CN112561865A - Constant molar position detection model training method, system and storage medium - Google Patents
Constant molar position detection model training method, system and storage medium Download PDFInfo
- Publication number
- CN112561865A CN112561865A CN202011406995.3A CN202011406995A CN112561865A CN 112561865 A CN112561865 A CN 112561865A CN 202011406995 A CN202011406995 A CN 202011406995A CN 112561865 A CN112561865 A CN 112561865A
- Authority
- CN
- China
- Prior art keywords
- image
- images
- feature
- detection model
- feature extraction
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 85
- 238000000034 method Methods 0.000 title claims abstract description 53
- 238000012549 training Methods 0.000 title claims abstract description 33
- 238000000605 extraction Methods 0.000 claims abstract description 39
- 230000004913 activation Effects 0.000 claims abstract description 23
- 230000003213 activating effect Effects 0.000 claims abstract description 13
- 210000000214 mouth Anatomy 0.000 claims abstract description 13
- 210000004513 dentition Anatomy 0.000 claims abstract description 12
- 230000036346 tooth eruption Effects 0.000 claims abstract description 12
- 238000011176 pooling Methods 0.000 claims description 20
- 230000007246 mechanism Effects 0.000 claims description 13
- 230000008197 oral dentition Effects 0.000 claims description 9
- 230000004927 fusion Effects 0.000 claims description 8
- 238000012545 processing Methods 0.000 claims description 8
- 238000005070 sampling Methods 0.000 claims description 7
- 230000008569 process Effects 0.000 abstract description 14
- 230000006870 function Effects 0.000 description 7
- 241000894006 Bacteria Species 0.000 description 5
- 230000009286 beneficial effect Effects 0.000 description 5
- 208000002925 dental caries Diseases 0.000 description 5
- 238000013528 artificial neural network Methods 0.000 description 4
- 238000012216 screening Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 238000007789 sealing Methods 0.000 description 3
- 230000002378 acidificating effect Effects 0.000 description 2
- 230000001013 cariogenic effect Effects 0.000 description 2
- 238000004590 computer program Methods 0.000 description 2
- 230000006378 damage Effects 0.000 description 2
- 201000010099 disease Diseases 0.000 description 2
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 2
- 235000013305 food Nutrition 0.000 description 2
- 238000001310 location test Methods 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 239000003566 sealing material Substances 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 208000017667 Chronic Disease Diseases 0.000 description 1
- 241000282414 Homo sapiens Species 0.000 description 1
- 239000002253 acid Substances 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000004888 barrier function Effects 0.000 description 1
- 150000001720 carbohydrates Chemical class 0.000 description 1
- 235000014633 carbohydrates Nutrition 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 239000011248 coating agent Substances 0.000 description 1
- 238000000576 coating method Methods 0.000 description 1
- 230000003670 easy-to-clean Effects 0.000 description 1
- 230000003628 erosive effect Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000013100 final test Methods 0.000 description 1
- 238000007499 fusion processing Methods 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 239000002207 metabolite Substances 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 239000012466 permeate Substances 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000001681 protective effect Effects 0.000 description 1
- 239000004576 sand Substances 0.000 description 1
- 230000028327 secretion Effects 0.000 description 1
- 235000021058 soft food Nutrition 0.000 description 1
- 235000021147 sweet food Nutrition 0.000 description 1
- 208000024891 symptom Diseases 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30036—Dental; Teeth
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Health & Medical Sciences (AREA)
- Quality & Reliability (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a method, a system and a storage medium for training a detection model of a constant molar position, wherein the method comprises the following steps: acquiring a plurality of oral cavity dentition images as first images; performing first feature extraction on the first image to obtain a first feature image; selectively activating the first feature image in a channel dimension and a space dimension respectively; performing second feature extraction on the first feature image after the selection activation; a detection frame for predicting the permanent molar position according to the image features extracted by the second features; fusing the information of the detection frame to generate a first confidence coefficient; taking the maximum first confidence as a prediction output result; and reversely updating the parameters of the detection model according to the prediction output result. The detection model disclosed by the invention is more targeted and more suitable for different application situations in the application process, so that the accuracy of the detection result of the detection model in the application process is improved. The invention can be widely applied to the technical field of model training.
Description
Technical Field
The invention relates to the technical field of model training, in particular to a method and a system for training a detection model of a constant molar position and a storage medium.
Background
Caries is a chronic disease in which bacteria (cariogenic bacteria) produce acid using carbohydrates in food, resulting in progressive destruction of hard tissues of teeth, is a major common disease of the oral cavity and is one of the most common diseases of human beings. Children of school age 6-12 years old are often fond of sweet food, soft food and sticky food and are easy to stick on teeth, but the oral hygiene habits of the children at this stage are often not good or the cleaning method is not well mastered, so that the teeth are difficult to clean effectively, and the children are high-incidence people suffering from caries. In addition, when the children are in the tooth replacement stage at this stage, the newly erupted permanent teeth, especially the newly erupted permanent molar fossae are often deep, bacteria are easy to accumulate and are not easy to clean, and the bacteria generate acidic secretion to damage the hard tissues of the teeth, so that decayed teeth are easy to generate. Therefore, timely screening of oral cavity conditions of children and early intervention are the primary and secondary symptoms of preventing permanent tooth caries. The pit and furrow closing is the best method recommended by the world health organization for preventing permanent tooth caries, is generally popularized in many countries and regions, and is also widely popularized in China for the purpose of closing the pit and furrow of school-age children. The pit and fissure sealing refers to a method for preventing pit and fissure caries by coating pit and fissure sealing material on pit and fissure points on occlusal surface and buccal and lingual surface of dental crown, solidifying and hardening after the pit and fissure sealing material flows into and permeates into the pit and fissure to form a protective barrier which is covered on deep pit and fissure and can prevent cariogenic bacteria and acidic metabolites from eroding tooth body.
The first step in the closure of the sulcus is to screen out the permanent molars for indications. The previous pit and fissure sealing projects are screened by the manual force of stomatologists, and a large amount of manpower, material resources and financial resources are required. In order to save the manpower, material resources and financial resources of the pit and fissure sealing project, a method for detecting and early warning teeth on line is provided. However, the current online detection model lacks pertinence and adaptive situations, so that the accuracy of the identification result of the permanent molars is not high in the application process.
Disclosure of Invention
To solve one of the above technical problems, the present invention aims to: provided are a constant molar position detection model training method, system and storage medium, which can improve the accuracy of a detection result of a detection model in an application process.
In a first aspect, an embodiment of the present invention provides:
a constant molar position detection model training method comprises the following steps:
acquiring a plurality of oral cavity dentition images as first images;
performing first feature extraction on the first image to obtain a first feature image;
selectively activating the first feature image in a channel dimension and a space dimension respectively;
performing second feature extraction on the first feature image after the selection activation;
a detection frame for predicting the permanent molar position according to the image features extracted by the second features;
fusing the information of the detection frame to generate a first confidence coefficient;
taking the maximum first confidence as a prediction output result;
and reversely updating the parameters of the detection model according to the prediction output result.
Further, the acquiring a plurality of oral dentition images as a first image includes:
acquiring a plurality of oral cavity dentition images;
processing the sizes of the oral cavity dentition images into preset sizes;
carrying out interlaced sampling on the dental image after the size processing to obtain a plurality of sub-images;
and taking the spliced image of the plurality of sub-images as a first image.
Further, the performing the first feature extraction on the first image to obtain a first feature image includes:
performing down-sampling feature extraction on the first image to generate feature images with different sizes;
and taking the characteristic images with different sizes as first characteristic images.
Further, the selectively activating the first feature image in a channel dimension includes:
performing average pooling and maximum pooling on the first feature images;
and selectively activating the first feature images after the average pooling and the maximum pooling in the channel dimension by using an attention mechanism. Further, the selectively activating the first feature image in the spatial dimension specifically includes:
the first image after attention mechanism processing is selectively activated in the spatial dimension using a convolution kernel.
Further, the second feature extraction is abstract feature extraction.
Further, the fusing the information of the detection frame to generate a first confidence level includes:
fusing the position information, the second confidence coefficient and the image characteristic information of the detection frame;
and generating a first confidence coefficient according to the fusion result.
In a second aspect, an embodiment of the present invention provides:
a constant molar position detection model training system, comprising:
the acquisition module is used for acquiring a plurality of oral cavity dentition images as first images;
the first feature extraction module is used for performing first feature extraction on the first image to obtain a first feature image;
the activation module is used for selectively activating the first characteristic image in a channel dimension and a space dimension respectively;
the second feature extraction module is used for performing second feature extraction on the selected and activated first feature image;
the prediction module is used for predicting a detection frame of the constant molar position according to the image features extracted by the second features;
the fusion module is used for fusing the information of the detection frame to generate a first confidence coefficient;
the confidence coefficient selection module is used for taking the maximum first confidence coefficient as a prediction output result;
and the parameter updating module is used for reversely updating the parameters of the detection model according to the prediction output result.
In a third aspect, an embodiment of the present invention provides:
a constant molar position detection model training system, comprising:
at least one memory for storing a program;
at least one processor for loading the program to perform the method for training a permanent molars location test model.
In a fourth aspect, an embodiment of the present invention provides:
a storage medium having stored therein a processor-executable program for performing the method for training a permanent molars location detection model when executed by a processor.
The embodiment of the invention has the beneficial effects that: according to the embodiment of the invention, the first feature extraction is carried out on a plurality of oral dentition images, then the first feature images are selectively activated in the channel dimension and the space dimension respectively, the second feature extraction is carried out on the first feature images after the activation is selected, then the detection frame of the position of the permanent molars is predicted according to the image features after the second feature extraction, the information of the detection frame is fused to generate the first confidence coefficient, and finally the maximum first confidence coefficient is used as the prediction output result, and the parameters of the detection model are reversely updated according to the prediction output result, so that the detection model trained by the embodiment is more targeted and more suitable for different application situations in the application process, and the accuracy of the detection result of the detection model in the application process is improved.
Drawings
Fig. 1 is a flowchart of a constant molar position detection model training method according to an embodiment of the present invention.
Detailed Description
The invention is described in further detail below with reference to the figures and the specific embodiments. The step numbers in the following embodiments are provided only for convenience of illustration, the order between the steps is not limited at all, and the execution order of each step in the embodiments can be adapted according to the understanding of those skilled in the art.
In the following description, reference is made to "some embodiments" which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of the present application only and is not intended to be limiting of the application.
Referring to fig. 1, an embodiment of the present invention provides a method for training a detection model of a constant molar position, and this embodiment may be applied to a server or a background processor of various platforms, so that in an application process, a user only needs to perform a specified step operation on a human-computer interaction interface, and thus, the detection of the constant molar position can be completed.
Wherein, this embodiment includes the following steps:
s1, acquiring a plurality of oral dentition images as first images; in this step, the oral dentition image may be an image of the person to be detected, which is acquired in an actual detection process, and the acquisition process may be obtained by shooting the person to be detected through a mobile phone or other devices with a camera function.
When the plurality of oral dentition images are applied to the training process of the detection model, in some embodiments, data enhancement modes such as data inversion and rotation can be adopted to perform data enhancement on the oral dentition images so as to provide more training sample data and improve the training precision.
In some embodiments, the step S1 can be implemented by:
acquiring a plurality of oral cavity dentition images; the plurality of oral dentition images are images shot by different mobile intelligent devices, and because the sizes of the shot images are different and the input size of the neural network is fixed, the sizes of the plurality of oral dentition images are processed into preset sizes. In this process, the image size may be set to 512 × 512, that is, the size of the input image data is C × W × H, where C ═ 3 indicates that the number of input image channels is 3, and each channel sequentially indicates the color value of the red, green, and blue components, respectively; w is 512 the width of the image; h512 is the height of the image. Secondly, carrying out interlaced sampling on the dental dentition image after size processing to obtain a plurality of sub-images; for example, for a 512 x 512 size image, the original image is sampled intermittently at intervals of length 2 starting from each point in a 2 x 2 region in the upper left corner of the image, resulting in 4 sub-images of 256 x 256. And finally, taking the image formed by splicing the plurality of sub-images as a first image input by the detection model so as to fully utilize the information on the image.
And S2, performing first feature extraction on the first image to obtain a first feature image.
In this step, the first image is subjected to first feature extraction by the feature extraction module. The feature extraction module is formed by superposing 8-layer network modules, and 2 layers in the 8-layer network form a down-sampling feature extraction module group which is divided into four groups in total, and each group of operations generates a feature map with corresponding size. After feature extraction is performed through the down-sampling feature extraction module group, 4 feature maps with sizes respectively 512 × 512, 256 × 256, 128 × 128 and 64 × 64 can be obtained, and finally feature images with different sizes are used as first feature images.
S3, selectively activating the first characteristic image in a channel dimension and a space dimension respectively;
in some embodiments, selective activation of the first feature image in the channel dimension has been achieved by:
performing average pooling and maximum pooling on the first feature images;
and selectively activating the first feature images after the average pooling and the maximum pooling in the channel dimension by using an attention mechanism.
In some embodiments, selective activation of the first feature image in the spatial dimension may be performed in the following manner:
the first image after attention mechanism processing is selectively activated in the spatial dimension using a convolution kernel.
In the above embodiment, the attention mechanism and the convolution kernel form an attention mechanism module, and specifically, selective activation of channel dimensions is performed first, in this portion, the feature maps are subjected to maximum pooling and average pooling respectively, and then the feature maps obtained by the two operations are selectively activated for corresponding feature channels respectively, where selection of an activation channel is a parameter that can be learned by neural network iteration, and finally two features obtained after activation of channel dimensions respectively are superimposed and fused to be used as features after activation of channel dimensions. And secondly, selective activation of spatial dimensions, namely adopting a strategy of simultaneously utilizing the feature maps after average pooling and maximum pooling as input, utilizing a 7-7 convolution kernel to enlarge a receptive field, and mapping the output after convolution to a numerical value in [0,1] through a sigmoid activation function to represent the importance degree of each position in the feature maps relative to a prediction target.
Specifically, in selective activation of channel dimensions, the calculation is performed using equation 1:
wherein, FSFor features that are selectively activated through the channel dimensions,
representing the image features after the average pooling,
representing features of the image after maximum pooling, ω0And ω1Two different matrix parameters are used for representing the weight size for different channel numbers, and sigma represents an activation function sigmoid.
In this step, two identical parameters are applied to the feature maps processed in different manners, so that the parameters have stronger robustness, and the feature maps continuously applied with the two parameters can contain more neurons to fit features as complex as possible on the basis of ensuring that the size of the feature maps is unchanged. Finally, the mapping from real domain R to [0,1] is done using an activation function represented by σ to achieve normalization of the output.
In selective activation of spatial dimensions, the calculation is performed using equation 2:
wherein f is7*7Represents the operation of 7 by 7 convolution,
representing image features obtained by averaging pooling based on features obtained by selective activation of channel dimensions,
the method includes the steps that image features obtained by maximum pooling are shown on the basis of features obtained by channel dimension selective activation, and sigma represents an activation function sigmoid.
In this step, the reception field is enlarged by using a larger convolution kernel, so that the model is helped to deduce the importance degree of the current position according to the current pixel point and the content of the field thereof.
S4, performing second feature extraction on the selected and activated first feature image; in this step, the second feature extraction is abstract feature extraction.
S5, predicting a detection frame of the permanent molar position according to the image features extracted by the second features; in this step, it specifically predicts the position and size of the detection frame on the first image.
And S6, fusing the information of the detection frame to generate a first confidence coefficient.
In some embodiments, step S6 may be specifically implemented by:
fusing the position information, the second confidence coefficient and the image characteristic information of the detection frame;
and generating a first confidence coefficient according to the fusion result.
In this embodiment, the second confidence is the confidence of the image itself, and the first confidence is the confidence after the same image fusion processing as the second confidence.
This is not possible in practice because the existing model predicts a deviation in the detection frame, for example, two teeth of the same type are present in the same side. Thus, in some embodiments, the mechanism is screened by a detection box employing multimodal information fusion. Specifically, the existing detection frame screening method only judges according to the confidence generated by the regression model and the overlapping area between different detection frames, and ignores the image feature information of the candidate detection frame itself. In this embodiment, the position information and the image features of the candidate frame are regarded as information of two different modalities, and the information of the two different modalities is fused, so that the detection frame can be effectively screened.
Specifically, when fusing information of different modalities, formula 3 and formula 4 are adopted:
wherein f isAFor image features after fusion of confidence information, fIFor the confidence information after fusing the image features,
and
the original image characteristics and the confidence coefficient information are respectively, and the other variables are parameters self-learned by the neural network.
On the basis of fusing information of different modes, calculating by combining position information of images to obtain the confidence coefficient of the detection frame fusing multi-mode information, wherein the specific calculation process is shown as a formula 5-a formula 8:
fS＝σ(WSftotal)+bSequation 8
Wherein,
representing the feature after fusing the confidence information and the image feature information,
for the coded detection frame position information, ftotal,kThen it is indicated at
Further fusing the characteristics of the position information of the detection frame fSAnd the new confidence coefficient finally obtained after the multi-mode information is fused is represented, and the rest letters or letter combinations represent parameters required by the neural network.
S7, taking the maximum first confidence as a prediction output result; in this step, the prediction output result with the highest confidence level is obtained for each region.
S8, reversely updating the parameters of the detection model according to the prediction output result, so that the detection model is more targeted and more adaptable to different application situations in the application process, and the accuracy of the detection result of the detection model in the application process is improved.
The embodiment is specifically applied, and comprises the following steps:
the method comprises the steps of collecting color images collected by the mobile intelligent equipment, wherein the angles, light rays and covered area conditions of all the images are different. In the present embodiment, the size of the images is unified to 512 × 512.
The image data set was divided into a training set and a test set at a 4:1 ratio, i.e., 3316 pictures in the data set were used to train the model and the remaining 829 pictures were used to verify the performance of the model. The final test results as shown in table 1:
TABLE 1
| Method | AP | AP50 | AP75 | AR | AR50 | SR75 | APmolar1 | APmolar2 | Time(ms) |
| Baseline(D) | 46.2 | 92.5 | 40.2 | 48.4 | 99.7 | 70.3 | 92.7 | 94.3 | 6.6 |
| Baseline(O) | 44.5 | 89.7 | 37.2 | 46.7 | 98.1 | 64.3 | - | 90.1 | 6.1 |
| Baseline(A+D) | 46.7 | 93.2 | 40.8 | 48.6 | 99.7 | 70.8 | 93.5 | 95.4 | 8.9 |
| Baseline(AN+D) | 47.9 | 94.5 | 41.6 | 48.5 | 98.9 | 72.5 | 94.2 | 97.3 | 10.4 |
| Baseline(AN+A+D) | 49.1 | 95.6 | 42.3 | 49.1 | 99.2 | 72.6 | 96.1 | 98.5 | 12.4 |
In table 1, AP and AR are used as evaluation indexes, that is, the IOU values of the detection box and the target box are considered to be correct when the values are greater than a threshold value. Wherein, Baseline (D) represents the result when the first premolar and the second premolar are detected simultaneously, Baseline (O) represents the result when only the second premolar is detected, A represents adding AN attention mechanism, and AN represents adding AN A-NMS detection frame screening mechanism.
In summary, the above embodiment limits the range of target detection by adding the attention mechanism, and greatly increases the adaptability and robustness of the method. Meanwhile, a detection frame screening mechanism is set, the image features and the position features of the target are integrated to further select the detection frame, and the candidate frame matched with the image features and the position features is reserved, so that the accuracy of the detection result is improved on the same data set.
The embodiment of the invention provides a constant molar position detection model training system corresponding to the method shown in figure 1, which comprises the following steps:
the acquisition module is used for acquiring a plurality of oral cavity dentition images as first images;
the first feature extraction module is used for performing first feature extraction on the first image to obtain a first feature image;
the activation module is used for selectively activating the first characteristic image in a channel dimension and a space dimension respectively;
the second feature extraction module is used for performing second feature extraction on the selected and activated first feature image;
the prediction module is used for predicting a detection frame of the constant molar position according to the image features extracted by the second features;
the fusion module is used for fusing the information of the detection frame to generate a first confidence coefficient;
the confidence coefficient selection module is used for taking the maximum first confidence coefficient as a prediction output result;
and the parameter updating module is used for reversely updating the parameters of the detection model according to the prediction output result.
The content of the embodiment of the method of the invention is all applicable to the embodiment of the system, the function of the embodiment of the system is the same as the embodiment of the method, and the beneficial effect achieved by the embodiment of the system is the same as the beneficial effect achieved by the method.
The embodiment of the invention provides a detection model training system for a constant molar position, which comprises:
at least one memory for storing a program;
at least one processor for loading the program to perform the method for training a permanent molars location test model.
The content of the embodiment of the method of the invention is all applicable to the embodiment of the system, the function of the embodiment of the system is the same as the embodiment of the method, and the beneficial effect achieved by the embodiment of the system is the same as the beneficial effect achieved by the method.
An embodiment of the present invention provides a storage medium in which a processor-executable program is stored, which, when being executed by a processor, is configured to perform the method for training a model for detecting permanent molar positions.
The embodiment of the invention also discloses a computer program product or a computer program, which comprises computer instructions, and the computer instructions are stored in a computer readable storage medium. The computer instructions may be read by a processor of a computer device from a computer-readable storage medium, and executed by the processor to cause the computer device to perform the method illustrated in fig. 1.
While the preferred embodiments of the present invention have been illustrated and described, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.
Claims (10)
1. A constant molar position detection model training method is characterized by comprising the following steps:
acquiring a plurality of oral cavity dentition images as first images;
performing first feature extraction on the first image to obtain a first feature image;
selectively activating the first feature image in a channel dimension and a space dimension respectively;
performing second feature extraction on the first feature image after the selection activation;
a detection frame for predicting the permanent molar position according to the image features extracted by the second features;
fusing the information of the detection frame to generate a first confidence coefficient;
taking the maximum first confidence as a prediction output result;
and reversely updating the parameters of the detection model according to the prediction output result.
2. The training method of the detection model for the permanent molar position according to claim 1, wherein the acquiring a plurality of oral dentition images as the first image comprises:
acquiring a plurality of oral cavity dentition images;
processing the sizes of the oral cavity dentition images into preset sizes;
carrying out interlaced sampling on the dental image after the size processing to obtain a plurality of sub-images;
and taking the spliced image of the plurality of sub-images as a first image.
3. The constant molar position detection model training method according to claim 1, wherein the performing a first feature extraction on the first image to obtain a first feature image comprises:
performing down-sampling feature extraction on the first image to generate feature images with different sizes;
and taking the characteristic images with different sizes as first characteristic images.
4. A method for training a permanent molars detection model according to claim 1, wherein said selectively activating the first feature image in the channel dimension comprises:
performing average pooling and maximum pooling on the first feature images;
and selectively activating the first feature images after the average pooling and the maximum pooling in the channel dimension by using an attention mechanism.
5. The training method for the detection model of the permanent molar position according to claim 4, wherein the selective activation of the first feature image in the spatial dimension is specifically:
the first image after attention mechanism processing is selectively activated in the spatial dimension using a convolution kernel.
6. A constant molar position detection model training method according to claim 1, wherein the second feature extraction is abstract feature extraction.
7. The training method of the detection model for the permanent molar position according to claim 1, wherein the fusing the information of the detection frame to generate the first confidence level comprises:
fusing the position information, the second confidence coefficient and the image characteristic information of the detection frame;
and generating a first confidence coefficient according to the fusion result.
8. A constant molar position detection model training system, comprising:
the acquisition module is used for acquiring a plurality of oral cavity dentition images as first images;
the first feature extraction module is used for performing first feature extraction on the first image to obtain a first feature image;
the activation module is used for selectively activating the first characteristic image in a channel dimension and a space dimension respectively;
the second feature extraction module is used for performing second feature extraction on the selected and activated first feature image;
the prediction module is used for predicting a detection frame of the constant molar position according to the image features extracted by the second features;
the fusion module is used for fusing the information of the detection frame to generate a first confidence coefficient;
the confidence coefficient selection module is used for taking the maximum first confidence coefficient as a prediction output result;
and the parameter updating module is used for reversely updating the parameters of the detection model according to the prediction output result.
9. A constant molar position detection model training system, comprising:
at least one memory for storing a program;
at least one processor for loading the program to perform a method of training a permanent molars location detection model according to any of claims 1-7.
10. A storage medium having stored therein a processor-executable program, wherein the processor-executable program, when executed by a processor, is adapted to perform a method of training a model for detection of permanent molar positions according to any of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011406995.3A CN112561865B (en) | 2020-12-04 | 2020-12-04 | Method, system and storage medium for training detection model of constant molar position |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011406995.3A CN112561865B (en) | 2020-12-04 | 2020-12-04 | Method, system and storage medium for training detection model of constant molar position |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112561865A true CN112561865A (en) | 2021-03-26 |
| CN112561865B CN112561865B (en) | 2024-03-12 |
Family
ID=75048486
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011406995.3A Active CN112561865B (en) | 2020-12-04 | 2020-12-04 | Method, system and storage medium for training detection model of constant molar position |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112561865B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113343853A (en) * | 2021-06-08 | 2021-09-03 | 深圳格瑞健康管理有限公司 | Intelligent screening method and device for child dental caries |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040029068A1 (en) * | 2001-04-13 | 2004-02-12 | Orametrix, Inc. | Method and system for integrated orthodontic treatment planning using unified workstation |
| US20130158958A1 (en) * | 2010-07-12 | 2013-06-20 | Alain Methot | Dental analysis method and system |
| US20180028294A1 (en) * | 2016-07-27 | 2018-02-01 | James R. Glidewell Dental Ceramics, Inc. | Dental cad automation using deep learning |
| US20190030371A1 (en) * | 2017-07-28 | 2019-01-31 | Elekta, Inc. | Automated image segmentation using dcnn such as for radiation therapy |
| CN110246571A (en) * | 2019-07-30 | 2019-09-17 | 深圳市倍康美医疗电子商务有限公司 | Tooth data processing method |
| US20190313963A1 (en) * | 2018-04-17 | 2019-10-17 | VideaHealth, Inc. | Dental Image Feature Detection |
| CN110349224A (en) * | 2019-06-14 | 2019-10-18 | 众安信息技术服务有限公司 | A kind of color of teeth value judgment method and system based on deep learning |
| CN110619350A (en) * | 2019-08-12 | 2019-12-27 | 北京达佳互联信息技术有限公司 | Image detection method, device and storage medium |
| CN110930421A (en) * | 2019-11-22 | 2020-03-27 | 电子科技大学 | Segmentation method for CBCT (Cone Beam computed tomography) tooth image |
| CN111882002A (en) * | 2020-08-06 | 2020-11-03 | 桂林电子科技大学 | MSF-AM-based low-illumination target detection method |
| CN111986217A (en) * | 2020-09-03 | 2020-11-24 | 北京大学口腔医学院 | Image processing method, device and equipment |
-
2020
- 2020-12-04 CN CN202011406995.3A patent/CN112561865B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040029068A1 (en) * | 2001-04-13 | 2004-02-12 | Orametrix, Inc. | Method and system for integrated orthodontic treatment planning using unified workstation |
| US20130158958A1 (en) * | 2010-07-12 | 2013-06-20 | Alain Methot | Dental analysis method and system |
| US20180028294A1 (en) * | 2016-07-27 | 2018-02-01 | James R. Glidewell Dental Ceramics, Inc. | Dental cad automation using deep learning |
| US20190030371A1 (en) * | 2017-07-28 | 2019-01-31 | Elekta, Inc. | Automated image segmentation using dcnn such as for radiation therapy |
| US20190313963A1 (en) * | 2018-04-17 | 2019-10-17 | VideaHealth, Inc. | Dental Image Feature Detection |
| CN110349224A (en) * | 2019-06-14 | 2019-10-18 | 众安信息技术服务有限公司 | A kind of color of teeth value judgment method and system based on deep learning |
| CN110246571A (en) * | 2019-07-30 | 2019-09-17 | 深圳市倍康美医疗电子商务有限公司 | Tooth data processing method |
| CN110619350A (en) * | 2019-08-12 | 2019-12-27 | 北京达佳互联信息技术有限公司 | Image detection method, device and storage medium |
| CN110930421A (en) * | 2019-11-22 | 2020-03-27 | 电子科技大学 | Segmentation method for CBCT (Cone Beam computed tomography) tooth image |
| CN111882002A (en) * | 2020-08-06 | 2020-11-03 | 桂林电子科技大学 | MSF-AM-based low-illumination target detection method |
| CN111986217A (en) * | 2020-09-03 | 2020-11-24 | 北京大学口腔医学院 | Image processing method, device and equipment |
Non-Patent Citations (2)
| Title |
|---|
| 段志伟;: "基于特征融合和中心预测的目标检测算法", 现代计算机, no. 09 * |
| 王心醉;董宁宁;李欢利;: "基于SIFT算法的图像特征点提取和匹配研究", 南京医科大学学报(自然科学版), no. 02 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113343853A (en) * | 2021-06-08 | 2021-09-03 | 深圳格瑞健康管理有限公司 | Intelligent screening method and device for child dental caries |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112561865B (en) | 2024-03-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20210174477A1 (en) | Domain specific image quality assessment | |
| US20210022833A1 (en) | Generation of synthetic post treatment images of teeth | |
| US20220218449A1 (en) | Dental cad automation using deep learning | |
| CN108416377B (en) | Information extraction method and device in histogram | |
| CN113194872B (en) | Identification device, scanner system, and identification method | |
| CN110473243B (en) | Tooth segmentation method and device based on depth contour perception and computer equipment | |
| WO2020063986A1 (en) | Method and apparatus for generating three-dimensional model, device, and storage medium | |
| CN105654436A (en) | Backlight image enhancement and denoising method based on foreground-background separation | |
| CN115362451A (en) | System and method for constructing three-dimensional model from two-dimensional image | |
| DE102012110491A1 (en) | Method and device for cosmetic tooth analysis and dental consultation | |
| CN106875361A (en) | A kind of method that poisson noise is removed based on depth convolutional neural networks | |
| US11887209B2 (en) | Method for generating objects using an hourglass predictor | |
| US20210097730A1 (en) | Face Image Generation With Pose And Expression Control | |
| CN107194426A (en) | A kind of image-recognizing method based on Spiking neutral nets | |
| CN107403065A (en) | Detection method, device and the server of dental problems | |
| CN112561865A (en) | Constant molar position detection model training method, system and storage medium | |
| CN106898011A (en) | A kind of method that convolutional neural networks convolution nuclear volume is determined based on rim detection | |
| CN103679710B (en) | The weak edge detection method of image based on multilayer neuron pool discharge information | |
| CN113516639B (en) | Training method and device for oral cavity abnormality detection model based on panoramic X-ray film | |
| CN114586069A (en) | Method for generating dental images | |
| CN106023238A (en) | Color data calibration method for camera module | |
| CN108735010A (en) | A kind of intelligent English teaching system for English teaching | |
| JP6831433B2 (en) | Identification device, tooth type identification system, identification method, and identification program | |
| US12020372B2 (en) | Method for analyzing a photo of a dental arch | |
| CN113343853B (en) | Intelligent screening method and device for dental caries of children |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information |
Address after: 805, Building B, Second Unified Building, Houhai Neighborhood Committee, No.1 Nanyou Huaming Road, Nanshan District, Guangdong Province, 518000 Applicant after: Shenzhen Gree Health Technology Co.,Ltd. Address before: 805, block B, No.2 Tongjian building, Houhai neighborhood committee, No.1 Huayou Huaming Road, Nanshan District, Shenzhen, Guangdong 518000 Applicant before: Shenzhen Gree Health Management Co.,Ltd. |
|
| CB02 | Change of applicant information | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |