Tooth defect identification method for children

Abstract

The invention relates to the technical field of oral medicine, in particular to a tooth defect identification method for children, which comprises the following steps: in a defective tooth detection process, determining whether model construction is qualified or not based on a defect feature surface proportion and a defect geometric distribution value, and when the model construction is judged to be abnormal, adjusting generation parameters of a tooth model, the accuracy of defective tooth detection is improved, and the detection efficiency of defective teeth is further improved.

Description

Technical Field

[0001] The present invention relates to the technical field of stomatology, and particularly relates to a method for identifying tooth defects in children. Background Art

[0002] The early diagnosis of defective teeth is a major problem in the field of stomatology. The affected area of early defective teeth is hidden, and accurate judgment is very important for the formulation of the treatment plan by clinicians. For the diseases of defective teeth, the common discrimination method is to take an X-ray film and then the clinician observes it with the naked eye, or directly discriminates by the clinician observing the teeth. However, due to the early stage of the lesion, the lesion is not obvious, making it difficult to distinguish with the naked eye. Currently in clinical practice, CT tomographic images are also used to discriminate the area of defective teeth. Compared with traditional X-ray films, although the accuracy rate is relatively higher when using CT tomographic images, due to its high price and strong side effects such as radiation, it is not easy to be popularized in clinical practice.

[0003] Chinese Patent Publication No.: CN109859203B discloses a method for identifying defective tooth images based on deep learning. First, the provided defective tooth image data set uses the watershed segmentation algorithm to segment the images in the data set to obtain the images after region segmentation; uses a Gaussian low-pass filter to filter out noise and extracts the image edges with the canny operator to obtain the images after edge extraction; performs image superposition respectively; then, establishes a deep learning image training data set and manually marks the defective areas of the images; selects the deep learning network model of cifar10Net and fine-tunes the parameters of the fully connected layer; performs transfer learning on the tooth image data set, trains and optimizes the network structure and parameters; finally, saves the trained neural network model, inputs the new image to be discriminated, and completes the image preprocessing and the suggestion of the defective area of the tooth image; thus, the above technical solution has the following problems: it is impossible to adjust the generation parameters of the tooth model based on the detection parameters during the detection of defective teeth, which affects the accuracy of the detection of defective teeth and further affects the detection efficiency of defective teeth. Summary of the Invention

[0004] Therefore, the present invention provides a method for identifying tooth defects in children to overcome the problem in the prior art that it is impossible to adjust the generation parameters of the tooth model based on the detection parameters during the detection of defective teeth, which affects the accuracy of the detection of defective teeth and further affects the detection efficiency of defective teeth.

[0005] To achieve the above object, the present invention provides a method for identifying tooth defects in children, including:

[0006] S1, collecting the image information of teeth;

[0007] S2, preprocessing the obtained image information to obtain grayscale image information;

[0008] S3. Construct a tooth model based on the grayscale image information;

[0009] S4. Determine the proportion of the defective feature surface based on the tooth model, and determine whether the model construction is qualified based on the proportion of the defective feature surface and the defective geometric distribution value, including:

[0010] Determine that the model construction is abnormal, and adjust the acquisition parameters of the tooth model based on the defective proportion deviation amount, including adjusting the brightness during the process of acquiring image information to the corresponding value, or adjusting the relevant influence coefficient of the relationship between the grayscale value for mapping the grayscale image information and the defective depth to the corresponding value during the process of constructing the tooth model based on the grayscale image information; or, determine that the model construction is qualified and send an indication message indicating that the tooth is qualified.

[0011] Further, in the S4, the process of determining the proportion of the defective feature surface is to identify the defective areas in the tooth model, calculate the ratio of the total area of each defective area to the surface area of the tooth model, and obtain the proportion of the defective feature surface;

[0012] Determine whether the model construction is qualified based on the proportion of the defective feature surface, including:

[0013] If the proportion of the defective feature surface is less than or equal to the first preset defective proportion, determine that the model construction is qualified and send an indication message indicating that the tooth is qualified;

[0014] If the proportion of the defective feature surface is less than or equal to the second preset defective proportion and greater than the first preset defective proportion, determine whether the model construction is qualified based on the defective geometric distribution value;

[0015] If the proportion of the defective feature surface is greater than the second preset defective proportion, determine that the model construction is abnormal and adjust the acquisition parameters of the tooth model based on the defective proportion deviation amount.

[0016] Further, the process of determining whether the model construction is qualified based on the defective geometric distribution value includes:

[0017] Establish a spatial rectangular coordinate system with the centroid of the tooth model as the origin, obtain the geometric center points of each of the defective areas, record the distance between a single geometric center point and the geometric center point with the smallest spatial distance from it as the distribution distance for this center point, solve the average value of the distribution distances of each geometric center point, and obtain the defective geometric distribution value;

[0018] If the defective geometric distribution value is greater than the preset defective geometric distribution value, determine that the model construction is abnormal and adjust the weighting coefficient of a single color channel to the corresponding value during the preprocessing of the image information based on the average area of each defective area;

[0019] If the defect geometric distribution value is less than or equal to the preset defect geometric distribution value, it is determined that the model construction is qualified, and an indication message that there are defects in the teeth is sent.

[0020] Further, based on the average area of each defect region, the weighting coefficient of a single color channel is adjusted to a corresponding value, where:

[0021] The increase amplitude of the weighting coefficient of a single color channel is proportional to the average area of each defect region.

[0022] Further, the process of adjusting the acquisition parameters of the tooth model based on the defect ratio deviation amount includes:

[0023] The difference between the calculated proportion of the defect feature surface and the second preset defect proportion is recorded as the defect ratio deviation amount;

[0024] If the defect ratio deviation amount is less than or equal to the first preset defect ratio deviation amount, the current acquisition parameters of the tooth model are continuously used, and an indication message that there are defects in the teeth is sent;

[0025] If the defect ratio deviation amount is less than or equal to the second preset defect ratio deviation amount and greater than the first preset defect ratio deviation amount, the acquisition parameters of the tooth model are adjusted based on the volume of the obtained defect region;

[0026] If the defect ratio deviation amount is greater than the second preset defect ratio deviation amount, the brightness in the process of collecting image information is adjusted to a corresponding value based on the area difference ratio.

[0027] Further, the process of adjusting the acquisition parameters of the tooth model based on the volume of the obtained defect region includes:

[0028] The sum of the volumes of the calculated and statistically obtained defect regions is recorded as the defect volume parameter;

[0029] If the defect volume parameter is less than or equal to the preset defect volume parameter, the current acquisition parameters of the tooth model are continuously used, and an indication message that there are defects in the teeth is sent;

[0030] If the defect volume parameter is greater than the preset defect volume parameter, the relevant influence coefficient that maps the relationship between the gray value and the defect depth in the process of constructing the tooth model based on the gray image information is adjusted to a corresponding value based on the defect volume parameter.

[0031] Further, adjusting the relevant influence coefficient to a corresponding value based on the defect volume parameter, where,

[0032] The reduction amplitude of the relevant influence coefficient is proportional to the defect volume parameter.

[0033] Further, adjusting the brightness in the process of collecting image information to a corresponding value based on the area difference ratio, where,

[0034] Solve for the ratio of the defect proportion deviation to the second preset defect proportion deviation to obtain the area difference ratio;

[0035] The increase in brightness is proportional to the area difference ratio.

[0036] Compared with the prior art, the beneficial effect of the present invention is that during the detection of defective teeth, it is determined whether the model construction is qualified based on the surface proportion of defect features and the defect geometric distribution value. When it is determined that the model construction is abnormal, the generation parameters of the tooth model are adjusted, improving the accuracy of defective tooth detection and thus enhancing the detection efficiency of defective teeth.

[0037] Furthermore, it is determined whether the model construction is qualified based on the surface proportion of defect features, which characterizes the proportion of the defect area. When the surface proportion of defect features is less than or equal to the second preset defect proportion and greater than the first preset defect proportion, it is comprehensively determined whether the model construction is qualified based on the defect geometric distribution value, which characterizes the distribution of the defect area. When the defect geometric distribution value is less than or equal to the preset defect geometric distribution value, the distribution is concentrated and the average distribution distance is small. In this case, the defects of the teeth are concentrated and the construction of the model is stable. At this time, it is determined that the teeth have defects. When the defect geometric distribution value is greater than the preset defect geometric distribution value, the distribution is dispersed and the average distribution distance is large. At this time, the image preprocessing parameters are abnormal. Image grayscale conversion is the process of converting a color image into a grayscale image with only brightness information. In a tooth image, grayscale conversion can highlight the brightness changes on the tooth surface, while tooth defects such as cavities and cracks will cause the local brightness to be different from the normal area. In this case, due to improper setting of the grayscale parameters, it is impossible to accurately highlight the brightness difference, resulting in difficult accurate identification of the defect area and misjudgment in a large range of multiple areas. In this case, for weighted average grayscale conversion, the weighting coefficients of the red, green, and blue channels are adjusted to increase the difference in grayscale values between the normal area and the defect area, reducing the probability of being misjudged as a defect in subsequent processing. The smaller the average area of each defect area, the more dispersed the areas determined as defect features, and the greater the adjustment amplitude of the weighting coefficient adjustment. While improving the accuracy of defective tooth detection, it further effectively enhances the detection efficiency of defective teeth.

[0038] Furthermore, in the process of the defect proportion deviation amount adjusting the acquisition parameters of the tooth model, when the defect proportion deviation amount is less than or equal to the first preset defect proportion deviation amount, the defect area of the tooth is within the normal range, and a defect existence notice is issued; when the defect proportion deviation amount is less than or equal to the second preset defect proportion deviation amount and greater than the first preset defect proportion deviation amount, the acquisition parameters of the tooth model are adjusted in combination with the volume of the defect area; when the defect volume parameter is greater than the preset defect volume parameter, at this time, due to a deviation in the model construction process, there is a misjudgment in the defect identification. For this situation where the defect area is too large due to the large defect volume in the model construction, the parameters of the model construction are adjusted; when the defect proportion deviation amount is greater than the second preset defect proportion deviation amount, at this time, due to problems in image acquisition, there are too many misjudgments, and there is an abnormality in the most fundamental acquisition stage. There are problems with the acquisition of the original image information. At this time, the brightness in the image acquisition process is adjusted, and by adjusting the brightness, the misjudgment of the shadow part in the image acquisition process is eliminated; based on the detection results, the situation of abnormal model construction is processed, which improves the accuracy of defective tooth detection and thus improves the detection efficiency of defective teeth. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 FIG. is a flowchart of the steps of the method for identifying tooth defects for children according to an embodiment of the present invention;

[0040] Figure 2 FIG. is a logical decision diagram for determining whether the model construction is qualified based on the surface proportion of defect features according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0041] In order to make the objectives and advantages of the present invention clearer and more understandable, the present invention will be further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0042] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are only used to explain the technical principles of the present invention and do not limit the protection scope of the present invention.

[0043] It should be noted that in the description of the present invention, the terms indicating directions or positional relationships such as "upper", "lower", "left", "right", "inner", "outer", etc. are based on the directions or positional relationships shown in the drawings. This is only for convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention.

[0044] In addition, it should be noted that in the description of the present invention, unless otherwise clearly specified and limited, the terms "installation", "connection", and "coupling" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, and it can be the communication inside two components. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific situations.

[0045] Please refer to Figure 1 and Figure 2 as shown, which are respectively the step flow chart of the tooth defect recognition method for children in the embodiment of the present invention and the logical decision chart for determining whether the model construction is qualified based on the surface proportion of defect features; an embodiment of the present invention provides a tooth defect recognition method for children, including:

[0046] S1, collecting image information of teeth;

[0047] S2, preprocessing the obtained image information to obtain grayscale image information;

[0048] S3, constructing a tooth model based on the grayscale image information;

[0049] S4, determining the surface proportion of defect features based on the tooth model, and determining whether the model construction is qualified based on the surface proportion of defect features and the defect geometric distribution value, including:

[0050] Determining that the model construction is abnormal, and adjusting the acquisition parameters of the tooth model based on the defect proportion deviation amount, including adjusting the brightness during the process of collecting image information to the corresponding value, or adjusting the relevant influence coefficient of the relationship between the grayscale value for mapping the grayscale image information and the defect depth during the process of constructing the tooth model based on the grayscale image information to the corresponding value; or, determining that the model construction is qualified and sending out an indication information that the tooth is qualified.

[0051] Specifically, during the detection of defective teeth, determining whether the model construction is qualified based on the surface proportion of defect features and the defect geometric distribution value, and adjusting the generation parameters of the tooth model when determining that the model construction is abnormal, improves the accuracy of defective tooth detection, and further improves the detection efficiency of defective teeth.

[0052] Specifically, the specific method for collecting the image information of the teeth is not limited. It may include a camera and a light source, and a robotic arm that can penetrate into the oral cavity to capture tooth images. The light source can be a ring-shaped LED lamp with adjustable illumination brightness. The analog signal collected by the camera is converted into a digital signal and transmitted for subsequent storage and processing. It can be understood that it is possible to collect the image information of children's teeth from multiple angles and simultaneously control the brightness to obtain clear, comprehensive, and properly illuminated tooth images. This will not be elaborated further.

[0053] Specifically, the specific process of preprocessing the obtained image information in S2 is not limited. It may include gray-scale processing, which uses image processing software to perform gray-scale conversion to reflect the brightness information of the teeth and highlight the outline and details of the teeth; noise reduction processing, which uses Gaussian filtering to remove Gaussian noise in the image to perform a filtering operation on the image information and make the image smoother; image enhancement, which uses histogram equalization technology to enhance the contrast of the image to make the detailed features of the teeth more obvious; and image normalization, which uniformly adjusts the size of the image to a fixed size for subsequent model processing and normalizes the pixel values of the image. It can be understood that it is possible to convert the image information into clear gray-scale image information. This will not be elaborated further.

[0054] Specifically, the process of gray-scale conversion using image processing software in gray-scale processing includes determining the weighted average gray-scale formula Gray = w1*R + w2*G + w3*B, where w1, w2, and w3 are the weighting coefficients of the red, green, and blue channels respectively, and w1 + w2 + w3 = 1. Gray is the gray value of the pixel point after gray-scale conversion by the weighted average method, and R, G, and B respectively represent the values of the red (Red), green (Green), and blue (Blue) color channels of the pixel point in the color image information.

[0055] Specifically, the construction of the tooth model in the step S3 is the construction of a single tooth model. The construction process of the tooth model can be as follows: based on the preprocessed grayscale image information, a single tooth model is constructed, and during the construction process, a construction relationship between the collected grayscale value and the defect depth is established. Specifically, in the data set preparation stage for model construction, for each grayscale image information with defects, the position and type of the defects are marked, and the actual depth h of the accurately measured defects is measured. At the same time, the grayscale value b is extracted from the grayscale image information. When training the model, b is used as the input feature. Using a deep learning model, a convolutional neural network, a multi-input network structure is designed. One input branch processes the conventional image space features, which is achieved by extracting through convolutional layers. The other input branch specifically processes the grayscale value feature b. The output of the network is the predicted defect depth h. Through data training, the model automatically learns the relationship between the grayscale value b and the defect depth h to determine the relevant influence coefficient k, where h = k * b.

[0056] Specifically, the relevant influence coefficient is used to represent the mapping relationship between the grayscale value of the grayscale image information and the defect depth of the tooth model during the construction process of the tooth model.

[0057] The process of constructing a single tooth model can include: dividing the training set, validation set, and test set from the data set; the training set is used for model training, the validation set is used to adjust model parameters to prevent overfitting, and the test set is used to evaluate the final performance of the model. Further enhancement processing is performed on the images in the data set, such as randomly rotating and flipping, to increase data diversity and improve the generalization ability of the model. Select a deep learning architecture, such as a convolutional neural network, define the loss function and optimizer, input the training set data into the model, and adjust the model parameters through the backpropagation algorithm to continuously reduce the loss function value of the model on the training set. During the training process, the performance of the model is regularly evaluated using the validation set, and hyperparameters such as the learning rate are adjusted to prevent model overfitting; the grayscale image information is input into the trained model to obtain the completed tooth model, which will not be elaborated here.

[0058] Specifically, in the step S4, the process of determining the defect feature surface ratio is to identify the defect areas in the tooth model, calculate the ratio of the total area of each defect area to the surface area of the tooth model to obtain the defect feature surface ratio;

[0059] Based on the defect feature surface ratio, it is determined whether the model construction is qualified, including:

[0060] If the defect feature surface ratio is less than or equal to the first preset defect ratio, it is determined that the model construction is qualified, and an indication information of qualified teeth is sent out;

[0061] If the surface proportion of the defect features is less than or equal to the second preset defect proportion and greater than the first preset defect proportion, determine whether the model construction is qualified based on the defect geometric distribution value;

[0062] If the surface proportion of the defect features is greater than the second preset defect proportion, determine that the model construction is abnormal, and adjust the acquisition parameters of the tooth model based on the defect proportion deviation.

[0063] Specifically, the first preset defect proportion B1 is selected in the interval [0.01, 0.03], and the second preset defect proportion B2 is selected in the interval [0.05, 0.07].

[0064] Specifically, the process of determining whether the model construction is qualified based on the defect geometric distribution value includes:

[0065] Establish a space rectangular coordinate system with the centroid of the tooth model as the origin, obtain the geometric center points of each defect area, and for a single geometric center point, record the distance between this geometric center point and the geometric center point with the smallest spatial distance from it as the distribution distance for this center point. Solve the average value of the distribution distances of each geometric center point to obtain the defect geometric distribution value;

[0066] If the defect geometric distribution value is greater than the preset defect geometric distribution value, determine that the model construction is abnormal, and adjust the weighting coefficient of a single color channel to the corresponding value during the preprocessing of the image information based on the average area of each defect area;

[0067] If the defect geometric distribution value is less than or equal to the preset defect geometric distribution value, determine that the model construction is qualified, and send an indication message that there are defects in the teeth.

[0068] Specifically, based on the surface proportion of defect features, it is determined whether the model construction is qualified. The surface proportion of defect features characterizes the proportion of the defect area. When the surface proportion of defect features is less than or equal to the second preset defect proportion and greater than the first preset defect proportion, it is comprehensively determined whether the model construction is qualified based on the defect geometric distribution value. The defect geometric distribution value characterizes the distribution of the defect area. When the defect geometric distribution value is less than or equal to the preset defect geometric distribution value, the distribution is concentrated and the average distribution distance is small. In this case, the defects of the teeth are concentrated and the construction of the model is stable. At this time, it is determined that the teeth have defects. When the defect geometric distribution value is greater than the preset defect geometric distribution value, the distribution is dispersed and the average distribution distance is large. At this time, the image preprocessing parameters are abnormal. Image grayscale conversion is the process of converting a color image into a grayscale image with only brightness information. In a tooth image, grayscale conversion can highlight the brightness changes on the tooth surface, while tooth defects such as cavities and cracks will cause the local brightness to be different from the normal area. In this case, due to improper setting of the grayscale parameters, it is impossible to accurately highlight the brightness difference, resulting in difficult accurate identification of the defect area and misjudgment in a large range of multiple areas. In this case, for the weighted average method of grayscale conversion, the weighting coefficients of the red, green, and blue channels are adjusted to increase the difference in grayscale values between the normal area and the defect area, and reduce the probability of being misjudged as a defect in subsequent processing. The smaller the average area of each defect area, the more dispersed the area determined as a defect feature, and the greater the adjustment amplitude of the weighting coefficient adjustment. While improving the accuracy of defect tooth detection, it further effectively improves the detection efficiency of defect teeth.

[0069] Specifically, the preset defect geometric distribution value is selected within the interval [0.11L0, 0.2L0], where L0 is the maximum length value around the outer periphery of the tooth model.

[0070] Specifically, when the adjustment of the weighting coefficient for a single color channel is completed, it is determined whether the model construction is qualified based on the newly obtained surface proportion of defect features. When it is still determined to adjust the weighting coefficient of a single color channel to the corresponding value, the weighting coefficient of the unadjusted color channel is adjusted; until it falls into other judgment conditions.

[0071] Specifically, based on the average area of each defect area, the weighting coefficient of a single color channel is adjusted to the corresponding value, where:

[0072] The increase amplitude of the weighting coefficient of a single color channel is proportional to the average area of each defect area.

[0073] In this implementation, optionally,

[0074] The average area is compared with the first preset area and the second preset area;

[0075] If the average area is less than or equal to the first preset area, adjust the weighting coefficient of a single color channel to 1.29 times the initial weighting coefficient;

[0076] If the average area is less than or equal to the second preset area and greater than the first preset area, adjust the weighting coefficient of a single color channel to 1.21 times the initial weighting coefficient;

[0077] If the average area is greater than the second preset area, adjust the weighting coefficient of a single color channel to 1.11 times the initial weighting coefficient;

[0078] When the adjustment of the weighting coefficient for a single color channel is completed, proportionally reduce the weighting coefficients of the other two channels accordingly to ensure that w1 + w2 + w3 = 1, and distribute the amount to be reduced according to the original proportional relationship of the weighting coefficients of the other two channels.

[0079] The first preset area is taken as 0.14S0, and the second preset area is taken as 0.23S0, where S0 is the total area of the tooth model.

[0080] Specifically, the process of adjusting the acquisition parameters of the tooth model based on the defect ratio deviation amount includes:

[0081] Record the difference between the calculated ratio of the defective feature surface and the second preset defect ratio as the defect ratio deviation amount;

[0082] If the defect ratio deviation amount is less than or equal to the first preset defect ratio deviation amount, continue to use the current acquisition parameters of the tooth model and send an indication message that the tooth has a defect;

[0083] If the defect ratio deviation amount is less than or equal to the second preset defect ratio deviation amount and greater than the first preset defect ratio deviation amount, adjust the acquisition parameters of the tooth model based on the volume of the obtained defective area;

[0084] If the defect ratio deviation amount is greater than the second preset defect ratio deviation amount, adjust the brightness during the process of collecting image information to the corresponding value based on the area difference ratio.

[0085] Specifically, the first preset defect ratio deviation amount Z1 is taken as 2.7B2, and the second preset defect ratio deviation amount Z2 is taken as 6.4B2.

[0086] Specifically, the process of adjusting the acquisition parameters of the tooth model based on the volume of the obtained defective area includes:

[0087] Record the sum of the volumes of the calculated and statistically analyzed defective areas as the defect volume parameter;

[0088] If the defect volume parameter is less than or equal to the preset defect volume parameter, continue to use the current acquisition parameters of the tooth model and send an indication message that the tooth has a defect;

[0089] If the defective body parameter is greater than the preset defective body parameter, the relevant influence coefficient that maps the relationship between the gray value and the defect depth in the process of constructing the tooth model based on the gray image information is adjusted to the corresponding value based on the defective body parameter.

[0090] Specifically, the preset defective body parameter Q0 is selected within the interval [0.3V0, 0.4V0], where V0 is the total volume of the tooth model.

[0091] Specifically, the relevant influence coefficient is adjusted to the corresponding value based on the defective body parameter, where

[0092] The reduction amplitude of the relevant influence coefficient is proportional to the defective body parameter.

[0093] In this embodiment, optionally,

[0094] The defective body parameter is compared with the first preset defect comparison threshold and the second preset defect comparison threshold;

[0095] If the defective body parameter is less than or equal to the first preset defect comparison threshold, the relevant influence coefficient is adjusted to 0.92 times the initial relevant influence coefficient;

[0096] If the defective body parameter is less than or equal to the second preset defect comparison threshold and greater than the first preset defect comparison threshold, the relevant influence coefficient is adjusted to 0.84 times the initial relevant influence coefficient;

[0097] If the defective body parameter is greater than the second preset defect comparison threshold, the relevant influence coefficient is adjusted to 0.77 times the initial relevant influence coefficient;

[0098] The first preset defect comparison threshold is taken as 1.5Q0, and the second preset defect comparison threshold is taken as 2Q0.

[0099] Specifically, the brightness in the process of collecting image information is adjusted to the corresponding value based on the area difference ratio, where

[0100] The ratio of the defect occupancy deviation to the second preset defect occupancy deviation is solved to obtain the area difference ratio;

[0101] The increase amplitude of the brightness is proportional to the area difference ratio.

[0102] In this embodiment, optionally,

[0103] The area difference ratio is compared with the first preset difference ratio and the second preset difference ratio;

[0104] If the area difference ratio is less than or equal to the first preset difference ratio, the brightness is adjusted to 1.13 times the initial brightness;

[0105] If the area difference ratio is less than or equal to the second preset difference ratio and greater than the first preset difference ratio, the brightness is adjusted to 1.23 times the initial brightness;

[0106] If the area difference ratio is greater than the second preset difference ratio, the brightness is adjusted to 1.28 times the initial brightness;

[0107] The first preset difference ratio is 1.3Z2, and the second preset difference ratio is 1.8Z2.

[0108] Specifically, in the process of adjusting the acquisition parameters of the tooth model by the defect ratio deviation, when the defect ratio deviation is less than or equal to the first preset defect ratio deviation, the defect area of ​​the tooth is within the normal range, and a defect notification is issued; when the defect ratio deviation is less than or equal to the second preset defect ratio deviation and greater than the first preset defect ratio deviation, the acquisition parameters of the tooth model are adjusted in combination with the volume of the defect area; when the defect volume parameter is greater than the preset defect volume parameter, at this time, there is a misjudgment in the identification of the defect due to a deviation in the process of model construction. In view of the situation that the defect area is large due to the large defect volume construction, the model construction parameters are adjusted; when the defect ratio deviation is greater than the second preset defect ratio deviation, at this time, there are too many misjudgments due to problems with image acquisition, and anomalies occur in the most basic acquisition stage. There are problems with the acquisition of the original image information. At this time, the brightness in the image acquisition process is adjusted, and the misjudgment of the shadow part in the image acquisition process is eliminated by adjusting the brightness; based on the detection results, the abnormal model construction is processed to improve the accuracy of defective tooth detection, thereby improving the detection efficiency of defective teeth.

[0109] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

[0110] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for identifying dental defects in children, characterized in that: include: S1, collect image information of teeth; S2, preprocessing the acquired image information to obtain grayscale image information; S3, constructing a tooth model based on grayscale image information; S4, determining the defect feature surface ratio based on the tooth model, and determining whether the model construction is qualified based on the defect feature surface ratio and the defect geometric distribution value, including: Determine whether the model construction is abnormal, and adjust the acquisition parameters of the tooth model based on the defect ratio deviation, including adjusting the brightness in the process of collecting image information to a corresponding value, or adjusting the relevant influence coefficient used to map the relationship between the grayscale value of the grayscale image information and the defect depth in the process of building the tooth model based on the grayscale image information to a corresponding value; Or, determine that the model construction is qualified and issue an indication that the teeth are qualified.

2. The method for identifying dental defects in children according to claim 1, characterized in that: In said S4, the process of determining the defect feature surface ratio is to identify defect regions in the tooth model, calculate the ratio of the total area of ​​each defect region to the surface area of ​​the tooth model, and obtain the defect feature surface ratio; Determine whether the model construction is qualified based on the defect feature surface ratio, including: If the defect feature surface ratio is less than or equal to the first preset defect ratio, the model construction is determined to be qualified, and an indication message that the tooth is qualified is issued; If the defect feature surface ratio is less than or equal to the second preset defect ratio and greater than the first preset defect ratio, determining whether the model construction is qualified based on the defect geometric distribution value; If the defect feature surface ratio is greater than the second preset defect ratio, the model construction is determined to be abnormal, and the acquisition parameters of the tooth model are adjusted based on the defect ratio deviation.

3. The method for identifying dental defects in children according to claim 2, characterized in that: The process of determining whether the model construction is qualified based on the defect geometric distribution value includes: A spatial rectangular coordinate system is established with the centroid of the tooth model as the origin, and the geometric center point of each defect area is obtained. For a single geometric center point, the distance between the geometric center point and the geometric center point with the smallest spatial distance from the center point is recorded as the distribution distance for the center point, and the average value of the distribution distance of each geometric center point is solved to obtain the defect geometric distribution value; If the defect geometric distribution value is greater than the preset defect geometric distribution value, the model construction is judged to be abnormal, and the weighting coefficient of a single color channel is adjusted to a corresponding value during the preprocessing of the image information based on the average area of ​​each defect area; If the defect geometry distribution value is less than or equal to the preset defect geometry distribution value, the model construction is determined to be qualified, and an indication message indicating that a tooth defect exists is issued.

4. The method for identifying dental defects in children according to claim 3, characterized in that: The weighting coefficients of individual color channels are adjusted to corresponding values ​​based on the average area of ​​each defect region, where: The increase in the weighting coefficient of a single color channel is proportional to the average area of ​​each defect region.

5. The method for identifying dental defects in children according to claim 4, characterized in that: The process of adjusting the acquisition parameters of the tooth model based on the defect ratio deviation includes: The difference between the calculated defect characteristic surface ratio and the second preset defect ratio is recorded as the defect ratio deviation; If the defect ratio deviation is less than or equal to the first preset defect ratio deviation, the acquisition parameters of the current tooth model are continuously used, and indication information indicating that there are defects in the teeth is issued; If the defect ratio deviation is less than or equal to the second preset defect ratio deviation and greater than the first preset defect ratio deviation, adjusting the acquisition parameters of the tooth model based on the acquired volume of the defect area; If the defect ratio deviation is greater than the second preset defect ratio deviation, the brightness during the image information acquisition process is adjusted to a corresponding value based on the area difference ratio.

6. The method for identifying dental defects in children according to claim 5, characterized in that: The process of adjusting acquisition parameters of the tooth model based on the acquired volume of the defect area includes: The sum of the volumes of each defect area calculated and counted is recorded as the defect volume parameter; If the defect volume parameter is less than or equal to the preset defect volume parameter, the acquisition parameters of the current tooth model are continuously used, and an indication message indicating that a tooth defect exists is issued; If the defect volume parameter is greater than the preset defect volume parameter, the relevant influence coefficient used to map the relationship between the grayscale value of the grayscale image information and the defect depth in the process of building a tooth model based on the grayscale image information is adjusted to a corresponding value based on the defect volume parameter.

7. The method for identifying dental defects in children according to claim 6, characterized in that: Based on the defect volume parameters, the relevant influence coefficients are adjusted to corresponding values, where: The reduction of the relevant influence coefficient is proportional to the defect volume parameter.

8. The method for identifying dental defects in children according to claim 7, characterized in that: The brightness in the process of collecting image information is adjusted to the corresponding value based on the area difference ratio, where: Solve the ratio of the defect proportion deviation to the second preset defect proportion deviation to obtain the area difference ratio; The increase in brightness is proportional to the area difference ratio.

Images