An image recognition-based oral medication application device

Abstract

This invention discloses an image recognition-based oral medication application device. The device includes a medication applicator with a wireless data link and a display device. The medication applicator includes a handle with a first wireless transmission module and a probe mounted on the handle, which has a light source and a camera. The display device includes a second wireless transmission module, a processor, and a display screen. The probe further includes a medication delivery part for applying medication to the target affected area. The first wireless transmission module wirelessly transmits images acquired by the camera to the second wireless transmission module. The processor processes the images acquired by the second wireless transmission module, identifies the target affected area based on the processed image, and provides medication application prompts. This invention automatically identifies the target affected area and issues medication application prompts, greatly facilitating self-medication and improving the user experience.

Description

Technical Field

[0001] This invention relates to the field of medical device technology, and more specifically to an oral medication application device based on image recognition. Background Technology

[0002] Oral ulcers are a widespread and prevalent oral disease, with no significant seasonal correlation. They can occur in children, young adults, and the elderly. Current treatments for oral ulcers typically involve oral medication and oral administration. Oral medication usually requires using a cotton swab or other object to depress the tongue and then using a flashlight to examine the patient's oral cavity before applying medication. This process is extremely inconvenient for patients to perform themselves. While visual oral medication applicators combined with mobile devices have been developed for visualized application, existing devices still require the operator to carefully determine the appropriate application site, making quick and accurate application difficult. Furthermore, discrepancies between the image angle and the operator's position on the device result in a poor user experience. Summary of the Invention

[0003] The present invention aims to provide an image recognition-based oral medication application device to solve the problem of poor self-medication experience for oral patients.

[0004] In a first aspect, the present invention provides an oral medication application device based on image recognition, comprising:

[0005] A wireless data link medication applicator and display device; the medication applicator includes a handle with a first wireless transmission module and a probe mounted on the handle and having a light source and a camera, and the display device includes a second wireless transmission module, a processor, and a display screen; wherein, the probe further includes a drug delivery part for applying medication to the target affected area, the first wireless transmission module is used to wirelessly transmit images acquired by the camera to the second wireless transmission module, and the processor is used to recognize and process the images acquired by the second wireless transmission module, and identify the target affected area based on the processed image to provide medication application prompts.

[0006] Optionally, the medication applicator also includes a distance sensor, which is used to obtain the distance range between the probe and the front obstacle. The medication application prompt information includes marking the identified affected area with a symbol and issuing a prompt sound when the distance range is within a preset range.

[0007] Optionally, the medication application prompt information includes marking the identified affected area with a symbol and emitting a prompt sound within a preset range, including:

[0008] Step 1: Establish an empirical database of oral ulcer images, set image pixel settings, select a class model with similar pattern abstract features for the images to be identified, and preprocess the images to be identified. The preprocessing includes setting the component R=G=B for each pixel of the image to be identified, thereby obtaining the grayscale feature vector of the image to be identified. ;

[0009] Step 2: Make

[0010] and ,

[0011] in yes The transpose of , For the vector of the class model, N is the dimension of the vector of the class model, and the adjoint vector is obtained. Make it satisfy the following formula

[0012] ,in Represents the adjoint vector correspond The eigenvectors in the vector, k' is The index corresponding to the internal feature vector, k=1……M, where M is the number of the class models, M≤N;

[0013] Step 3: Input the initial value q(0) of the feature vector of the model to be tested in the image to be recognized, and use the formula Find the ordered parameters initial value ;

[0014] Step 4: Using parameters Establish the following Iterative formula:

[0015] ,in The model is selected based on the similarity between the class model and the vector of the model to be tested. b and c are coefficients preset based on experience, and d satisfies: ;

[0016] Step 5: Establish a three-layer neural network model. The first to third layers of the neural network model are the input layer, the second layer, and the output layer, respectively. Each receiving unit in the input layer receives the component q(0) of the feature vector q(0) of the model to be recognized. j (0); The second layer of the neural network model consists of various parameters. Neuron, in which parameters It is each q j (0) multiplied by q j(0) Connected The sum of the products, where adjoint vector The j-th component; the second layer from the formula in step 4 (0) Start iterative running, and in If the result is ≠0, return to step 4 and run the process again until... When =0, proceed to step 6 to obtain the final image recognition result from the output layer;

[0017] Step 6: Obtain the final image recognition result according to the following formula:

[0018] ,in For the elements of output layer unit j, For parameters The t-th vector, t=0,1,2,……T, where T is The total number of vectors, where M is the number of class models;

[0019] The identified The image is input into an oral ulcer image experience database, and compared with the experience data in the database to determine whether the image to be identified is the target lesion.

[0020] Optionally, the medication prompt information includes displaying the required medication dosage based on the preset medication dosage and / or the previous patient's condition. The required medication dosage includes the amount of medication to be dispensed each time and the current remaining number of dispensing sessions.

[0021] Optionally, the drug delivery portion of the probe is a spoon-shaped component and / or a hollow portion. When it is a hollow portion, it works in conjunction with the drug storage portion and pressure generation module inside the handle to deliver the drug to the target affected area.

[0022] Optionally, the preset range is obtained based on the infrared sensor and the processor.

[0023] Optionally, the infrared sensor is integrated into the camera or located within the probe.

[0024] Optionally, the processor is an FPGA or an MCU.

[0025] In this invention, by processing the acquired oral cavity image to obtain the affected area, i.e., the target medication application location information, medication application prompts can be provided based on this information. This avoids inconvenience caused by directional operation due to the angle deviation between the handle and the image displayed on the display device during user operation, as well as inconvenience caused by visual impairment or lack of professional judgment leading to the inability to promptly locate the target medication application location. Furthermore, the prompts provided by this invention include visual prompts (symbols marking the affected area) and auditory prompts (sound alerts), effectively improving the accuracy of medication application when the operator applies medication independently. To broaden the applicability of the medication application device, the probe of this invention can be designed as a spoon shape and / or a hollow part, making this invention applicable to solid or liquid medications. The affected area localization of this invention combines a neural network model for image processing, solving the shortcomings of existing medication application devices in obtaining affected area information, and addressing the problems of long training times and insufficient accuracy of existing image recognition models. Moreover, the distance acquisition of this invention incorporates temperature information, achieving precise positioning of the affected area, more efficiently acquiring affected area information, and better improving the user experience. Attached Figure Description

[0026] The invention will be further understood from the following description taken in conjunction with the accompanying drawings. The components in the drawings are not necessarily drawn to scale, but rather the focus is on illustrating the principles of the embodiments.

[0027] Figure 1 A structural diagram of an oral medication application device system based on image recognition provided by the present invention;

[0028] Figure 2 This is a diagram of a three-layer neural network structure provided in Embodiment 1 of the present invention;

[0029] Figure 3 This is a display effect diagram after obtaining the target affected area, provided by the present invention;

[0030] Figure 4 This is one circuit connection diagram provided by the present invention;

[0031] Figure 5 This is a schematic diagram of the image recognition method provided in Embodiment 1 of the present invention. Detailed Implementation

[0032] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to its embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the present invention. Other systems, methods, and / or features of this embodiment will become apparent to those skilled in the art after reviewing the following detailed description. All such additional systems, methods, features, and advantages are intended to be included within this specification, within the scope of the invention, and protected by the appended claims. Further features of the disclosed embodiments are described in the following detailed description, and these features will become apparent from the following detailed description.

[0033] In the accompanying drawings of the embodiments of the present invention, the same or similar reference numerals correspond to the same or similar components. In the description of the present invention, it should be understood that if terms such as "upper," "lower," "left," "right," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, they are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or component referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, the terms used to describe positional relationships in the drawings are only for illustrative purposes and should not be construed as limiting the present patent. For those skilled in the art, the specific meaning of the above terms can be understood according to the specific circumstances.

[0034] Example 1: Figure 1This invention provides an image recognition-based oral medication application device. By incorporating image recognition of the affected area, this embodiment facilitates quick and easy self-application of medication by the user. The oral medication application device includes a wirelessly linked applicator and a display device. The applicator includes a handle with a first wireless transmission module and a probe mounted on the handle, which includes a light source and a camera. The display device includes a second wireless transmission module, a processor, and a display screen. The probe further includes a medication delivery part for applying medication to the target affected area. The first wireless transmission module wirelessly transmits images acquired by the camera to the second wireless transmission module. The processor processes the images acquired by the second wireless transmission module and identifies the target affected area based on the processed image, providing medication application prompts. The handle also includes a microcontroller, a speaker, and a user operation switch (optional). The light source, camera, first wireless transmission module, speaker, and user operation switch are all connected to the microcontroller. The microcontroller processes all data within the applicator, and the speaker emits the medication application prompts. Optionally, the speaker is not located within the probe but integrated with the display device, such as a mobile phone, tablet, or computer. The notification information can be delivered via the speaker on the mobile phone, tablet, or computer. The first wireless transmission module and the second wireless transmission module are wirelessly connected. Both the display screen and the second wireless transmission module are connected to the processor. (See attached diagram) Figure 4 The diagram shows the circuit module connection of the speaker inside the handle. The left half of the diagram is the medicine applicator, and the right half is the display device.

[0035] The medication applicator also includes a distance sensor connected to a microcontroller. The distance sensor is used to acquire the distance range between the probe and a front-end obstacle. The medication application prompts include marking the identified affected area with a symbol and emitting an audible alert when the distance range is within a preset range. In this embodiment, the distance sensor can be an infrared sensor or an ultrasonic sensor.

[0036] The medication application prompt information includes marking the identified affected area with a symbol and emitting a prompt sound within a preset range, as shown in the attached figure. Figure 5 The process steps are as follows:

[0037] Step 1: Establish an empirical database of oral ulcer images. Set the image pixel size and segment the image acquired by the camera into images to be identified. Specifically, this can be done by segmenting the image into several images to be identified, or by using semantic segmentation to segment the image to obtain images to be identified, or by using spatial domain filtering and threshold segmentation algorithms to separate two regions based on the visual saliency map, and then combining morphological operations to obtain a triadic image. Gradient calculation is performed on each pixel in the unknown region based on the triadic image. Sample points are obtained for the current unknown region pixels based on the gradient direction and saliency magnitude. Then, the opacity and confidence of each sample point are calculated. The sample with the highest confidence is taken as the best sample pair for final matting. The opacity of the local area is then smoothed to obtain the final estimated opacity. Based on the final estimated opacity and the color value of the best sample pair, matting is performed on the original image to extract the images to be identified. A class model with abstract features of similar patterns is selected for the images to be identified. This class model reflects the common attributes of similar patterns. The images to be identified are preprocessed to make each... Pixels The components R=G=B, where R, G, and B represent the color brightness of the red, green, and blue channels, respectively, thereby obtaining the grayscale feature vector of the image to be recognized. Specifically, this can be done as follows: Extracting texture features from the image to be recognized, and extracting feature values ​​to form the feature vector; or obtaining the feature vector from the image to be recognized based on a convolutional neural network trained on samples; or starting from a preset pixel in the grayscale image, establishing a selection area in the grayscale image according to a preset size, judging whether the grayscale value of the pixel is less than the grayscale mean based on the grayscale mean and standard deviation value. If it is less, the feature value of the point is marked as 0; if it is greater, the feature value of the point is marked as non-zero. Concatenating the feature values ​​within the selection area by row / column, obtaining the feature vector of the current preset pixel; obtaining the next pixel along the image coordinate system direction, and repeating the above steps for obtaining the feature vector of the preset pixel based on the next pixel until the feature vector of the last pre-selected pixel is obtained; only retaining the feature vectors with a deviation value greater than a preset value W, and performing weighted processing on the retained feature vectors, concatenating all the weighted feature vectors to finally obtain the global feature vector, wherein the feature vector... Represented by pixel grayscale values;

[0038] Step 2: Make and ,

[0039] in yes The transpose of , For the vector of the class model, N is the dimension of the vector of the class model of the image to be identified, and the accompanying vector is obtained. Make it satisfy the following formula: ,in Represents the adjoint vector correspond The eigenvectors in the vector, k ’ for The index corresponding to the internal feature vector, k=1……M, where M is the number of the class models, M≤N;

[0040] Step 3: Input the initial value q(0) of the feature vector of the model to be tested in the image to be recognized, and use the formula Find the ordered parameters initial value ;

[0041] Step 4: Using parameters Establish the following Iterative formula:

[0042] ,

[0043] in The model is selected based on the similarity between the class model and the vector of the model to be tested. b and c are coefficients preset based on experience. Generally, b=c>0, but in specific applications, b=c=1 can be taken. d satisfies: ; This was obtained by those skilled in the art based on historical experience and extensive repeated experimental training, and will not be elaborated further here.

[0044] Step 5: Establish a three-layer neural network model. The first to third layers of the neural network model are the input layer, the second layer, and the output layer, respectively. Each receiving unit in the input layer receives the component q(0) of the feature vector q(0) of the model to be identified in the image to be recognized. j (0); The second layer of the neural network model consists of various parameters. Neuron, in which parameters It is each q j (0) multiplied by q j (0) Connected The sum of the products, where adjoint vector The j-th component; the second layer from the formula in step 4 (0) Start iterative running, and in If the result is ≠0, return to step 4 and run the process again until... When =0, proceed to step 6 to obtain the final image recognition result from the output layer;

[0045] Step 6: Obtain the final image recognition result according to the following formula:

[0046] ,

[0047] in Let t be the t-th element of the output layer unit j. For parameters The t-th vector, t=0,1,2,……T, where T is The total number of vectors, where M is the number of class models;

[0048] The identified The image is input into an oral ulcer image experience database, and compared with the experience data in the database to determine whether the image to be identified is an image of the target ulcer. Specifically, this can be done by... The feature vector of the m-th image in the database Calculate Euclidean distance Determine if the Euclidean distance is less than a threshold. If the image is smaller than the m-th image, it is determined that the image to be identified belongs to the same category as the m-th image. If the m-th image is an image of an oral ulcer, then the image to be identified is identified as the image of the target lesion. When the result is an image of the target lesion, the image recognition result is marked with a symbol. (See attached figure) Figure 3 It is a diagram after marking, where the symbols can also be circles, ovals or squares.

[0049] The medication application prompt information includes displaying the required medication amount based on the preset medication dosage and / or the previous patient's condition. The required medication amount includes the amount of medication to be applied each time and the current remaining number of applications.

[0050] The probe's drug delivery portion is a spoon-shaped component and / or a hollow portion. When it is hollow, it works in conjunction with the drug storage portion and pressure generating module within the handle to deliver medication to the target affected area. In this embodiment, when using the device, the user presses the switch after hearing a prompt and confirming the location according to the prompt information displayed on the display device. The handle then generates spray pressure to spray solid or liquid medication onto the target affected area. To reduce the cost of the medication application device, the switch can be omitted, and the probe can be a spoon-shaped portion. In this embodiment, when the probe is a spoon-shaped portion, the user can apply the medication originally placed in the spoon to the target affected area after hearing a prompt and confirming the location according to the prompt information displayed on the display device.

[0051] The preset range is obtained based on an infrared sensor or an ultrasonic sensor and the processor. The processor determines whether the range is within the preset range based on the range data obtained by the infrared sensor or the ultrasonic sensor. If it is, the processor sends a signal to the speaker to activate the medication application prompt.

[0052] The infrared sensor is integrated into the camera or located inside the probe.

[0053] The processor is either an FPGA or an MCU.

[0054] The image recognition method used in this embodiment not only enables existing medication application devices to automatically identify affected areas, but also improves the efficiency of oral lesion image recognition. It avoids problems such as noise interference, image defects, and angle transformation that exist in ordinary image recognition algorithms. The algorithm of this invention is simple and easy to develop and apply to microprocessors, which improves the feasibility of applying oral lesion image recognition in medication application devices.

[0055] Example 2: This example can be combined with Example 1 or implemented independently. This example achieves precise distance acquisition of the affected area, improving the user experience. The oral medication application device of this example includes a wirelessly linked medication applicator and a display device. The medication applicator includes a handle with a first wireless transmission module and a probe mounted on the handle, which has a light source and a camera. The display device includes a second wireless transmission module, a processor, and a display screen. The probe also includes a medication delivery part for applying medication to the target affected area. The first wireless transmission module wirelessly transmits the image acquired by the camera to the second wireless transmission module. The processor processes the image acquired by the second wireless transmission module, identifies the target affected area based on the processed image, and provides medication application prompts. The probe also includes a microcontroller, a speaker, and a user operation switch (optional). The light source, camera, first wireless transmission module, speaker, and user operation switch are all connected to the microcontroller. The microcontroller processes all data within the medication applicator, and the speaker emits the medication application prompts. The first wireless transmission module is wirelessly connected to the second wireless transmission module, and both the display screen and the second wireless transmission module are connected to the processor.

[0056] The medication applicator also includes a distance sensor connected to a microcontroller. The medication application prompts include marking the identified affected area with a symbol and emitting a prompt sound within a preset range. Since infrared imaging can effectively reflect the temperature difference between the background and the target, and the imaging varies depending on the location and condition of different parts of the body, the distance sensor in this embodiment can be an infrared sensor capable of acquiring infrared images. This sensor can be combined with a camera, utilizing an existing infrared camera technology.

[0057] The medication application prompt information includes marking the identified affected area with a symbol and issuing a prompt sound within a preset range, which includes: converting the acquired image into a grayscale image, identifying the pixel area with grayscale value within the threshold area in the grayscale image as the target area, and the pixel area with other grayscale value as the background area; calculating the centroid of the target area, and issuing a prompt sound if the centroid pixel meets the preset distance condition.

[0058] Specifically, pixels with gray values ​​within the threshold range [t1, t2] in the grayscale image are identified as the target region, while pixels with other gray values ​​are identified as the background region; the centroid of the target region is calculated using the formula: Where D is the range of the x and y coordinates of the target region; (X m Y m () represents the centroid coordinates of the target region D; The number of all pixels within the target region D; It is the sum of the x-coordinates of all pixels within the target region D; Let be the sum of the ordinates of all pixels D within the target region; assuming the acquired image size is X*Y, with its center pixel at the origin (X0, Y0), let the maximum deviation region be a rectangular region △x*△y centered at the origin. If the centroid pixel (X0, Y0) m Y m Satisfy: |X m -X0|≤ 、|Y m -Y0|≤ If the condition is met, a prompt sound will be emitted. The values ​​t1, t2, Δx, and Δy were obtained by those skilled in the art based on historical experience and extensive repeated experimental training, and will not be elaborated upon here.

[0059] The probe's drug delivery portion is a spoon-shaped component and / or a hollow portion. When it is hollow, it works in conjunction with the drug storage portion and pressure generating module within the handle to deliver medication to the target affected area. In this embodiment, when using the device, the user presses the switch after hearing a prompt and confirming the location according to the prompt information displayed on the display device. The handle then generates spray pressure to spray solid or liquid medication onto the target affected area. To reduce the cost of the medication application device, the switch can be omitted, and the probe can be a spoon-shaped portion. In this embodiment, when the probe is a spoon-shaped portion, the user can apply the medication originally placed in the spoon to the target affected area after hearing a prompt and confirming the location according to the prompt information displayed on the display device.

[0060] The preset range is obtained based on the infrared sensor and the processor. The processor determines whether the range is within the preset range based on the range data obtained by the infrared sensor. If it is, the processor sends a signal to the speaker to activate the medication application prompt.

[0061] The infrared sensor is integrated into the camera or located inside the probe.

[0062] The processor is either an FPGA or an MCU.

[0063] Since the infrared images of the affected area are significantly different from those of the normal area, this embodiment uses the image detected by the infrared sensor to locate the affected area. This allows users to quickly and easily obtain the location of the affected area and promptly alerts the user when the medication application site is within the effective medication application range, thereby improving medication application efficiency.

[0064] Example 3: This example can be combined with the previous examples or implemented independently. This example further defines the prompting information, making it easier for the operator to quickly locate and apply medication to the affected area when applying medication independently. The oral medication application device of this example includes a wirelessly linked medication applicator and a display device. The medication applicator includes a handle with a first wireless transmission module and a probe mounted on the handle, which has a light source and a camera. The display device includes a second wireless transmission module, a processor, and a display screen. The probe also includes a medication delivery part for applying medication to the target affected area. The first wireless transmission module wirelessly transmits the image acquired by the camera to the second wireless transmission module. The processor processes the image acquired by the second wireless transmission module, identifies the target affected area based on the processed image, and provides medication application prompts. The probe also includes a microcontroller, a speaker, and a user operation switch (optional). The light source, camera, first wireless transmission module, speaker, and user operation switch are all connected to the microcontroller. The microcontroller processes all data within the medication applicator, and the speaker emits the medication application prompts. The first wireless transmission module is wirelessly connected to the second wireless transmission module, and both the display screen and the second wireless transmission module are connected to the processor.

[0065] The medication applicator also includes a distance sensor connected to a microcontroller. The medication application prompts include marking the identified affected area with a symbol and emitting a prompt sound within a preset range. In this embodiment, the distance sensor can be an infrared sensor or an ultrasonic sensor. This embodiment can determine whether the medication application area is within the effective application range by checking whether the distance between the medication application area and the obstacle is within a preset range, based on feedback from the infrared or ultrasonic sensor. There are no limitations on this.

[0066] The medication application prompt information includes marking the identified affected area with a symbol and emitting a prompt sound within a preset range. The medication application prompt information includes displaying the required medication amount for this application based on the preset medication dosage and / or the previous disease condition. The required medication amount for this application includes the amount of medication to be applied each time and the current remaining number of applications. Specifically, the processor can save the first image of the affected area. During subsequent medication applications, the processor can compare the previously acquired image with the current image and issue a prompt based on the comparison result. Alternatively, it can issue a prompt based on the preset dosage, number of applications, and the current application number. For example, if a total of 6 applications are required, twice a day, 0.3g each time, and this is the third application, the speaker will announce, "Affected area reached, medication can be applied. This is the third application; please apply the fourth application in three hours," or "Affected area reached, medication can be applied. This is the first application today; please apply the second application in three hours." If the comparison between the previously acquired image and the current image shows that the affected area has shrunk, the speaker will announce, "Improvement has been observed; please continue the course of treatment. We wish you a speedy recovery." Such information is not limited in this invention.

[0067] The probe's drug delivery portion is a spoon-shaped component and / or a hollow portion. When it is hollow, it works in conjunction with the drug storage portion and pressure generating module within the handle to deliver medication to the target affected area. In this embodiment, when using the device, the user presses the switch after hearing a prompt and confirming the location according to the prompt information displayed on the display device. The handle then generates spray pressure to spray solid or liquid medication onto the target affected area. To reduce the cost of the medication application device, the switch can be omitted, and the probe can be a spoon-shaped portion. In this embodiment, when the probe is a spoon-shaped portion, the user can apply the medication originally placed in the spoon to the target affected area after hearing a prompt and confirming the location according to the prompt information displayed on the display device.

[0068] The preset range is obtained based on an infrared sensor or an ultrasonic sensor and the processor. The processor determines whether the range is within the preset range based on the range data obtained by the infrared sensor or the ultrasonic sensor. If it is, the processor sends a signal to the speaker to activate the medication application prompt.

[0069] The infrared sensor is integrated into the camera or located inside the probe.

[0070] The processor is either an FPGA or an MCU.

[0071] This embodiment implements intelligent voice prompts based on the previous embodiment, making it easier for users to obtain medication information and recovery status, and further improving the user experience.

[0072] Note that the above description is merely a preferred embodiment of the present invention and the technical principles employed. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions can be made without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments, and may include many other equivalent embodiments without departing from the concept of the present invention, the scope of which is determined by the scope of the appended claims.

Claims

1. An oral medication application device based on image recognition, characterized in that, The oral medication delivery device includes: A wireless data link medication applicator and display device; the medication applicator includes a handle with a first wireless transmission module and a probe mounted on the handle and having a light source and a camera; the display device includes a second wireless transmission module, a processor, and a display screen; wherein, the probe further includes a drug delivery part for applying medication to the target affected area; the first wireless transmission module is used to wirelessly transmit images acquired by the camera to the second wireless transmission module; the processor is used to process the images acquired by the second wireless transmission module, identify the target affected area based on the processed image, and provide medication application prompts; The medication applicator also includes a distance sensor, which is used to obtain the distance range between the probe and the front obstacle. The medication application prompt information includes marking the identified affected area with a symbol and emitting a prompt sound when the distance range is within a preset range. The step of marking the identified affected area with a symbol and emitting a prompt sound within a preset range includes: Step 1: Establish an empirical database of oral ulcer images, set image pixels, segment the images acquired by the camera into images to be recognized, select class models with similar pattern abstract features for the images to be recognized, and preprocess the images to be recognized. The preprocessing includes setting the component R=G=B for each pixel of the image to be recognized, thereby obtaining the feature vector of the grayscale value of the image to be recognized. ; Step 2: Make and , in yes The transpose of , For the vector of the class model, N is the dimension of the vector of the class model of the image to be identified, and the accompanying vector is obtained. Make it satisfy the following formula: ,in Represents the adjoint vector correspond The eigenvectors in the vector, k' is The index corresponding to the internal feature vector, k=1……M, where M is the number of the class models, M≤N; Step 3: Input the initial value q(0) of the feature vector of the model to be tested in the image to be recognized, and use the formula Find the ordered parameters initial value ; Step 4: Using parameters Establish the following Iterative formula: , in The model is selected based on the similarity between the class model and the vector of the model to be tested. b and c are coefficients preset based on experience, and d satisfies: ； Step 5: Establish a three-layer neural network model. The first to third layers of the neural network model are the input layer, the second layer, and the output layer, respectively. Each receiving unit in the input layer receives the component q(0) of the feature vector q(0) of the model to be recognized. j (0); The second layer of the neural network model consists of various parameters. Neuron, in which parameters It is each q j (0) multiplied by q j (0) Connected The sum of the products, where adjoint vector The j-th component; the second layer from the formula in step 4 (0) Start iterative running, and in If the result is ≠0, return to step 4 and run the process again until... When =0, proceed to step 6 to obtain the final image recognition result from the output layer; Step 6: Obtain the final image recognition result according to the following formula: ,in Let t be the t-th element of the output layer unit j. For parameters The t-th vector, t=0,1,2,……T, where T is The total number of vectors, where M is the number of class models; The identified The image is input into an oral ulcer image experience database and compared with the experience data in the database to determine whether the image to be identified is an image of the target lesion. If the result is an image of the target lesion, the image identification result is marked with a symbol.

2. The oral medication applicator according to claim 1, characterized in that, The medication application prompt information includes marking the identified affected area with a symbol and issuing a prompt sound within a preset range, which includes: converting the acquired image into a grayscale image, identifying the pixel area with grayscale value within the threshold area in the grayscale image as the target area, and the pixel area with other grayscale value as the background area; calculating the centroid of the target area, and issuing a prompt sound if the centroid pixel meets the preset distance condition.

3. The oral medication applicator according to claim 2, characterized in that, The medication application prompt information includes displaying the required medication amount based on the preset medication dosage and / or the previous patient's condition. The required medication amount includes the amount of medication to be applied each time and the current remaining number of applications.

4. The oral medication applicator according to claim 3, characterized in that, The drug delivery part of the probe is a spoon-shaped component and / or a hollow part. When it is a hollow part, it works together with the drug storage part and pressure generation module in the handle to deliver the drug to the target affected area.

5. The oral medication applicator according to claim 4, characterized in that, The preset range is obtained based on the infrared sensor and the processor.

6. The oral medication applicator according to claim 5, characterized in that, The infrared sensor is integrated into the camera or located inside the probe.

7. The oral medication applicator according to claim 6, characterized in that, The processor is either an FPGA or an MCU.

Images