the water outlet of the cleaning device is relatively provided with an image acquisition module and a background module; the image acquisition module is used to collect the water quality image of the water outlet flowing into the surface of the background module; the control module is respectively connected to the image acquisition module and the control unit of the cleaning device by signal, and the control module is used to obtain the water quality image of the water outlet collected by the image acquisition module and the working state parameters of the cleaning device in the current washing mode; the control module is also used to identify the water quality image of the water outlet to obtain the current water quality information of the water outlet, and adjust the working state parameters in the current washing mode according to the current water quality information of the water outlet, and feed back the adjusted working state parameters to the control unit of the cleaning device to control the cleaning device to perform the corresponding washing work with the adjusted working state parameters. In addition, similar prior art also includes a Chinese patent with publication number CN115774407A, which proposes a control method for a shoe washing machine, aiming to solve the problem that the existing shoe washing machine requires the user to manually set the shoe washing program. To this end, the control method of the present invention includes: acquiring an image of the shoe; determining the information of the shoe based on the image; when the shoe is in the washing tub, controlling the water level in the washing tub to a preset water level; after the water level in the washing tub reaches the preset water level, acquiring the turbidity information of the washing water; judging the degree of dirtiness of the shoe based on the turbidity information; determining the washing program of the shoe washing machine based on the shoe information and the degree of dirtiness of the shoe; wherein the shoe information includes at least one of the type and size of the shoe, and the washing program can be automatically determined based on the image of the shoe and the turbidity information of the washing water, so that the washing program determined in this way can be more suitable for the actual state of the shoe for washing, so that the shoe can be better cleaned and a better washing effect can be obtained. Both of the above patent applications generate control information for cleaning equipment through image recognition and clean the target cleaning objects, but do not adopt different cleaning solutions based on the different conditions of different areas, nor do they control the cleaning process of the cleaning equipment to prevent the cleaning agent for a specific area from causing damage to other cleaning areas. Summary of the invention

[0003] The present application provides an intelligent nasal cleaning pressure pump control method based on image processing, wherein a camera unit is provided on the cleaning nozzle of the cleaning pressure pump, and the method comprises:

[0004] Step S1: acquiring first images at multiple angles and positions within a target object through the camera unit, and creating a three-dimensional model of the target object based on the first images;

[0005] Step S2: dividing the target object into a plurality of grids based on the three-dimensional model, obtaining a pre-cleaning position according to the pre-cleaning pressure corresponding to the grid and the characteristics of the cleaning nozzle, and pre-cleaning the target object at the pre-cleaning pressure and the pre-cleaning position corresponding to each grid;

[0006] Step S3: after the pre-cleaning, the three-dimensional model is updated, the three-dimensional image in the three-dimensional model is clustered according to the pixel value of each pixel, and a plurality of target clusters are obtained, and the sub-image corresponding to each of the target clusters is input into the classification model, and the state type and cleaning information of the sub-region corresponding to each of the sub-images are obtained;

[0007] Step S4: each of the sub-areas is divided into a plurality of cleaning units, and control information and a flow path of the corresponding first cleaning agent when the cleaning nozzle cleans each of the cleaning units are simulated through a three-dimensional model according to the cleaning information of the sub-areas, the cleaning units are cleaned based on the control information, and the cleaning path of the cleaning nozzle is adjusted according to the conflict between the characteristics of the first cleaning agent and the state types of other sub-areas through which the flow path passes;

[0008] Step S5: Repeat the sub-area cleaning method in step S4 to complete the cleaning of the target object.

[0009] As a preferred technical solution of the present invention, step S4 further includes:

[0010] When cleaning the sub-area, the airflow information of the target object when breathing is also obtained through the detection unit on the cleaning nozzle, and correction information of the cleaning unit is obtained based on the airflow information and the three-dimensional model, and the control information of the cleaning unit is corrected based on the correction information.

[0011] As a preferred technical solution of the present invention, step S2 includes:

[0012] The three-dimensional model is divided into a plurality of grids based on the spraying area of ​​the cleaning nozzle, and the normal vector at the center position of each grid is obtained, the cleaning distance of the grid is obtained according to the pre-cleaning pressure corresponding to the grid and the characteristics of the cleaning nozzle, the cleaning position of the cleaning nozzle on the normal vector is determined according to the cleaning distance, and the grid is pre-cleaned with the pre-cleaning pressure, wherein the characteristic of the cleaning nozzle is the correspondence between the cleaning distance of the cleaning nozzle from the target object and the cleaning pressure when the cleaning nozzle sprays the cleaning agent.

[0013] As a preferred technical solution of the present invention, step S3 includes:

[0014] After the target object has been pre-cleaned, a second image of the target object at different positions is reacquired, and the three-dimensional model is updated through the second image. The three-dimensional image corresponding to the three-dimensional model is clustered according to pixel values, and a plurality of target clustering clusters with different pixel values ​​are acquired. A pixel area corresponding to each of the target clustering clusters and the number of continuous pixels therein is greater than or equal to N is taken as the sub-image, and the sub-image is input into the classification model to acquire a state type and cleaning information of a sub-area of ​​the target object corresponding to the sub-image, wherein the state type is the state type of the nasal mucosa of the sub-area, and the cleaning information includes a cleaning agent and a cleaning pressure of the sub-area.

[0015] As a preferred technical solution of the present invention, step S4 includes:

[0016] Divide each of the sub-regions into a plurality of cleaning units according to the spraying area of ​​the cleaning nozzle, and simulate the control information of the cleaning nozzle when cleaning each of the cleaning units and the flow path of the corresponding first cleaning agent through the three-dimensional model according to the cleaning information corresponding to the sub-region, wherein the control information includes the cleaning direction, the cleaning position and the cleaning duration, and sort the plurality of cleaning units according to the distance from the flow path outside the sub-region, and obtain a cleaning list, wherein the farther the distance from the flow path outside the sub-region, the higher the sorting, and judge whether the characteristics of the first cleaning agent conflict with the state type of other sub-regions through which the flow path corresponding to each cleaning unit passes based on the characteristics of the first cleaning agent in the cleaning information of each sub-region;

[0017] When there is no conflict, the cleaning nozzle performs normal cleaning based on the control information corresponding to each cleaning unit in the cleaning list; when the characteristics of the first cleaning agent conflict with the state type in the other sub-areas, the other sub-areas are used as conflicting sub-areas, and when the cleaning nozzle cleans the cleaning units based on the control information corresponding to each cleaning unit in the cleaning list, when the distance between the real-time flow path of the first cleaning agent of any cleaning unit and the conflicting sub-area is less than or equal to the set distance, the first cleaning agent is switched to the second cleaning agent and the conflicting sub-area is cleaned based on the control information corresponding to the cleaning unit in the conflicting sub-area until the first cleaning agent is flushed out of the conflicting sub-area; repeat this step to complete the cleaning of the sub-area.

[0018] As a preferred technical solution of the present invention, step S4 further includes:

[0019] When cleaning the sub-area, the detection unit on the cleaning nozzle is also used to detect the airflow information of the target object when breathing, and the airflow information includes the airflow size, airflow direction and corresponding airflow time period distribution;

[0020] Before cleaning the cleaning unit in the sub-area, a cleaning airflow time period corresponding to the cleaning unit is calculated according to the cleaning duration in the control information corresponding to the cleaning unit and the current airflow time period, and when the cleaning airflow time period is the second time period, the cleaning nozzle performs normal cleaning based on the control information corresponding to the cleaning unit;

[0021] When the airflow time period is the first time period and the third time period, the airflow size and airflow direction corresponding to the airflow time period are simulated by the three-dimensional model based on the control information of the cleaning nozzle, and the correction information of the cleaning nozzle is obtained, and the control information is corrected based on the correction information, and the cleaning unit is cleaned according to the corrected control information, wherein the airflow time period includes the first time period, the second time period and the third time period, the first time period is the inhalation time period, the second time period is the gap time period, and the third time period is the exhalation time period.

[0022] As a preferred technical solution of the present invention, the classification model is a machine learning model trained with historical sample data, wherein the historical sample data includes nasal sample images and corresponding label information, and the label information includes the nasal state type and cleaning information corresponding to the nasal sample images.

[0023] As a preferred technical solution of the present invention, the nasal cleaning pressure pump can switch between the first cleaning agent and the second cleaning agent by controlling the cleaning agent switching port.

[0024] The present invention also provides an intelligent nasal cleaning pressure pump control system based on image processing, the system is used to implement the above method, and the system includes:

[0025] A camera unit, used to obtain first images at different angles and positions within a target object;

[0026] a model creation unit, configured to create a three-dimensional model of the target object based on the first image;

[0027] A pre-cleaning unit, configured to divide the target object into a plurality of grids based on the three-dimensional model, obtain a pre-cleaning position according to a pre-cleaning pressure corresponding to the grids and a characteristic of a cleaning nozzle, and pre-clean the target object at the pre-cleaning pressure and the pre-cleaning position corresponding to each grid;

[0028] a cleaning information acquisition unit, configured to update the three-dimensional model after the pre-cleaning, cluster the three-dimensional image in the three-dimensional model according to the pixel value of each pixel, and obtain a plurality of target clusters, input the sub-image corresponding to each of the target clusters into the classification model, and obtain the state type and cleaning information of the sub-region corresponding to each of the sub-images;

[0029] The sub-area cleaning unit is configured as follows: each of the sub-areas is divided into a plurality of cleaning units, and according to the cleaning information of the sub-areas, control information and a flow path of a corresponding first cleaning agent when the cleaning nozzle cleans each of the cleaning units are simulated through a three-dimensional model, the cleaning unit is cleaned based on the control information, and the cleaning path of the cleaning nozzle is adjusted according to a conflict between the characteristics of the first cleaning agent and the state types of other sub-areas through which the flow path passes; the cleaning method of the sub-areas is repeated to complete the cleaning of the target object.

[0030] The present invention also provides a computer-readable storage medium, on which instructions are stored, and the above method is implemented when the instructions are executed by a processor.

[0031] Technical Effects

[0032] The present invention significantly improves the cleaning efficiency and user comfort through intelligent technology. First, a multi-angle first image of the inside of the nasal cavity is obtained through a camera unit, and a three-dimensional model is created based on the first image, so that the cleaning process can accurately target the specific structure in the nasal cavity, effectively avoiding the blind spots that may exist in traditional cleaning methods, and improving the accuracy and efficiency of cleaning. In the pre-cleaning step, the system divides the nasal cavity into multiple grids according to the three-dimensional model, and determines the cleaning position according to the pre-cleaning pressure corresponding to each grid and the characteristics of the cleaning nozzle, so as to perform a preliminary cleaning of the nasal cavity, which not only cleans the secretions, but also provides a basis for subsequent precise cleaning. After pre-cleaning, the three-dimensional model is updated, and clustering is performed based on the pixel value of the three-dimensional image in the three-dimensional model to obtain multiple target clusters, and the sub-images corresponding to the target clusters are input into the classification model to obtain the state type and cleaning information of each sub-area, so that the system can identify and select the appropriate cleaning agent and cleaning pressure, which improves the intelligent level of cleaning. In addition, the system can also select the state type and cleaning information of each sub-area according to the characteristics of the cleaning agent and the state of other sub-areas through which the flow path passes. The cleaning path of the cleaning nozzle is adjusted according to the conflict between the state types, which effectively prevents the cleaning agent in a specific area from causing damage to other sub-areas, thereby improving the safety of cleaning. The airflow information of the target object during breathing is monitored in real time by the detection unit, and the correction information of the cleaning unit is obtained based on this information and the three-dimensional model, and the control information of the cleaning unit is corrected to make the cleaning more precise, further improving the cleaning effect and user comfort. Through the mutual cooperation of the above-mentioned technical solutions, the present invention realizes precise control and optimization of the nasal cleaning process through intelligent image processing and cleaning path control, improves the cleaning efficiency and user comfort, and reduces the potential damage to the nasal mucosa, demonstrating its innovation and practicality in the field of nasal cleaning. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0034] Figure 1 Flow chart of the intelligent nasal cleaning pressure pump control method based on image processing in an embodiment of the present invention;

[0035] Figure 2 Schematic diagram of the structure of the intelligent nasal cleaning pressure pump control system based on image processing in an embodiment of the present invention. DETAILED DESCRIPTION

[0036] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0037] For ease of understanding, the specific process of the embodiment of the present application is described below. Figure 1 As shown, an embodiment of the intelligent nasal cleaning pressure pump control method based on image processing in the embodiment of the present application, wherein a camera unit is provided on the cleaning nozzle of the cleaning pressure pump, and the method comprises:

[0038] Step S1: acquiring first images at multiple angles and positions within a target object through the camera unit, and creating a three-dimensional model of the target object based on the first images;

[0039] Specifically, the first images of the target object, i.e., the inside of the user's nasal cavity, at multiple different angles and positions are acquired through the camera unit on the cleaning pressure pump, and a three-dimensional model of the target object is constructed based on the first image, so that the target object can be pre-cleaned and precisely cleaned according to the three-dimensional model.

[0040] Step S2: dividing the target object into a plurality of grids based on the three-dimensional model, obtaining a pre-cleaning position according to the pre-cleaning pressure corresponding to the grid and the characteristics of the cleaning nozzle, and pre-cleaning the target object at the pre-cleaning pressure and the pre-cleaning position corresponding to each grid;

[0041] Specifically, when cleaning the nasal cavity, since there may be multiple mixed states in the nasal cavity, and the above states may be accompanied by the generation of secretions, since the secretions are attached to the nasal mucosa, the true state of the above nasal mucosa cannot be obtained, and in order to accurately determine the distribution position of each of the above states, it is necessary to perform the above-mentioned pre-cleaning on the secretions in the nasal cavity, and divide the above-mentioned three-dimensional model into multiple grids according to the single spraying area of ​​the cleaning nozzle of the nasal pressure pump, and obtain the normal vector at the center position of each of the above grids. When cleaning the secretions in the nasal cavity, the same pre-cleaning pressure is used to clean different positions of the entire nasal cavity, and the pre-cleaning position of the cleaning nozzle on the normal vector corresponding to the above grid is obtained through the above-mentioned pre-cleaning pressure and the characteristics of the above-mentioned cleaning nozzle, and the above-mentioned pre-cleaning is performed on the above-mentioned target object based on the above-mentioned pre-cleaning pressure and the above-mentioned cleaning position. Through the above-mentioned technical solution, the cleaning of the secretions in the target object can be completed, laying a foundation for further accurately obtaining the state type of different areas of the target object.

[0042] Step S3: after the pre-cleaning, the three-dimensional model is updated, the three-dimensional image in the three-dimensional model is clustered according to the pixel value of each pixel, and a plurality of target clusters are obtained, and the sub-image corresponding to each of the target clusters is input into the classification model, and the state type and cleaning information of the sub-region corresponding to each of the sub-images are obtained;

[0043] Specifically, after the target object has undergone the pre-cleaning, the state of each position in the nasal cavity can be clearly seen, and the three-dimensional image corresponding to the updated three-dimensional model can be obtained. The three-dimensional image is also clustered according to the pixel values ​​of the pixels, and multiple target clustering clusters are obtained. Since the target clustering clusters may contain continuous pixels and discrete pixels, the purpose of clustering is to obtain the same state type area, and each of the sub-images is input into the classification model to obtain the state type of the sub-area corresponding to the sub-image and the cleaning agent and cleaning pressure corresponding to the state type. Through the above technical solution, not only the state type of each of the sub-areas can be accurately identified, but also the cleaning information corresponding to the sub-area can be obtained, laying the foundation for accurately cleaning each of the sub-areas and improving the nasal environment.

[0044] Step S4: each of the sub-areas is divided into a plurality of cleaning units, and control information and a flow path of the corresponding first cleaning agent when the cleaning nozzle cleans each of the cleaning units are simulated through a three-dimensional model according to the cleaning information of the sub-areas, the cleaning units are cleaned based on the control information, and the cleaning path of the cleaning nozzle is adjusted according to the conflict between the characteristics of the first cleaning agent and the state types of other sub-areas along the flow path;

[0045] Specifically, each of the above-mentioned sub-areas is divided into a plurality of the above-mentioned cleaning units through the spraying area of ​​the above-mentioned cleaning nozzle, and the above-mentioned cleaning units are cleaned one by one during cleaning. In the above-mentioned three-dimensional model, a cleaning simulation is performed on each of the above-mentioned cleaning units based on the cleaning information of the above-mentioned sub-area where each of the above-mentioned cleaning units is located, and the above-mentioned cleaning control information corresponding to each of the above-mentioned cleaning units and the flow path of the first cleaning agent corresponding to the above-mentioned sub-area are obtained, and the above-mentioned cleaning units are also sorted according to the distance from the sub-flow path outside the above-mentioned sub-area. Since the state type of each sub-area is different, the corresponding cleaning agents are also different. The first cleaning agents corresponding to sub-areas of different state types may cause irritation to other sub-areas, which is not conducive to the improvement of the nasal environment. Therefore, based on the characteristics of the above-mentioned first cleaning agent corresponding to each of the above-mentioned sub-areas and whether the state type of the flow path of the cleaning unit in the above-mentioned sub-area through other sub-areas conflicts, for example: when the state of the sub-area is ulcer, the first cleaning agent is a medicine for improving ulcers, and the state type of other sub-areas of the above-mentioned sub-area may be allergy. At this time, if When the first cleaning agent corresponding to the above-mentioned sub-area flows through the above-mentioned other sub-areas, it may irritate the mucous membranes in the above-mentioned other sub-areas and cause discomfort to the user. Therefore, it is considered that the characteristics of the first cleaning agent in the above-mentioned sub-area conflict with the state type of the above-mentioned other sub-areas. At this time, when the above-mentioned cleaning units are cleaned one by one based on the control information of the cleaning nozzle corresponding to each of the above-mentioned cleaning units, when the distance between the real-time flow path of the above-mentioned first cleaning agent of any of the above-mentioned cleaning units and the above-mentioned conflicting sub-area is less than or equal to the above-mentioned set distance, the cleaning agent of the above-mentioned cleaning nozzle is switched, and the above-mentioned conflicting sub-area is cleaned in time through the control information of the cleaning unit in the above-mentioned conflicting sub-area, so as to avoid irritating the mucous membranes of the above-mentioned other sub-areas or causing damage. When there is no conflict, the above-mentioned first cleaning agent can flow normally through the above-mentioned other sub-areas. Through the above-mentioned technical solution, through the precise cleaning of the above-mentioned sub-areas, not only the above-mentioned sub-areas can be cleaned in a targeted manner, but also the above-mentioned sub-areas can be effectively prevented from being irritated or damaged when cleaning the above-mentioned sub-areas, thereby improving user comfort and improving cleaning efficiency.

[0046] Step S5: Repeat the sub-area cleaning method in step S4 to complete the cleaning of the target object.

[0047] Specifically, by using the cleaning method for the above-mentioned sub-areas in the above-mentioned step S4, the above-mentioned other sub-areas are cleaned until the above-mentioned target object is accurately cleaned. By pre-cleaning and accurately cleaning the above-mentioned sub-areas, not only the cleaning effect can be improved, but also the user comfort can be improved.

[0048] Furthermore, when cleaning the sub-area, the airflow information of the target object when breathing is obtained through the detection unit on the cleaning nozzle, and correction information of the cleaning unit is obtained based on the airflow information and the three-dimensional model, and the control information of the cleaning unit is corrected based on the correction information.

[0049] Specifically, since when the above-mentioned sub-area is cleaned, since there will be airflow passing through when the nasal cavity breathes, when the nasal cavity is cleaned by the cleaning nozzle of the cleaning pressure pump, the path of the first cleaning agent in the air will be changed, and it may not accurately fall on the position corresponding to the predetermined cleaning unit, thereby causing the above-mentioned first cleaning agent to stimulate the mucous membrane in the surrounding conflicting sub-area. Therefore, the airflow information of the above-mentioned nasal cavity during breathing, that is, the airflow size, airflow direction and corresponding time period distribution are detected by the detection unit on the above-mentioned cleaning nozzle. Before cleaning the cleaning unit in the above-mentioned sub-area, the airflow size and airflow direction corresponding to the airflow time period are simulated based on the control information of the above-mentioned cleaning nozzle for cleaning the above-mentioned cleaning unit through the above-mentioned three-dimensional model, and the correction information of the above-mentioned cleaning nozzle is obtained. The above-mentioned corrected control information is obtained based on the above-mentioned correction information, and the above-mentioned cleaning unit is cleaned according to the corrected control information. Through the above-mentioned technical solution, each of the above-mentioned cleaning units can be accurately cleaned, thereby improving the comfort of the user.

[0050] Furthermore, the step S2 includes: dividing the three-dimensional model into a plurality of grids based on the spraying area of ​​the cleaning nozzle, and obtaining the normal vector at the center position of each grid, obtaining the cleaning distance of the grid according to the pre-cleaning pressure corresponding to the grid and the characteristics of the cleaning nozzle, determining the pre-cleaning position of the cleaning nozzle on the normal vector according to the cleaning distance, and pre-cleaning the grid with the pre-cleaning pressure, wherein the characteristic of the cleaning nozzle is the correspondence between the cleaning distance of the cleaning nozzle from the target object and the cleaning pressure when the cleaning nozzle sprays the cleaning agent.

[0051] Specifically, when cleaning the nasal cavity, different cleaning agents and cleaning methods should be used according to the different states of each position in the nasal cavity to improve the nasal environment. However, since there may be multiple mixed states in the nasal cavity, such as allergies, ulcers or infections, and the above states may be accompanied by the production of secretions, in order to accurately determine the distribution position of each of the above states, it is necessary to perform the above-mentioned pre-cleaning of the secretions in the nasal cavity, divide the above-mentioned three-dimensional model into multiple grids according to the single spraying area of ​​the cleaning nozzle of the nasal pressure pump, and obtain the normal vector at the center position of each of the above grids. When cleaning the secretions in the nasal cavity, the same pre-cleaning pressure is used to clean different positions of the entire nasal cavity. The cleaning agent is pure water at a set temperature. Because different cleaning agents are used when the cleaning nozzle sprays the cleaning agent, The spraying position corresponding to the pressure is also different. Therefore, the pre-cleaning position of the cleaning nozzle on the normal vector corresponding to the above-mentioned grid is obtained through the above-mentioned pre-cleaning pressure and the characteristics of the above-mentioned cleaning nozzle. The characteristics of the above-mentioned cleaning nozzle are the correspondence between the distance between the nozzle and the target object and the cleaning pressure when the above-mentioned cleaning nozzle is in operation, and the above-mentioned pre-cleaning is performed on the above-mentioned target object based on the above-mentioned pre-cleaning pressure and the above-mentioned pre-cleaning position, wherein the area of ​​each of the above-mentioned grids is less than or equal to the spraying area of ​​the above-mentioned cleaning nozzle, thereby ensuring that each position of the above-mentioned target object can be cleaned, the accuracy requirement of the above-mentioned pre-cleaning is not high, and the pre-cleaning pressure is small, and will not cause damage to the target object. Through the above-mentioned technical scheme, the cleaning of secretions in the target object can be completed, laying the foundation for further accurate acquisition of the state type of different areas of the target object.

[0052] Furthermore, the step S3 includes: after the target object is pre-cleaned, re-obtaining a second image of the target object at different positions, and updating the three-dimensional model through the second image, clustering the three-dimensional image corresponding to the three-dimensional model according to the pixel values ​​of the pixels, and obtaining multiple target clustering clusters with different pixel values, and also taking the pixel area corresponding to the number of continuous pixels greater than or equal to N in each of the target clustering clusters as the sub-image, and inputting the sub-image into the classification model, obtaining the state type and cleaning information of the sub-area of ​​the target object corresponding to the sub-image, wherein the state type is the state type of the nasal mucosa of the sub-area, and the cleaning information includes the cleaning agent and cleaning pressure of the sub-area.

[0053] Specifically, after the target object has undergone the pre-cleaning, the state of each position in the nasal cavity can be clearly seen, the second image in the nasal cavity is reacquired through the camera unit on the cleaning nozzle, and the three-dimensional model is updated based on the second image, and the three-dimensional image corresponding to the updated three-dimensional model is acquired, and the three-dimensional image is clustered according to the pixel values ​​of the pixels, and a plurality of target clustering clusters are acquired. Since the target clustering clusters may contain continuous pixels and discrete pixels, and the purpose of clustering is to acquire areas of the same state type, the discrete pixels in the target clustering clusters are excluded, and the areas with a large number of continuous pixels in each of the clustering clusters are clustered. A pixel area of ​​N is taken as the above-mentioned sub-image, wherein the value of N is a positive integer greater than or equal to 10, and each of the above-mentioned sub-images is input into the above-mentioned classification model to obtain the state type of the sub-area corresponding to the above-mentioned sub-image and the cleaning agent and cleaning pressure corresponding to the above-mentioned state type, wherein the above-mentioned state type is the state type of the above-mentioned nasal mucosa, for example: allergy, ulcer or infection, and the above-mentioned cleaning information includes the cleaning agent and cleaning pressure corresponding to the above-mentioned sub-area. Through the above-mentioned technical scheme, not only the state type of each of the above-mentioned sub-areas can be accurately identified, but also the cleaning information corresponding to the above-mentioned sub-areas can be obtained, thereby laying a foundation for targeted cleaning of each of the above-mentioned sub-areas and thus improving the nasal environment.

[0054] Further, the step S4 comprises:

[0055] Divide each of the sub-regions into a plurality of cleaning units according to the spraying area of ​​the cleaning nozzle, and simulate the control information of the cleaning nozzle when cleaning each of the cleaning units and the flow path of the corresponding first cleaning agent through the three-dimensional model according to the cleaning information corresponding to the sub-region, wherein the control information includes the cleaning direction, the cleaning position and the cleaning duration, and sort the plurality of cleaning units according to the distance from the flow path outside the sub-region, and obtain a cleaning list, wherein the farther the distance from the flow path outside the sub-region, the higher the sorting, and judge whether the characteristics of the first cleaning agent conflict with the state type of other sub-regions through which the flow path corresponding to each cleaning unit passes based on the characteristics of the first cleaning agent in the cleaning information of each sub-region;

[0056] When there is no conflict, the cleaning nozzle performs normal cleaning based on the control information corresponding to each cleaning unit in the cleaning list; when the characteristics of the first cleaning agent conflict with the state type in the other sub-areas, the other sub-areas are used as conflicting sub-areas, and when the cleaning nozzle cleans the cleaning units based on the control information corresponding to each cleaning unit in the cleaning list, when the distance between the real-time flow path of the first cleaning agent of any cleaning unit and the conflicting sub-area is less than or equal to the set distance, the first cleaning agent is switched to the second cleaning agent and the conflicting sub-area is cleaned based on the control information corresponding to the cleaning unit in the conflicting sub-area until the first cleaning agent is flushed out of the conflicting sub-area; repeat this step to complete the cleaning of the sub-area.

[0057] Specifically, since the spraying area of ​​the cleaning nozzle is small, partial cleaning is performed when the sub-area is cleaned. Therefore, each of the sub-areas is divided into a plurality of the cleaning units according to the spraying area of ​​the cleaning nozzle, and the cleaning units are cleaned one by one during cleaning. In order to obtain the cleaning position and cleaning angle of each of the cleaning units, a cleaning simulation is performed on each of the cleaning units based on the cleaning information of the sub-area where each of the cleaning units is located in the three-dimensional model, and the cleaning control information corresponding to each of the cleaning units and the flow path of the first cleaning agent corresponding to the sub-area are obtained, wherein the cleaning information passing through the sub-area includes the first cleaning agent, that is, the cleaning angle, the cleaning position and The cleaning duration is determined by the cleaning units being sorted according to the distance from the sub-flow path outside the sub-region, and the cleaning list is obtained. The farther the distance from the sub-flow path, the higher the sorting. When cleaning is performed according to the list, the cleaning efficiency can be improved. Since the state type of each sub-region is different, the corresponding cleaning agent is also different. The cleaning agents corresponding to sub-regions of different state types may irritate other sub-regions, which is not conducive to the improvement of the nasal environment. Therefore, based on whether the characteristics of the first cleaning agent corresponding to each of the above sub-regions conflict with the state type of the flow path of the cleaning unit in the above sub-region passing through other sub-regions, the characteristics of the first cleaning agent are the components in the cleaning agent. characteristics, for example: when the state of the sub-area is ulcer, the cleaning agent is a medicine for improving ulcers, and the state type of other sub-areas of the above sub-area may be allergy. At this time, if the first cleaning agent corresponding to the above sub-area flows through the above other sub-areas, it may stimulate the mucous membrane in the above other sub-areas and cause discomfort to the user. Therefore, it is considered that the characteristics of the first cleaning agent in the above sub-area conflict with the state type of the above other sub-areas. At this time, when the above cleaning units are cleaned one by one based on the control information of the cleaning nozzle corresponding to each of the above cleaning units, when the distance between the real-time flow path of the above-mentioned first cleaning agent of any of the above cleaning units and the above-mentioned first sub-area is less than or equal to the above-mentioned set distance, wherein the real-time flow path The third image taken in real time by the camera unit is acquired, and the cleaning agent type of the cleaning nozzle is switched, for example, the first cleaning agent is switched to low-concentration physiological saline, i.e., the second cleaning agent, and the conflicting sub-region is cleaned in time through the control information of the cleaning unit in the conflicting sub-region, that is, the first cleaning agent is diluted by spraying the second cleaning agent through the cleaning nozzle to avoid irritating the mucous membrane of the other sub-regions or causing damage. When there is no conflict, the first cleaning agent can flow normally through the other sub-regions. Through the above technical solution, it can be effectively prevented that the mucous membrane of the conflicting sub-region is irritated or damaged when cleaning the above sub-region, thereby improving user comfort and improving cleaning efficiency.

[0058] Furthermore, when cleaning the sub-area, the airflow information of the target object when breathing is also detected by the detection unit on the cleaning nozzle, and the airflow information includes the airflow size, airflow direction and corresponding airflow time period distribution;

[0059] Before cleaning the cleaning unit in the sub-area, a cleaning airflow time period corresponding to the cleaning unit is calculated according to the cleaning duration in the control information corresponding to the cleaning unit and the current airflow time period, and when the cleaning airflow time period is the second time period, the cleaning nozzle is normally cleaned based on the control information corresponding to the cleaning unit;

[0060] When the airflow time period is the first time period and the third time period, the airflow size and airflow direction corresponding to the airflow time period are simulated by the three-dimensional model based on the control information of the cleaning nozzle, and the correction information of the cleaning nozzle is obtained, and the control information is corrected based on the correction information, and the cleaning unit is cleaned according to the corrected control information, wherein the airflow time period includes the first time period, the second time period and the third time period, the first time period is the inhalation time period, the second time period is the gap time period, and the third time period is the exhalation time period.

[0061] Specifically, since when the above-mentioned sub-area is cleaned, since there will be airflow passing through when the nasal cavity breathes, when the nasal cavity is cleaned by the cleaning nozzle of the cleaning pressure pump, the path of the first cleaning agent in the air will be changed, and it may not accurately fall on the position corresponding to the predetermined cleaning unit, thereby causing the above-mentioned first cleaning agent to stimulate the mucous membrane in the surrounding conflicting sub-area. Therefore, the airflow information of the above-mentioned nasal cavity during breathing, that is, the airflow size, airflow direction and corresponding time period distribution, is detected by the detection unit on the above-mentioned cleaning nozzle. Before cleaning the cleaning unit in the above-mentioned sub-area, the above-mentioned cleaning airflow time period corresponding to the above-mentioned cleaning unit is obtained according to the above-mentioned cleaning duration and the current airflow time period in the above-mentioned control information corresponding to the above-mentioned cleaning unit, wherein the above-mentioned cleaning airflow time period only corresponds to one of the above-mentioned airflow time periods, and in the above-mentioned cleaning airflow When the time period is the second time period, that is, the gap time period, there is no airflow passing through at this time. Therefore, the nozzle of the cleaning pressure pump cleans the cleaning unit with the above-mentioned control information. When the airflow time period is the first time period or the third time period, that is, there is airflow passing through, in order to enable the first cleaning agent to accurately fall into the cleaning unit, the airflow size and airflow direction corresponding to the airflow time period are simulated on the basis of the control information of the cleaning nozzle cleaning the cleaning unit through the three-dimensional model, and the correction information of the cleaning nozzle is obtained. The corrected control information is obtained based on the correction information, and the cleaning unit is cleaned according to the corrected control information. Through the above-mentioned technical solution, each of the above-mentioned cleaning units can be accurately cleaned, thereby improving the comfort of users.

[0062] Furthermore, the classification model is a machine learning model trained with historical sample data, wherein the historical sample data includes nasal sample images and corresponding label information, and the label information includes the nasal state type and cleaning information corresponding to the nasal sample images.

[0063] Furthermore, the nasal cleaning pressure pump can switch between the first cleaning agent and the second cleaning agent by controlling the cleaning agent switching port.

[0064] The present invention also provides an intelligent nasal cleaning pressure pump control system based on image processing, the system is used to implement the above method, such as Figure 2 As shown, the system comprises:

[0065] A camera unit, used to obtain first images at different angles and positions within a target object;

[0066] a model creation unit, configured to create a three-dimensional model of the target object based on the first image;

[0067] A pre-cleaning unit, configured to divide the target object into a plurality of grids based on the three-dimensional model, obtain a pre-cleaning position according to a pre-cleaning pressure corresponding to the grids and a characteristic of a cleaning nozzle, and pre-clean the target object at the pre-cleaning pressure and the pre-cleaning position corresponding to each grid;

[0068] a cleaning information acquisition unit, configured to update the three-dimensional model after the pre-cleaning, cluster the three-dimensional image in the three-dimensional model according to the pixel value of each pixel, and obtain a plurality of target clusters, input the sub-image corresponding to each of the target clusters into the classification model, and obtain the state type and cleaning information of the sub-region corresponding to each of the sub-images;

[0069] The sub-area cleaning unit is configured as follows: each of the sub-areas is divided into a plurality of cleaning units, and according to the cleaning information of the sub-areas, control information and a flow path of a corresponding first cleaning agent when the cleaning nozzle cleans each of the cleaning units are simulated through a three-dimensional model, the cleaning unit is cleaned based on the control information, and the cleaning path of the cleaning nozzle is adjusted according to a conflict between the characteristics of the first cleaning agent and the state types of other sub-areas along the flow path; the cleaning method of the sub-areas is repeated to complete the cleaning of the target object.

[0070] The present invention also provides a computer-readable storage medium, on which instructions are stored, and wherein the instructions implement the above method when executed by a processor.

[0071] In summary, the present invention significantly improves the cleaning efficiency and user comfort through intelligent technology. First, a multi-angle first image of the inside of the nasal cavity is obtained through a camera unit, and a three-dimensional model is created based on the first image, so that the cleaning process can accurately target the specific structure in the nasal cavity, effectively avoiding the blind spots that may exist in traditional cleaning methods, and improving the accuracy and efficiency of cleaning. In the pre-cleaning step, the system divides the nasal cavity into multiple grids according to the three-dimensional model, and determines the pre-cleaning position according to the pre-cleaning pressure corresponding to each grid and the characteristics of the cleaning nozzle, so as to perform a preliminary cleaning of the nasal cavity, which not only cleans the secretions, but also provides a basis for subsequent precise cleaning. After pre-cleaning, the three-dimensional model is updated, and the three-dimensional image in the three-dimensional model is clustered according to the pixel value to obtain multiple target clustering clusters, and the sub-images corresponding to the target clustering clusters are input into the classification model to obtain the state type and cleaning information of each sub-area, so that the system can identify and select the appropriate cleaning agent and cleaning pressure, which improves the intelligent level of cleaning. In addition, the system can also select the cleaning agent according to the characteristics of the cleaning agent and the other sub-areas through which the flow path passes. The cleaning path of the cleaning nozzle is adjusted according to the conflict between state types, which effectively prevents the cleaning agent in a specific area from causing damage to other sub-areas, thereby improving the safety of cleaning. The airflow information of the target object during breathing is monitored in real time by the detection unit, and the correction information of the cleaning unit is obtained based on this information and the three-dimensional model. The control information of the cleaning unit is corrected to make the cleaning more precise, further improving the cleaning effect and user comfort. Through the mutual cooperation of the above-mentioned technical solutions, the present invention realizes precise control and optimization of the nasal cleaning process through intelligent image processing and cleaning path control, improves the cleaning efficiency and user comfort, and reduces the potential damage to the nasal mucosa, demonstrating its innovation and practicality in the field of nasal cleaning.

[0072] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0073] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program codes.

[0074] As described above, the above embodiments 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, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these 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.

Claims

1. An intelligent nasal cleaning pressure pump control method based on image processing, characterized in that: A camera unit is provided on the cleaning nozzle of the cleaning pressure pump, and the method comprises: Step S1: acquiring first images at multiple angles and positions within a target object through the camera unit, and creating a three-dimensional model of the target object based on the first images; Step S2: dividing the target object into a plurality of grids based on the three-dimensional model, obtaining a pre-cleaning position according to the pre-cleaning pressure corresponding to the grid and the characteristics of the cleaning nozzle, and pre-cleaning the target object at the pre-cleaning pressure and the pre-cleaning position corresponding to each grid; Step S3: after the pre-cleaning, the three-dimensional model is updated, the three-dimensional image in the three-dimensional model is clustered according to the pixel value of each pixel, and a plurality of target clusters are obtained, and the sub-image corresponding to each of the target clusters is input into the classification model, and the state type and cleaning information of the sub-region corresponding to each of the sub-images are obtained; Step S4: each of the sub-areas is divided into a plurality of cleaning units, and control information and a flow path of the corresponding first cleaning agent when the cleaning nozzle cleans each of the cleaning units are simulated through a three-dimensional model according to the cleaning information of the sub-areas, the cleaning units are cleaned based on the control information, and the cleaning path of the cleaning nozzle is adjusted according to the conflict between the characteristics of the first cleaning agent and the state types of other sub-areas through which the flow path passes; Step S5: Repeat the sub-area cleaning method in step S4 to complete the cleaning of the target object.

2. The method according to claim 1, characterized in that , the step S4 also includes: When cleaning the sub-area, the airflow information of the target object when breathing is also obtained through the detection unit on the cleaning nozzle, and correction information of the cleaning unit is obtained based on the airflow information and the three-dimensional model, and the control information of the cleaning unit is corrected based on the correction information.

3. The method according to claim 1, characterized in that , the step S2 comprises: The three-dimensional model is divided into a plurality of grids based on the spraying area of ​​the cleaning nozzle, and the normal vector at the center position of each grid is obtained, the cleaning distance of the grid is obtained according to the pre-cleaning pressure corresponding to the grid and the characteristics of the cleaning nozzle, the cleaning position of the cleaning nozzle on the normal vector is determined according to the cleaning distance, and the grid is pre-cleaned with the pre-cleaning pressure, wherein the characteristic of the cleaning nozzle is the correspondence between the cleaning distance of the cleaning nozzle from the target object and the cleaning pressure when the cleaning nozzle sprays the cleaning agent.

4. The method according to claim 1, characterized in that , the step S3 comprises: After the target object has been pre-cleaned, a second image of the target object at different positions is reacquired, and the three-dimensional model is updated through the second image. The three-dimensional image corresponding to the three-dimensional model is clustered according to pixel values, and a plurality of target clustering clusters with different pixel values ​​are acquired. A pixel area corresponding to each of the target clustering clusters and the number of continuous pixels therein is greater than or equal to N is taken as the sub-image, and the sub-image is input into the classification model to acquire a state type and cleaning information of a sub-area of ​​the target object corresponding to the sub-image, wherein the state type is the state type of the nasal mucosa of the sub-area, and the cleaning information includes a cleaning agent and a cleaning pressure of the sub-area.

5. The method according to claim 1, characterized in that , the step S4 comprises: Divide each of the sub-regions into a plurality of cleaning units according to the spraying area of ​​the cleaning nozzle, and simulate the control information of the cleaning nozzle when cleaning each of the cleaning units and the flow path of the corresponding first cleaning agent through the three-dimensional model according to the cleaning information corresponding to the sub-region, wherein the control information includes the cleaning direction, the cleaning position and the cleaning duration, and sort the plurality of cleaning units according to the distance from the flow path outside the sub-region, and obtain a cleaning list, wherein the farther the distance from the flow path outside the sub-region, the higher the sorting, and judge whether the characteristics of the first cleaning agent conflict with the state type of other sub-regions through which the flow path corresponding to each cleaning unit passes based on the characteristics of the first cleaning agent in the cleaning information of each sub-region; When there is no conflict, the cleaning nozzle performs normal cleaning based on the control information corresponding to each cleaning unit in the cleaning list; when the characteristics of the first cleaning agent conflict with the state type in the other sub-areas, the other sub-areas are used as conflicting sub-areas, and when the cleaning nozzle cleans the cleaning units based on the control information corresponding to each cleaning unit in the cleaning list, when the distance between the real-time flow path of the first cleaning agent of any cleaning unit and the conflicting sub-area is less than or equal to the set distance, the first cleaning agent is switched to the second cleaning agent and the conflicting sub-area is cleaned based on the control information corresponding to the cleaning unit in the conflicting sub-area until the first cleaning agent is flushed out of the conflicting sub-area; repeat this step to complete the cleaning of the sub-area.

6. The method according to claim 2, characterized in that , the step S4 also includes: When cleaning the sub-area, the detection unit on the cleaning nozzle is also used to detect the airflow information of the target object when breathing, and the airflow information includes the airflow size, airflow direction and corresponding airflow time period distribution; Before cleaning the cleaning unit in the sub-area, a cleaning airflow time period corresponding to the cleaning unit is calculated according to the cleaning duration in the control information corresponding to the cleaning unit and the current airflow time period, and when the cleaning airflow time period is the second time period, the cleaning nozzle is normally cleaned based on the control information corresponding to the cleaning unit; When the airflow time period is the first time period and the third time period, the airflow size and airflow direction corresponding to the airflow time period are simulated by the three-dimensional model based on the control information of the cleaning nozzle, and the correction information of the cleaning nozzle is obtained, and the control information is corrected based on the correction information, and the cleaning unit is cleaned according to the corrected control information, wherein the airflow time period includes the first time period, the second time period and the third time period, the first time period is the inhalation time period, the second time period is the gap time period, and the third time period is the exhalation time period.

7. The method according to claim 1, characterized in that The classification model is a machine learning model trained with historical sample data, wherein the historical sample data includes nasal sample images and corresponding label information, and the label information includes the nasal state type and cleaning information corresponding to the nasal sample image.

8. The method according to claim 1, characterized in that ,The nasal cleaning pressure pump can switch between the first cleaning agent and the second cleaning agent by controlling the cleaning agent switching port.

9. An intelligent nasal cleaning pressure pump control system based on image processing, the system is used to implement the method according to any one of claims 1 to 8, characterized in that: The system comprises: A camera unit, used to obtain first images at different angles and positions within a target object; a model creation unit, configured to create a three-dimensional model of the target object based on the first image; A pre-cleaning unit, configured to divide the target object into a plurality of grids based on the three-dimensional model, obtain a pre-cleaning position according to a pre-cleaning pressure corresponding to the grids and a characteristic of a cleaning nozzle, and pre-clean the target object at the pre-cleaning pressure and the pre-cleaning position corresponding to each grid; a cleaning information acquisition unit, configured to update the three-dimensional model after the pre-cleaning, cluster the three-dimensional image in the three-dimensional model according to the pixel value of each pixel, and obtain a plurality of target clusters, input the sub-image corresponding to each of the target clusters into the classification model, and obtain the state type and cleaning information of the sub-region corresponding to each of the sub-images; The sub-area cleaning unit is configured as follows: each of the sub-areas is divided into a plurality of cleaning units, and according to the cleaning information of the sub-areas, control information and a flow path of a corresponding first cleaning agent when the cleaning nozzle cleans each of the cleaning units are simulated through a three-dimensional model, the cleaning unit is cleaned based on the control information, and the cleaning path of the cleaning nozzle is adjusted according to a conflict between the characteristics of the first cleaning agent and the state types of other sub-areas through which the flow path passes; the cleaning method of the sub-areas is repeated to complete the cleaning of the target object.

10. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by a processor, the method according to any one of claims 1 to 8 is implemented.

Images