Dental arch line generation method based on dental arch region skeleton branch deletion

Abstract

The application provides a dental arch line generation method based on dental arch region skeleton branch deletion, which comprises the following steps: extracting a dental arch region skeleton, generating a foreground point dictionary, finding out end points and branch points, deleting skeleton branches according to conditions, and generating dental arch lines through equi-arc length sampling; according to the characteristics of the dental arch region skeleton branches, the application can automatically process complex and changeable skeleton branches, quickly find and delete the branches, and generate dental arch lines through the equi-arc length sampling method; the dental arch line generation method provided by the application does not need manual intervention, can quickly process skeleton branches, and has high accuracy of the generated dental arch lines.

Description

TECHNICAL FIELD

[0001] The application relates to a dental arch line generation method based on deletion of dental arch region skeleton branches, and belongs to the field of computer image processing. BACKGROUND

[0002] With the continuous development and progress of oral medical technology, the application of oral cone beam computed tomography (CBCT) in the clinic is also more and more extensive, for example, used for preoperative and postoperative analysis of dental implant surgery, orthodontic analysis and oral disease diagnosis. In the process of oral diagnosis and treatment, in order to better observe the oral state of the patient, the oral panoramic film needs to be observed to obtain more information, and the CBCT image can generate a panoramic film through image processing technology, and the panoramic film needs to be generated first from the CBCT image, and then the accurate dental arch line is extracted.

[0003] The traditional method for obtaining the dental arch line is to rely on the clinical experience of the oral doctor, to manually draw the points of the dental arch line, and then the oral image processing software automatically fits the dental arch line according to the points drawn by the doctor. This method is very time-consuming and labor-intensive for the doctor, and requires high clinical experience of the doctor, and the accuracy cannot be guaranteed, and the dental arch line generation efficiency is low. Due to the influence of factors such as non-smooth edges in the dental arch region, the generated dental arch region skeleton branches are also various, and there is no special method for deleting the complex and variable branches of the dental arch region skeleton.

[0004] Therefore, in order to solve the above technical problems, it is very necessary to provide a dental arch line generation method based on deletion of dental arch region skeleton branches to solve the deficiencies of the prior art. SUMMARY

[0005] The purpose of the present application is to provide a dental arch line generation method based on deletion of dental arch region skeleton branches, which realizes the full automation of dental arch line generation by automatically extracting the dental arch region skeleton, finding out the branch point and the end point, and deleting the dental arch region skeleton branch.

[0006] To achieve the above purpose, the present application provides the following technical scheme: a dental arch line generation method based on deletion of dental arch region skeleton branches, a dental arch region image with points with a pixel value of 1 as foreground points and points with a pixel value of 0 as background points generated based on an adaptive method, and the dental arch line generation is realized according to the following steps:

[0007] Step 1: create a temporary image with the same size as the dental arch region image, copy the content of the dental arch region image into the temporary image to obtain a first image; and starting from the periphery of the dental arch region in the first image, traverse each point of the first image, replace the foreground points meeting the conditions as background points according to the following steps 100 to 111, and iteratively operate in a loop until there is no foreground point that can be replaced in the first image, i.e. the dental arch region skeleton extraction is completed;

[0008] Step 100: randomly select a point in the first image as a current point, and enter step 101 to start traversing;

[0009] Step 101: judge the pixel value of the current point, if the pixel value is 1, enter step 103; if the pixel value is 0, enter step 102;

[0010] Step 102: judge whether the traversal is completed, if the traversal is completed, enter step 105; if the traversal is not completed, update the next point as the current point and enter step 101;

[0011] Step 103: in the nine-square grid points centered on the current point, judge whether the adjacent points of the current point meet conditions A, B, C and D at the same time, if they meet the conditions at the same time, enter step 104, if they do not meet the conditions at the same time, enter step 102;

[0012] Condition A: the sum of the pixel values of the 8 adjacent points is greater than or equal to 2 and less than or equal to 6;

[0013] Condition B: traverse the 8 adjacent points clockwise, the number of times that the pixel values of two consecutive points are 0 and 1 in turn is 1;

[0014] Condition C: among the points directly above, directly right and directly below the current point, at least one point has a pixel value of 0;

[0015] Condition D: among the points directly right, directly below and directly left of the current point, at least one point has a pixel value of 0;

[0016] Step 104: mark the current point as deletable, and enter step 102;

[0017] Step 105: update the pixel value of all points marked as deletable in the first image to 0 to obtain a second image, and enter step 106;

[0018] Step 106: randomly select a point in the second image as a current point, and enter step 101 to traverse again;

[0019] Step 107: judge the pixel value of the current point, if the pixel value is 1, enter step 109; if the pixel value is 0, enter step 108;

[0020] Step 108: Determine if the traversal is complete. If the traversal is complete, proceed to step 111; if the traversal is not complete, update the next point to the current point and proceed to step 107.

[0021] Step 109: In the 3x3 grid centered on the current point, the current point has 8 neighboring points. Determine whether the neighboring points of the current point simultaneously satisfy conditions A, B, E and F. If they all satisfy conditions, proceed to step 110; otherwise, proceed to step 108.

[0022] Condition A: The sum of the pixel values ​​of the 8 adjacent points is greater than or equal to 2 and less than or equal to 6;

[0023] Condition B: Traverse 8 adjacent points clockwise, and the number of times two consecutive points have pixel values ​​of 0 and 1 respectively is 1;

[0024] Condition E: Among the points directly above, to the right, and to the left of the current point, at least one point has a pixel value of 0;

[0025] Condition F: Among the points directly above, below, and to the left of the current point, at least one point has a pixel value of 0;

[0026] Step 110: Mark the point as deletable, proceed to step 108;

[0027] Step 111: In the second image, update the pixel values ​​of all points marked as removable to 0 to obtain the third image, which is the skeleton image of the dental arch region, and then proceed to step 2;

[0028] Step 2: Traverse the dental arch region skeleton image, find all foreground points with a pixel value of 1 and record their coordinates. Traverse each foreground point and record the coordinates of the foreground points in its 8 neighboring points to generate a foreground point dictionary.

[0029] Step 3: Based on the foreground point dictionary obtained in Step 2, find the endpoints and bifurcation points; where each endpoint has only one foreground point among its adjacent points; and bifurcation points are found from foreground points with at least three adjacent points.

[0030] Step 4: Based on the foreground points and their adjacent foreground points recorded in the foreground point dictionary, starting from any bifurcation point, connect the bifurcation point to each adjacent point to form a branch. Then, connect the adjacent points of each foreground point in turn along the direction of each branch. Each branch extends gradually until the adjacent point is another bifurcation point or endpoint. The branch is then extended. Traverse all foreground points to obtain all branches of the dental arch region skeleton. Among them, branches with both ends being bifurcation points are bifurcation point branches; branches with one end being a bifurcation point and the other end being an endpoint are endpoint branches.

[0031] Step 5: based on all branches of the dental arch region skeleton obtained in step 4, the branches are deleted case by case, and finally a new dental arch region skeleton without branch points and only two end points is obtained;

[0032] Step 6: all foreground points contained in the new dental arch region skeleton are sequentially numbered, starting from one end point, numbered 1, and the other end point numbered m , all foreground points contained in the new dental arch region skeleton are changed into ordered points, and the coordinates of each foreground point are obtained from the foreground point dictionary to form an ordered foreground point list;

[0033] Step 7: based on the preset number of sampling points n , the new dental arch region skeleton is divided into n samples, and the sample length is calculated by the following formula a :

[0034] a =[ m / n ]

[0035] wherein, m is the number of foreground points contained in the new dental arch region skeleton; the operation x represents taking the integer part of x ;

[0036] Step 8: one sample point is taken at the same position of each sample, and the corresponding coordinates are obtained in the ordered foreground point list, that is, the n sample points and their coordinates are obtained, and the n sample points are subjected to spline interpolation for a preset number of times to obtain a smooth curve, which is the dental arch line.

[0037] Further, the method for finding the branch point in step 3 is as follows:

[0038] Step 301: find out the foreground points with 3 or more neighbor points, and identify the coordinates of each foreground point respectively x , y ;

[0039] Step 302: compare the x values of each foreground point two by two, x the points with a difference within 2 are divided into the same group, and the same point does not participate in grouping repeatedly;

[0040] Step 303: judge the number of points in each group, if the number of points in the group exceeds 4, the points in the group are further grouped, the y values of the coordinates of each point in the group are compared two by two, y the points with a difference within 2 are divided into the same group, and the same point does not participate in grouping repeatedly;

[0041] Step 304: find out the bifurcation point from each group respectively; if the group has only one point, the point is the bifurcation point; if the number of points in the group is two or more, take the most frequent x value and the most frequent y value in the group, and the combined coordinate point is the bifurcation point of the group; if the x value or y value in the group has no repeated value, the group has no bifurcation point.

[0042] Further, the step of deleting branches according to conditions in the foregoing step 5 is specifically as follows:

[0043] Step 501: count the number of bifurcation points in the dental arch region skeleton, if there is no bifurcation point in the dental arch region skeleton, no branch needs to be deleted, and step 508 is entered; if there is at least one bifurcation point in the dental arch region skeleton, step 502 is entered;

[0044] Step 502: arbitrarily select one bifurcation point as a current point, and step 503 is entered to start traversal;

[0045] Step 503: extract a branch with the current point as a starting point, if there are at least two bifurcation point branches in the branch, step 504 is entered; if there is only one bifurcation point branch in the branch, step 505 is entered; if there is no bifurcation point branch in the branch, step 506 is entered;

[0046] Step 504: in the bifurcation point branch with the current point as a starting point, it is judged whether there is a loop branch; if there is a loop branch, the shortest bifurcation point branch is retained, and other bifurcation point branches are deleted; if there is no loop branch, no deletion is made; wherein the branches with the same starting point and ending point are loop branches; in the endpoint branch with the current point as a starting point, if the y value of the endpoint coordinate is less than a preset segmentation value, the endpoint branch is deleted; and then step 507 is entered;

[0047] Step 505: it is judged whether there is a neighboring endpoint branch in the endpoint branch with the current point as a starting point; if there is a neighboring endpoint branch, the longer endpoint branch is retained, and the shorter endpoint branch is deleted; if there is no neighboring endpoint branch, no deletion is made; wherein the two endpoint branches are neighboring endpoint branches if the x value of the endpoint coordinates of the two endpoint branches differ by within 1 / 2 of the image width, and the y value differs by within 1 / 2 of the image height; the image width is calculated by the difference between the maximum x value and the minimum x value in the foreground point list, and the image height is calculated by the difference between the maximum y value and the minimum yThe difference of values is calculated; then step 507 is entered;

[0048] Step 506: compare the lengths of each end point branch with the current point as the starting point, keep the two longest end point branches, and delete other end point branches; then step 507 is entered;

[0049] Step 507: judge whether the traversal is completed, if the traversal is completed, update the dental arch region skeleton and the foreground point dictionary, and then step 501 is entered; if the traversal is not completed, update the next bifurcation point as the current point, and enter step 503;

[0050] Step 508: record the current dental arch region skeleton as a new dental arch region skeleton.

[0051] Further, the preset segmentation value in step 504 is 4 / 5 of the image height.

[0052] The dental arch line generation method based on dental arch region skeleton branch deletion has the following technical effects compared with the prior art by adopting the above technical scheme:

[0053] 1. The dental arch region skeleton branch deletion method is specially designed for the dental arch region skeleton, can process complex and variable skeleton branches according to the characteristics of the dental arch region skeleton branches, and the generated dental arch line has high accuracy;

[0054] 2. The present application records all foreground points and neighbor points through the foreground point dictionary, can quickly find and delete branches, and improves the speed of generating dental arch lines;

[0055] 3. The present application adopts the equal arc length sampling method, and the generated dental arch line has high fitting degree;

[0056] 4. The dental arch line generation process proposed by the present application is completely automatic and does not need human intervention, and has high robustness. BRIEF DESCRIPTION OF DRAWINGS

[0057] Figure 1 is a schematic diagram of the dental arch line generation method based on dental arch region skeleton branch deletion of the present application;

[0058] Figure 2 is the first image obtained from the image of the dental arch region in the embodiment of the present application;

[0059] Figure 3 is the dental arch region skeleton graph extracted in the embodiment of the present application;

[0060] Figure 4 is a schematic diagram of the end point and bifurcation point of the dental arch region skeleton in the embodiment of the present application; wherein the circle is the position of the end point, and the rectangular frame is the position of the bifurcation point;

[0061] Figure 5 is a skeleton graph of an embodiment of the present application, which contains loop branches in the dental arch region skeleton;

[0062] Figure 6 is a skeleton graph of an embodiment of the present application, which contains loop branches in the dental arch region skeleton;

[0063] Figure 7 is a dental arch curve generated by equi-arc length sampling of an embodiment of the present application; in the figure, the numbers mark the positions of the sampling points. DETAILED DESCRIPTION

[0064] In order to better understand the technical content of the present application, specific embodiments are described below with reference to the accompanying drawings.

[0065] Aspects of the present application are described in the detailed description, drawings, and claims set forth below. The embodiments of the present application are not limited to the described embodiments but are applicable to any of the concepts and embodiments described above and in the detailed description set forth below. In addition, some aspects of the present application can be used separately or in any appropriate combination with other aspects of the present application.

[0066] In the present embodiment, a dental arch line generation method based on dental arch region skeleton branch deletion, a dental arch region image with points with a pixel value of 1 as foreground points and points with a pixel value of 0 as background points is generated based on an adaptive method, as shown in Figure 1 The dental arch line generation is realized by the following steps:

[0067] Step 1: Extract the dental arch region skeleton;

[0068] As shown in Figure 2 , a temporary image with the same size as the dental arch region image is created, the content of the dental arch region image is copied into the temporary image to obtain a first image; and the dental arch region skeleton as shown in Figure 3 is extracted from the first image;

[0069] The extraction of the dental arch region skeleton specifically includes the following steps:

[0070] Step 100: Arbitrarily select a point in the first image as a current point and enter step 101 to start traversal;

[0071] Step 101: Determine the pixel value of the current point, if the pixel value is 1, enter step 103; if the pixel value is 0, enter step 102;

[0072] Step 102: Determine if the traversal is complete. If the traversal is complete, proceed to step 105; if the traversal is not complete, update the next point to the current point and proceed to step 101.

[0073] Step 103: In the 3x3 grid centered on the current point, the current point has 8 neighboring points. Determine whether the neighboring points of the current point simultaneously satisfy conditions A, B, C and D. If they all satisfy conditions A, B, C and D, proceed to step 104; otherwise, proceed to step 102.

[0074] Condition A: The sum of the pixel values ​​of the 8 adjacent points is greater than or equal to 2 and less than or equal to 6;

[0075] Condition B: Traverse 8 adjacent points clockwise, and the number of times two consecutive points have pixel values ​​of 0 and 1 respectively is 1;

[0076] Condition C: Among the points directly above, to the right, and below the current point, at least one point has a pixel value of 0;

[0077] Condition D: Among the points directly to the right, directly below, and directly to the left of the current point, at least one point has a pixel value of 0;

[0078] Step 104: Mark the current point as deletable, proceed to step 102;

[0079] Step 105: In the first image, update the pixel values ​​of all points marked as deletable to 0 to obtain the second image, and proceed to step 106;

[0080] Step 106: Select any point in the second image as the current point, and proceed to step 101 to traverse again;

[0081] Step 107: Determine the pixel value of the current point. If the pixel value is 1, proceed to step 109; if the pixel value is 0, proceed to step 108.

[0082] Step 108: Determine if the traversal is complete. If the traversal is complete, proceed to step 111; if the traversal is not complete, update the next point to the current point and proceed to step 107.

[0083] Step 109: In the 3x3 grid centered on the current point, the current point has 8 neighboring points. Determine whether the neighboring points of the current point simultaneously satisfy conditions A, B, E and F. If they all satisfy conditions, proceed to step 110; otherwise, proceed to step 108.

[0084] Condition A: The sum of the pixel values ​​of the 8 adjacent points is greater than or equal to 2 and less than or equal to 6;

[0085] Condition B: Traverse 8 adjacent points clockwise, and the number of times two consecutive points have pixel values ​​of 0 and 1 respectively is 1;

[0086] Condition E: At least one of the pixel values of the points directly above, directly right and directly left of the current point is 0;

[0087] Condition F: At least one of the pixel values of the points directly above, directly below and directly left of the current point is 0;

[0088] Step 110: Mark the point as deletable, and go to step 108;

[0089] Step 111: Update the pixel values of all the points marked as deletable in the second image to 0, as shown in Figure 3 , to obtain a third image, which is the skeleton image of the dental arch region.

[0090] Second step: Generate a foreground point dictionary;

[0091] Traverse the skeleton image of the dental arch region, find all foreground points with pixel value 1 and record the coordinates, traverse each foreground point, and record the coordinates of the foreground points in its 8 adjacent points to generate a foreground point dictionary.

[0092] Third step: Based on the foreground point dictionary, find the end points and branch points, and then obtain all branches of the skeleton of the dental arch region;

[0093] As shown in Figure 4 , the end point marked with a circle has only one adjacent foreground point; the branch point marked with a rectangular frame is found from the points with at least 3 adjacent foreground points;

[0094] The step of finding the branch point is specifically as follows:

[0095] Step 301: Find the foreground points with 3 or more adjacent points, and identify the coordinates of each foreground point, x , y respectively;

[0096] Step 302: Compare the x values of each foreground point with each other, x the points with a difference within 2 are grouped into the same group, and the same point does not participate in grouping repeatedly;

[0097] Step 303: Determine the number of points in each group, if the number of points in the group exceeds 4, re-group the points in the group, and compare the y values of the coordinates of each point in the group with each other, y the points with a difference within 2 are grouped into the same group, and the same point does not participate in grouping repeatedly;

[0098] Step 304: find out the bifurcation point from each group respectively; if the group has only one point, the point is the bifurcation point; if the number of points in the group is two or more, the coordinate point with the most frequent value and the most frequent value in the group is combined, and the combined coordinate point is the bifurcation point of the group; if the values of the coordinate points in the group are not repeated, the group has no bifurcation point. x y x y

[0099] Further, based on each foreground point and its adjacent foreground point recorded in the foreground point dictionary, from any bifurcation point, the bifurcation point and each adjacent point are connected to form a branch, and then the adjacent points of each foreground point are connected in turn along the direction of each branch, and each branch is gradually extended until the adjacent point is another bifurcation point or an end point, and the branch is completed. The extension; traverse all foreground points to obtain all branches of the dental arch region skeleton, as shown in Figure 5 The branch with both ends being bifurcation points is a bifurcation point branch, and the branch with one end being a bifurcation point and the other end being an end point is an end point branch.

[0100] Fourth step: deleting skeleton branches according to conditions;

[0101] Based on all branches of the dental arch region skeleton, the branches are deleted according to conditions, and finally a new dental arch region skeleton with no bifurcation points and only two end points is obtained;

[0102] The step of deleting branches according to conditions is specifically as follows:

[0103] Step 501: count the number of bifurcation points in the dental arch region skeleton; if there is no bifurcation point in the dental arch region skeleton, there is no need to delete the branch, and step 508 is entered; if there is at least one bifurcation point in the dental arch region skeleton, step 502 is entered;

[0104] Step 502: arbitrarily select one bifurcation point as a current point, and enter step 503 to start traversal;

[0105] Step 503: extract a branch with the current point as the starting point; if there are at least two bifurcation point branches in the branch, step 504 is entered; if there is only one bifurcation point branch in the branch, step 505 is entered; if there is no bifurcation point branch in the branch, step 506 is entered;

[0106] Step 504: in the bifurcation point branch with the current point as the starting point, it is judged whether there is a loop branch; if there is a loop branch, the shortest bifurcation point branch is retained, and other bifurcation point branches are deleted; if there is no loop branch, no deletion is made; wherein the branch with the same starting point and end point is a loop branch; in the end point branch with the current point as the starting point, if the y ​​​​If the value is less than the preset segmentation value, the endpoint branch is deleted; then step 507 is entered;

[0107] Step 505: judging whether there is a neighboring endpoint branch in the endpoint branch with the current point as the starting point, if there is a neighboring endpoint branch, the longer endpoint branch is kept and the shorter endpoint branch is deleted; if there is no neighboring endpoint branch, no deletion is made; wherein the x value of the endpoint coordinates of the two endpoint branches differs within 1 / 2 of the image width, and y the value of the endpoint coordinates of the two endpoint branches differs within 1 / 2 of the image height, the two endpoint branches are neighboring endpoint branches; the image width is calculated by the difference between the maximum x value and the minimum x value in the foreground point list, and the image height is calculated by the difference between the maximum y value and the minimum y value in the foreground point list; then step 507 is entered;

[0108] Step 506: comparing the lengths of the endpoint branches with the current point as the starting point, keeping the two longest endpoint branches and deleting the other endpoint branches; then step 507 is entered;

[0109] Step 507: judging whether the traversal is completed, if the traversal is completed, updating the dental arch region skeleton and the foreground point dictionary, and then step 501 is entered; if the traversal is not completed, updating the next bifurcation point as the current point and entering step 503;

[0110] Step 508: recording the current dental arch region skeleton as a new dental arch region skeleton.

[0111] Fifth step: generating a dental arch line through equi-arc length sampling;

[0112] Firstly, all the foreground points contained in the new dental arch region skeleton are sequentially numbered, one of the endpoints is numbered as 1 and the other endpoint is numbered as m , all the foreground points contained in the new dental arch region skeleton are changed into ordered points, and the coordinates of each foreground point are obtained from the foreground point dictionary to form an ordered foreground point list;

[0113] Then, the preset number of sampling points is 5, the new dental arch region skeleton is segmented into 5 samples, and the sample length is calculated through the following formula a :

[0114] a =[ m / 5]

[0115] wherein, n is the number of foreground points contained in the new dental arch region skeleton; the operation m represents taking the integer part of x ; x ​

[0116] Finally, one sample point is taken at the same position of each sample respectively, and the corresponding coordinates are obtained in the ordered foreground point list, that is, five sample points and their coordinates are obtained, and the five sample points are subjected to three times of spline interpolation to obtain a smooth curve as shown in FIG. 6, which is the dental arch line. Figure 7

[0117] Although the present application has been described above with reference to a preferred embodiment, it is not intended to limit the present application. Those skilled in the art, without departing from the spirit and scope of the present application, can make various modifications and improvements. Therefore, the scope of protection of the present application shall be subject to the scope defined by the claims.​

Claims

1. A dental arch line generation method based on dental arch region skeleton branch deletion, characterized by, Based on the dental arch region image generated using an adaptive method, with pixels of 1 as the foreground and pixels of 0 as the background, the dental arch line is generated according to the following steps: Step 1: Create a temporary image of the same size as the dental arch region image, copy the contents of the dental arch region image into the temporary image to obtain the first image; and traverse each point of the first image starting from the outer periphery of the dental arch region, replacing the foreground points that meet the conditions with background points according to steps 100 to 111 below, and iterate until there are no more foreground points that can be replaced in the first image, thus completing the extraction of the dental arch region skeleton. Step 100: Select any point in the first image as the current point, and proceed to Step 101 to start traversal; Step 101: Determine the pixel value of the current point. If the pixel value is 1, proceed to Step 103; if the pixel value is 0, proceed to Step 102; Step 102: Determine whether the traversal is complete. If the traversal is complete, proceed to Step 105. If the traversal is not complete, update the next point to the current point and proceed to step 101; Step 103: In the 3x3 grid centered on the current point, the current point has 8 adjacent points. Determine whether the adjacent points of the current point simultaneously satisfy conditions A, B, C, and D. If they all satisfy, proceed to step 104; otherwise, proceed to step 102; Condition A: The sum of the pixel values ​​of the 8 adjacent points is greater than or equal to 2 and less than or equal to 6; Condition B: When traversing the 8 adjacent points clockwise, the number of times two consecutive points have pixel values ​​of 0 and 1 is 1; Condition C: Among the points directly above, to the right, and below the current point, at least one point has a pixel value of 0; Condition D: At least one of the points directly to the right, directly below, and directly to the left of the current point has a pixel value of 0; Step 104: Mark the current point as deletable, proceed to step 102; Step 105: Update the pixel values ​​of all points marked as deletable in the first image to 0, obtain the second image, and proceed to step 106; Step 106: Randomly select one point in the second image as the current point, proceed to step 101 and traverse again; Step 107: Determine the pixel value of the current point. If the pixel value is 1, proceed to step 109; if the pixel value is 0, proceed to step 108; Step 108: Determine if the traversal is complete. If the traversal is complete, proceed to step 111; If the traversal is not completed, the next point is updated as the current point, and step 107 is entered; step 109: in the nine-square grid centered on the current point, there are eight adjacent points of the current point, and it is judged whether the adjacent points of the current point simultaneously satisfy conditions A, condition B, condition E and condition F, and if they are simultaneously satisfied, step 110 is entered, and if they are not simultaneously satisfied, step 108 is entered; condition A: the total sum of the pixel values of the eight adjacent points is greater than or equal to 2 and less than or equal to 6; condition B: the eight adjacent points are traversed clockwise, and the number of times that the pixel values of two points appearing in succession are 0 and 1 in sequence is 1; condition E: among the points directly above, directly to the right and directly to the left of the current point, at least one point has a pixel value of 0; condition F: among the points directly above, directly below and directly to the left of the current point, at least one point has a pixel value of 0; step 110: the point is marked as deletable, and step 108 is entered; step 111: in the second image, the pixel values of all points marked as deletable are updated to 0, a third image is obtained, which is a skeleton image of the dental arch region, and then step 2 is entered; Step 2: traverse the dental arch region skeleton image, find all foreground points with a pixel value of 1 and record the coordinates, traverse each foreground point, and record the coordinates of the foreground points in the eight adjacent points, to generate a foreground point dictionary; Step 3: based on the foreground point dictionary obtained in step 2, find the end points and bifurcation points; wherein the adjacent points of the end point only have one foreground point; the bifurcation points are found from the foreground points with at least three adjacent points; the method for finding the bifurcation points in step 3 is as follows: step 301: find the foreground points with three or more adjacent points, and identify the coordinates (x, y) of each foreground point; step 302: compare the x values of each foreground point pairwise, and group the foreground points with x values within 2 into the same group, and the same point does not participate in grouping repeatedly; step 303: judge the number of points in each group, if the number of points in each group exceeds 4, the points in the group are regrouped, and the y values of the coordinates of the points in the group are compared pairwise, and the points with y values within 2 are grouped into the same group, and the same point does not participate in grouping repeatedly; step 304: find the bifurcation points from each group respectively; if a group has only one point, the point is the bifurcation point; if the number of points in a group is two or more, the coordinate point composed of the most frequently occurring x value and the most frequently occurring y value in the group is the bifurcation point of the group; if there is no repeated x value or y value in a group, the group has no bifurcation point; Step 4: based on the foreground points and their adjacent foreground points recorded in the foreground point dictionary, starting from any bifurcation point, the bifurcation point and each adjacent point are connected to form a branch, and then each adjacent point of each foreground point is connected in turn along the direction of each branch, and each branch is gradually extended until the adjacent point is another bifurcation point or an end point, and the branch is completed; all foreground points are traversed to obtain all branches of the dental arch region skeleton; wherein the branch with two end points being bifurcation points is a bifurcation point branch; the branch with one end being a bifurcation point and the other end being an end point is an end point branch; Step 5: based on all branches of the dental arch region skeleton obtained in step 4, branches are deleted according to conditions, and finally a new dental arch region skeleton without branch points and with only two endpoints is obtained; the step of deleting branches according to conditions in step 5 is as follows: step 501: count the number of branch points in the dental arch region skeleton, if there is no branch point in the dental arch region skeleton, there is no need to delete branches, and step 508 is entered; if there is at least one branch point in the dental arch region skeleton, step 502 is entered; step 502: arbitrarily select one branch point as the current point, and step 503 is entered to start traversal; step 503: extract the branch with the current point as the starting point, if there are at least two branch points in the branch, step 504 is entered; if there is only one branch point in the branch, step 505 is entered; if there is no branch point in the branch, step 506 is entered; step 504: in the branch point branch with the current point as the starting point, it is judged whether there is a loop branch; if there is a loop branch, the shortest branch point branch is retained, and the other branch point branches are deleted; if there is no loop branch, no deletion is made; wherein the branches with the same starting point and endpoint are loop branches; in the endpoint branch with the current point as the starting point, if the y value of the endpoint coordinate is less than the preset segmentation value, the endpoint branch is deleted; then step 507 is entered; step 505: it is judged whether there is a neighboring endpoint branch in the endpoint branch with the current point as the starting point, if there is a neighboring endpoint branch, the longer endpoint branch is retained, and the shorter endpoint branch is deleted; if there is no neighboring endpoint branch, no deletion is made; wherein the two endpoint branches are neighboring endpoint branches if the x value of the endpoint coordinate of the two endpoint branches differs by within 1 / 2 of the image width, and the y value differs by within 1 / 2 of the image height; the image width is calculated from the difference between the maximum x value and the minimum x value in the foreground point list, and the image height is calculated from the difference between the maximum y value and the minimum y value in the foreground point list; then step 507 is entered; step 506: compare the lengths of each endpoint branch with the current point as the starting point, retain the longest two endpoint branches, and delete other endpoint branches; then step 507 is entered; step 507: it is judged whether the traversal is completed, if the traversal is completed, the dental arch region skeleton and the foreground point dictionary are updated, and then step 501 is entered; if the traversal is not completed, the next branch point is updated as the current point, and step 503 is entered; step 508: the current dental arch region skeleton is recorded as the new dental arch region skeleton; Step 6: all foreground points contained in the new dental arch region skeleton are sequentially numbered, one of the endpoints is numbered as 1, and the other endpoint is numbered as m, all foreground points contained in the new dental arch region skeleton are changed into ordered points, the coordinates of each foreground point are obtained from the foreground point dictionary, and an ordered foreground point list is constructed; Step 7: based on the preset number of sampling points n, the new dental arch region skeleton is segmented into n samples, and the sample length a is calculated by the following formula: a = [m / n] wherein m is the number of foreground points contained in the new dental arch region skeleton; and [x] represents the integer part of x. Step 8: 1 sample point is taken at the same position of each sample respectively, and the corresponding coordinates are obtained in the ordered foreground point list, that is, n sample points and their coordinates are obtained, and the n sample points are subjected to spline interpolation for a preset number of times to obtain a smooth curve, which is the dental arch line.

2. The dental arch line generation method based on the dental arch region skeleton branch deletion according to claim 1, characterized in that, The preset segmentation value in the step 504 is 4 / 5 of the image height.

Images