CN111222407B - Palate wrinkle identification method adopting uniform slicing and inflection point characteristic extraction - Google Patents
Palate wrinkle identification method adopting uniform slicing and inflection point characteristic extraction Download PDFInfo
- Publication number
- CN111222407B CN111222407B CN201911130386.7A CN201911130386A CN111222407B CN 111222407 B CN111222407 B CN 111222407B CN 201911130386 A CN201911130386 A CN 201911130386A CN 111222407 B CN111222407 B CN 111222407B
- Authority
- CN
- China
- Prior art keywords
- inflection point
- palate
- tangent line
- inflection
- points
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
- G06V10/462—Salient features, e.g. scale invariant feature transforms [SIFT]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
- G06V10/751—Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention belongs to the technical field of legal medical expert identification of human biological characteristics, and the specific technical scheme is as follows: a method for identifying pucker palate by adopting uniform slicing and inflection point characteristic extraction comprises the following specific steps: firstly, acquiring three-dimensional puckery palate data; uniformly slicing the three-dimensional puckery palate data, and preprocessing all slices to obtain a coordinate sequence of a puckery palate tangent line; thirdly, extracting the inflection point and corner characteristics of each slice of the tangent line of the ruffle palate, wherein the coordinate position, the inflection point and the corner of all the tangents of the ruffle palate jointly form a feature vector of the ruffle palate; comparing the test sample characteristics with the characteristics in the sample characteristic library one by one, and outputting a matching result; the method for extracting the uniform slicing and inflection point features, which is provided by the method, makes up the problem that the existing three-dimensional puckery palate data identification method is deficient, and has considerable identification effect.
Description
Technical Field
The invention belongs to the technical field of forensic identification of human body biological characteristics, and particularly relates to a pucker palate identification method by adopting uniform slicing and inflection point characteristic extraction.
Background
Identity determination is a forensic operation that infers whether an object under study is from cognitive activities of the same object by comparing and identifying similarities and differences between two or more characteristic indicators of known and unknown objects. The characteristic indexes adopted by the same identification need to meet the conditions of uniqueness, universality, permanence, collectability, identifiability and the like. At present, with the rapid development of modern forensic medicine and artificial intelligence technology, some intrinsic physiological characteristics of human (such as tooth, craniofacial shape, fingerprint, palm print, DNA, etc.) have become important same identification characteristic indexes. Although some studies have been made by the scholars on the above characteristic indexes, some disadvantages still exist. On one hand, due to the characteristic indexes of fingerprints, palm prints, craniofacial surfaces and the like of the anatomical position on the surface of a human body, the anatomical position is easily influenced by environmental factors such as fire, chemical corrosion or external trauma to generate information loss; environmental and economic factors also often limit the implementation of large-scale DNA assays in victims in major natural disasters and terrorist attacks. On the other hand, with the widespread commercial use of biometric technology (e.g., face recognition, fingerprint recognition, iris recognition, etc.), some criminals evade jurisdictions by making false fingerprints, facial cosmetic, dental surgery, etc. Therefore, the research on human characteristic indexes which are not easy to damage and forge and have low cost is a practical requirement for improving crime identification accuracy and perfecting the same identification.
The Palatal Rugae (PR), also known as the transverse Palatal folds, is an irregular soft tissue ridge located in the anterior third of the hard palate, asymmetrically distributed from the median palatine suture to the sides. Since it is located inside the oral cavity and protected by the surrounding tissues of the oral cavity such as the cheek, tongue, teeth, and maxilla and mandible, it has strong tolerance to the environment such as high temperature and trauma, and is relatively easy to preserve in cases of severe soft tissue destruction such as serious natural disasters, incineration, ossification, putrefaction, and the like. The palate winkles are formed from the third month of the embryonic stage to the final death of the human body, the palate winkles only have length change due to normal growth and development, the shape, the arrangement and the relative position of the palate winkles are always stable, the shape and the position of the palate winkles are difficult to change even if the palate winkles are subjected to operation, trauma or chemical corrosion, and the palate winkles can resist the decomposition of the putrefaction within 7 days of death and keep the shape of the palate winkles unchanged. Meanwhile, the systematic research aiming at the palate wrinkle morphology shows that the palate wrinkle is formed by controlling and developing by genetic genes, the morphology patterns of the palate wrinkle are different among different individuals, and the palate wrinkle has universality, uniqueness and stability. Pucker palate can be a single identifying characteristic index.
Although the description and analysis of the morphology of the wrinkled palate, numerous studies have been conducted by various researchers. However, due to the variability and complexity of irregular lines, many problems still exist in the same identified studies of pucker palate. First, the acquisition and storage of data on wrinkles palate is backward and lacks a simple and quick description. Scholars at home and abroad try to research the coding method of the wrinkle palate shape from different angles, but the description of the wrinkle palate shape still lacks a uniform standard due to the difference and complex variability of the wrinkle palate shape of people of different ethnic groups and countries in the world, and most researches focus on rough comparison of the wrinkle palate shape. The plaster model adopted for collecting the data of the wrinkles palate is easy to damage, difficult in physical storage, difficult in data transmission and sharing, high in time cost and the like. Second, the palatoglossus morphology description has not been digitized. At present, the palate wrinkle research mostly adopts artificial marking, the palate wrinkle line shape is drawn artificially, the palate wrinkle form is directly distinguished by eyes, marking results are easily influenced by personal understanding of a marker, and the measuring method has strong subjectivity; in addition, the manual labeling speed is slow, the accuracy is not guaranteed, and the batch processing is inconvenient; when the same identification is carried out on the pucker palate, the collected new sample can be identified only by being labeled by a labeling expert, and the real-time performance cannot be guaranteed. Third, studies relating to three-dimensional wrinkle recognition are still in the stage of onset. The three-dimensional digital analysis of the wrinkled palate morphology is still in the initial stage, a large number of relatively mature repeated researches and verifications are not carried out on a plurality of samples, and at present, no digital wrinkled palate morphology description standard which is relatively accurate and can be repeated for many times can be put into the commercial biological recognition or judicial practice (identity identification of individuals).
Disclosure of Invention
The invention provides a palate wrinkle identification method for forensic identification by adopting uniform slicing and inflection point extraction, aiming at solving the defect of the lack of the existing three-dimensional palate wrinkle identification technology, and the method is easy to understand the treatment process and good in identification effect.
In order to achieve the purpose, the technical scheme adopted by the invention is as follows: a method for identifying pucker palate by adopting uniform slicing and inflection point characteristic extraction comprises the following specific steps:
step one, acquiring three-dimensional puckery palate data information;
step two, uniformly slicing and preprocessing the three-dimensional puckery palate data to obtain a coordinate sequence of the puckery palate slice;
extracting inflection point characteristics of the tangent line of each sliced palate wrinkle, wherein the coordinate positions, inflection points and corners of all the tangent lines of the palate wrinkles jointly form a feature vector of the palate wrinkles;
and step four, matching the palate wrinkle tangent line based on inflection point feature identification, and outputting a matching result.
In the first step, the volunteer lies flat and disinfects in the mouth, the volunteer opens the mouth to the maximum opening degree, non-contact scanning is carried out on the pucker palate in the oral cavity by using 3Shape iTero, after the scanning is finished, a complete pucker palate scanning sample is led into a computer for storage, and is sent into a measuring system in stl format, and data format conversion is carried out through an Ortho Analyzer, so that three-dimensional pucker palate digital data is obtained.
The principles underlying non-contact scanning are: where the curvature is large, there are many measurement points and few data points in the uniform or smooth area.
The specific scanning sequence is as follows: firstly, from the right occlusal surface to the left occlusal surface of the upper jaw of the volunteer; second, from the left buccal surface to the right buccal surface of the upper jaw of the volunteer; and thirdly, from the right lingual side to the left lingual side of the maxillary maxilla of the volunteer.
And in the second step, uniformly slicing the three-dimensional puckery palate data, and further preprocessing the obtained slices, wherein the preprocessing comprises removing background colors, extracting a target puckery palate tangent line and calibrating a meaningful identification area of the puckery palate tangent line.
In the third step, a point with severe change in the palate fold tangent line is regarded as an inflection point of the palate fold tangent line, skeleton information of the palate fold tangent line is converted into region chain code information, and the region chain code information acquisition method comprises the following steps: processing the coordinate sequence one-dimensional array which is obtained in the step two and represents the position of the tangent line of the puccinia palatogenes, wherein L is a determined value, the unit of L is the number of pixels, and a pixel point a on the tangent line of the puccinia palatogenes isMRepresenting the Mth pixel from the left end, pixel point aM+LA pixel representing a distance L from the pixel point, pixel point aMTo pixel point aM+LThe vector direction angle of (a) is defined as a pixel point (a)MTraversing the meaningful identification area of the whole palate wrinkle tangent line to obtain all pixel point region chain code direction angles, wherein the vector direction angles form region chain code direction angle information of the palate wrinkle tangent line; preliminarily determining the inflection point of a tangent line of the pucker palate by the following method: traversing all pixel points of meaningful identification areas of palate wrinkle tangent from left end point, and defining vectors from pixel point to pixel pointThe quantity direction angle is the corner of a pixel point, if the corner of the pixel point is greater than a given threshold value by 10 degrees, the pixel point is considered as the inflection point of the palate wrinkle tangent line, and is the corner of the inflection point, all the pixel points are sequentially processed, so that the inflection point sequence of the palate wrinkle tangent line can be obtained, the inflection points are sparsely distributed in an approximately linear region on the palate wrinkle tangent line, the inflection point sequence is densely distributed in a bending region of the palate wrinkle tangent line, and when a plurality of continuous points in a certain section of tangent line are the inflection points, the representative inflection point is selected as the inflection point of the section of tangent line; screening the preliminarily determined inflection points, and removing redundant inflection points on the tangent line of the palate wrinkles, wherein the removing principle is that only one representative inflection point is considered in a certain pixel length, and the removing method comprises the following steps: sequencing all inflection points on the crease palate tangent line according to the sizes of corners, inserting the inflection points from large to small in sequence, if a certain inflection point is inserted, the inserted inflection point is not provided within the distance of 5 pixels on the left and right of the inflection point, the insertion of the inflection point is permitted, otherwise, the insertion is not performed, and the next inflection point is continuously processed, so that an inflection point sequence with the distance between the inflection points not less than 5 pixels is finally obtained, and thus, the effective inflection point of the crease palate tangent line is obtained; acquiring adjacent inflection point characteristics and separated inflection point characteristics of a crease palate tangent line, and determining the adjacent inflection point characteristics: starting from the 2 nd inflection point from left to right, if the inflection point P is the current inflection point, calculating the distance L from the inflection point P to the adjacent inflection point P1 before the inflection point PP1Distance L of P from its next adjacent inflection point P2P2Direction angle theta from inflection point P1 to inflection point PP1Direction angle θ from inflection point P to inflection point P2P2，ΔθP＝θP1-θP2And the ratio of the distances between the front and rear of the inflection point P is LP1/LP2The distance and the direction angle are adjacent inflection point characteristics; the determination method of the characteristic of inflection point separation comprises the following steps: starting from the 3 rd inflection point from left to right, if the inflection point P is the current inflection point, calculating the distance L from the inflection point P to the previous inflection point which is separated from the inflection point Q1Q1P to the next distance L from inflection point Q2Q2Direction angle theta from inflection point Q1 to inflection point QQ1Direction angle theta from inflection point Q to inflection point Q2Q2，ΔθQ＝θQ1-θQ2And the distance ratio DeltaL' between the front and back of the inflection point P is equal to LQ1/LQ2These areThe quantities are all characteristic of spaced inflection points.
In the fourth step, the characteristics of the test sample and the characteristics in the sample characteristic library are compared one by one to obtain a matching error, and if the matching error is smaller than the matching error, the matching is considered to be successful.
The method for extracting the uniform slicing and inflection point features, which is provided by the method, makes up the problem that the existing three-dimensional puckery palate data identification method is deficient, and has considerable identification effect.
Drawings
The present invention will be described in further detail with reference to the accompanying drawings.
FIG. 1 is an overall flow chart of the present invention.
Fig. 2 is a schematic representation of a three-dimensional pucker palate.
Fig. 3 is a schematic representation of a three-dimensional crumpled-palate slice.
Fig. 4 is a schematic diagram of the procedure of tangent pretreatment of the puccinia palate for each section.
Fig. 5 is a diagram showing calibration results of significant identified regions of the tangent to the pucker palate of different sections of sample No. 2.
Fig. 6 is a schematic diagram of pixel points of significant identified regions of the crease-palate tangency line of sample No. 2, slice 10.
Fig. 7 is a chain code of the palatopathic fold tangent area of sample No. 2 slice No. 10.
FIG. 8 is a schematic diagram of the initial inflection point of the crease line tangent of sample No. 2, section 10.
Fig. 9 is a graph showing effective inflection points of the crease line tangent of different slices of sample No. 2.
Fig. 10 is a graph showing inflection points of the crease palate tangent in sample No. 2, section 10.
Detailed Description
The present invention will now be described in further detail with reference to the accompanying drawings, which are simplified schematic drawings that illustrate only the basic structure of the invention and therefore show only the components that are relevant to the invention.
A method for identifying pucker palate by adopting uniform slicing and inflection point characteristic extraction comprises the following specific steps:
the method comprises the following steps: collecting three-dimensional puckery palate data, specifically, disinfecting oral instruments and collecting equipment, enabling an oral cavity volunteer to lie on a dental chair, adjusting the chair position of the volunteer to be a standard position (namely: the lower jaw of the volunteer is parallel to the ground, and the upper jaw of the volunteer is vertical to the ground), carrying out oral disinfection on the volunteer, and enabling the mouth of the volunteer to be as large as possible, as shown in fig. 1, carrying out non-contact scanning by adopting 3Shape iTero, wherein the scanning principle is that the part with large curvature has more measuring points, the uniform or smooth area data points are fewer, and the specific scanning sequence is as follows: firstly, from a right side biting occlusal surface to a left side occlusal surface of the upper jaw of the volunteer; from the left buccal surface to the right buccal surface of the upper jaw; and thirdly, from the right lingual side to the left lingual side of the upper jaw, finally checking whether the pucker palate of each face is complete, if not, performing complementary scanning, introducing a sample into a computer for storage after scanning is completed, sending the sample into a measurement system in stl format, and performing data format conversion through an ortho analyzer to obtain three-dimensional pucker palate digital data (as shown in fig. 2).
Step two: the three-dimensional wrinkle palate data is sliced and preprocessed, for convenience of identification, the three-dimensional data is uniformly sliced, the slicing effect is shown in fig. 3, and the obtained slices are preprocessed, wherein the specific process is shown in fig. 4.
The preprocessing comprises removing background color, extracting a target crease palate tangent and calibrating a meaningful identification area of the crease palate tangent. Specifically, the method comprises the following steps: the operation method for removing the background color comprises the following steps: traversing the palate wrinkle slice image by adopting a template with the size of 10 multiplied by 10, sequencing the gray values of pixels in a region to be processed, processing the pixels with the maximum gray value and the difference between the maximum gray value and the gray value less than 3 as a background, setting the gray values of all background pixels as 255, and obtaining a slice image containing a cell line and a palate wrinkle tangent line foreground; ② the operation method for extracting the target palate fold tangent line is: taking 168 as a threshold value, regarding pixels with the gray value less than 168 as palate wrinkle tangent line prospects, regarding the remaining pixels as unit grid lines, obtaining an image only containing a unit grid and only containing a curve, setting the unit grid line pixels as 0 in the unit grid image, then performing vertical projection superposition on the unit grid line pixels to obtain a projection vector, counting the position exceeding the mean value of 3 delta in the projection vector, wherein the position is the position of the vertical unit grid line, and similarly, performing horizontal projection superposition and statistics on the unit grid to obtain the position of the horizontal unit grid line. Obtaining the size of the cell according to the position of the grid line, wherein the position of the thickest cell grid line is the position of the central cell grid line; the specific operation method for calibrating the significant identification area of the palate wrinkle tangent line comprises the following steps: firstly, determining the leftmost end and the rightmost end of a curve in a foreground image of a palate wrinkle tangent line to obtain a curve width, removing 25% width parts from the left and the right respectively (eliminating the upward tilting influence of two ends of some curves), obtaining a curve vector according to the pixel points of the rest palate wrinkle tangent line parts, performing secondary curve fitting on the vector to obtain the center of a secondary curve as the center of the palate wrinkle tangent line, and taking the cell size expanded from the center to the left and the right by 1.3 times as the boundary of a curve central area, so that a meaningful identification area of the palate wrinkle tangent line is calibrated, but before identification, further pretreatment is still needed to ensure the identification effect, specifically, a palate wrinkle tangent line breakpoint needs to be bridged: judging whether the curve in the central area contains a plurality of sections of curves, if so, bridging single-pixel breakpoints, ensuring that no bifurcation or ring occurs after bridging, further refining the crease palate tangent line to obtain a curve skeleton, finding all endpoints of the curve, wherein the leftmost endpoint and the rightmost endpoint of the curve are curve endpoints, the other endpoints are bifurcation, gradually setting the other endpoints as backgrounds, sequentially inspecting the neighborhood of the curve from the left endpoint of the curve, and obtaining a coordinate sequence of the curve from one end to the other end, wherein the sequence represents the position of each pixel point on the crease palate tangent line. As shown in fig. 5, a graph showing the calibration results of the significant recognition area of the crease-palate tangent line of different slices of sample No. 2 is shown, wherein the black thin line part represents the crease-palate tangent line of different slices, the black thick line part represents the significant recognition area of the calibrated crease-palate tangent line, the semi-arc black line represents the fitted quadratic curve, and the position shown by the vertical box in the first line of the graph represents the position where the bridging of the break points is performed.
As shown in fig. 6, a schematic pixel point diagram of a significant identification region of the incision of the ruffle cut on sample No. 2, slice 10 is given.
Step three: and extracting inflection point characteristics of the cutting line of the pucker palate of each slice, and after obtaining the cutting line of the pucker palate of each slice, carrying out pucker palate identification by extracting the inflection point characteristics of the cutting line of the pucker palate. The inflection point of the tangent line of the pucker palate has the following characteristics: the inflection point sequence has invariance, and the positions, the sequence and the directions of the inflection points are basically the same, and the point with more severe change in the palate fold tangent line is regarded as the inflection point of the palate fold tangent line by the method.
(1) Converting framework information of a palate wrinkle tangent line into region chain code information, wherein the region chain code information acquisition method comprises the following steps: processing the coordinate sequence one-dimensional array which is obtained in the step two and represents the position of the tangent line of the puccinia palatogenes, wherein L is a determined value, the unit of L is the number of pixels, and a pixel point a on the tangent line of the puccinia palatogenes isMRepresenting the Mth pixel from the left end, pixel point aM+LA pixel representing a distance L from the pixel point, pixel point aMTo pixel point aM+LThe vector direction angle of (a) is defined as a pixel point (a)MThe region chain code direction angles of the palate winkle tangent line are traversed through the meaningful identification region of the whole palate winkle tangent line, the region chain code direction angles of all pixel points are obtained, and the vector direction angles form the region chain code direction angle information of the palate winkle tangent line (as shown in fig. 7).
(2) The inflection point of the tangent line of the pucker palate is preliminarily determined, and the determination method comprises the following steps: traversing all pixel points of the meaningful identification area of the palate wrinkle tangent from the left end point, and recording the pixel point aMTo pixel point aM+LHas a vector direction angle of thetaa1Recording pixel point aM+LTo pixel point aM+2LHas a vector direction angle of thetaa2Changing Δ θ to θa1-θa2Is defined as a pixel point aM+LIf pixel point a at the cornerM+LIs greater than the given threshold value by 10 deg., the pixel point a is considered to beM+LIs the inflection point of the tangent line of the palate fold, and Delta theta is the inflection point aM+LAs shown in fig. 8, a schematic diagram of inflection points preliminarily determined for the 10 th slice palate wrinkle tangent line of sample No. 2 is given, and it can be seen that these inflection points are distributed sparsely in an approximately straight region on the palate wrinkle tangent line and are distributed more densely in a curved region of the palate wrinkle tangent line, and when a continuous number of points in a certain section of tangent line are inflection points, a representative inflection point should be selected as the inflection point of the section of tangent line.
Screening out the inflection points preliminarily determined in the step (2), and removing redundant inflection points on the palate wrinkle cutting, wherein the removing principle is that only one representative inflection point is considered in a certain pixel length, and the removing method comprises the following steps: sequencing all inflection points on the palate wrinkle tangent line according to the sizes of corners, sequentially inserting the inflection points from large to small, if a certain inflection point is inserted, if no inserted inflection point exists in the distance of 5 pixels on the left and right, allowing the inflection point to be inserted, otherwise, not inserting, and continuously processing the next inflection point, so that a sparse inflection point sequence with the distance between the inflection points not less than 5 pixels is finally obtained, and thus, the effective inflection point of the palate wrinkle tangent line is obtained, and as shown in fig. 9, a schematic diagram of the effective inflection points of different slices of the palate wrinkle tangent line of a No. 2 sample is given.
(3) Acquiring adjacent inflection point characteristics and separated inflection point characteristics of a crease palate tangent line, and determining the adjacent inflection point characteristics: starting from the 2 nd inflection point from left to right, if the inflection point P is the current inflection point, calculating the distance L from the inflection point P to the adjacent inflection point P1 before the inflection point PP1Distance L of P from its next adjacent inflection point P2P2Direction angle θ from inflection point P to inflection point P2P2，ΔθP＝θP1-θP2And the ratio of the distances between the front and rear of the inflection point P is LP1/LP2These quantities are all adjacent corner features; the determination method of the characteristic of inflection point separation comprises the following steps: starting from the 3 rd inflection point from left to right, if the inflection point P is the current inflection point, calculating the distance L from the inflection point P to the previous inflection point which is separated by an inflection point Q1Q1P to the next distance L from inflection point Q2Q2Direction angle theta from inflection point Q1 to inflection point QQ1Direction angle theta from inflection point Q to inflection point Q2Q2，ΔθQ＝θQ1-θQ2And the distance ratio DeltaL' between the front and back of the inflection point P is LQ1/LQ2These quantities are characteristic of inflection points apart. As shown in fig. 10, a schematic diagram of the inflection point characteristics of the 10 th slice of ruffles palatine of sample No. 2 is shown, and as shown in table 1, a statistical table of the inflection point characteristics of the 10 th slice of ruffles palatine of sample No. 2 is shown.
TABLE 1.2 statistical table of inflection points of tangent line of No. 10 sample
Step four: crease-palate tangent matching based on inflection point feature recognition
Establishing a database, wherein the three-dimensional wrinkle palate database is derived from sample collection individuals consisting of 91 persons, each person collects five three-dimensional wrinkle palate data samples, two three-dimensional wrinkle palate data are randomly selected to be placed in a training database, and the remaining three-dimensional wrinkle palate data are placed in a testing database; in the testing process, the method of the invention is needed to match the test sample data with the sample data in the three-dimensional puckery palate sample database.
The samples were evenly sliced 40 times for the pucker palate test and the slice image is denoted as Ai(i-1, 2, …,40), the set of all slice images belonging to one test specimen is denoted as a-a ═ a1,A2,…,Ai,…,A40And a set of 182 three-dimensional crumple samples in the training database is denoted as B, each sample is numbered as BJ, J is 1,2, …,182, and B is { B1, B2, …, BJ, …, B182}, the three-dimensional crumple data samples BJ are uniformly sliced, and each slice image is numbered as BJi。
Available BJ ═ BJ1,…,BJi,…,BJ40Then the dataset consisting of all slices is B ═ B11,…,B140,B21,…,B240,…,BJ1,…,BJ40,…,B1821,…,B18240}；
The three-dimensional pucker palate matching problem described in step four can be equivalently described as: finding a three-dimensional crease palate in dataset B that matches test data a, specifically: and calculating the corner features of each three-dimensional pucker palate data BJ in the test data A and the sample data set B, matching the corner features of the test data A and the sample data BJ to obtain a matching error, and if the matching error is smaller than a matching threshold value, considering that the BJ is matched with the A, otherwise, considering that the BJ is not matched with the A.When the BJ is matched with the A, the BJ and the A need to be found firstiIdentically or closely positioned sections BJjTo A, aiAnd BJjAnd matching to obtain a matching error, and solving the average value of the obtained matching errors, wherein the value is the matching error of the BJ and the A.
(4)AiAnd BJjThe matching error calculation method of (2): first, according to slice AiThe position in the three-dimensional puckering data A, the position in B and A are foundiSets BJ of identically or closely located slicesj(ii) a Secondly, respectively adding AiSection and BJjSorting the corner features of the slices from large to small, selecting five inflection points with the largest corners as respective main matching inflection points, and expressing as { A ] according to the sequence of the inflection points in the palate fold tangent linei1,Ai2,Ai3,Ai4,Ai5} and { BJj1,BJj2,BJj3,BJj4,BJj5 }; thirdly, combining and pairing the main matching inflection points to form a plurality of matching queues, wherein each matching queue comprises 2-5 pairs of matching inflection point pairs, and each pair of matching inflection point pairs comprises AiAnd BJjOf main matching inflection points of (a), of the form { a }im,BJjn },1 is not less than m, n is not more than 5, and the two inflection points meet two conditions: A. the corner difference of the two inflection points is not more than the corner matching error of 22.5 degrees; B. the distance difference between the front points of the two inflection points is not more than 80% of the distance matching error. The matching corner pairs in the matching queue need to satisfy a certain sequence requirement, and 3 matching corner pairs { A } in a certain matching queue are assumedim1,BJjn1},{Aim2,BJjn2},{Aim3,BJjn3, and m1 is not less than m2 is not less than m3, then BJjThe 3 inflection points sequence of (a) should satisfy the condition: n1<n2<n 3; again, the match error is calculated for each match queue, assuming there are 3 pairs of matching inflection points in the match queue { A }im1,BJjn1},{Aim2,BJjn2},{Aim3,BJjn3, the calculation method is as follows: respectively calculating A in the matching queueiAnd BJjMean values of the horizontal and vertical coordinates of the inflection points (xA, yA) and (xB, yB); a is to beim1,Aim2,AiX and y are subtracted from the abscissa and ordinate of m3As new coordinates, BJ is also expressedjn1,BJjn2,BJjSubtracting xB and yB from the horizontal and vertical coordinates of n3 to obtain new coordinates; calculating the distance of new coordinates of the inflection point of each team and averaging to obtain the error of the matching queue: match queue error ═ aim1 and BJjn1 distance + Aim2 and BJjn2 distance + Aim3 and BJjn3 distance)/3; thirdly, A is mixediAnd BJjThe minimum value of all matching queue errors is used as the matching error of the slice; and finally, taking the average value of all slice matching errors of A as the matching error of A and BJ, if the matching error is smaller than a matching error threshold value 13, considering the two to be matched, and otherwise, judging the two not to be matched.
The embodiments of the present invention have been described in detail with reference to the accompanying drawings, but the present invention is not limited to the above embodiments, and various changes can be made within the knowledge of those skilled in the art without departing from the gist of the present invention. For example, the string-like member of the present invention may be in any shape that can fulfill its function.
Claims (5)
1. A method for identifying puckery palate by adopting uniform slicing and inflection point feature extraction is characterized by comprising the following specific steps:
step one, acquiring three-dimensional puckery palate data information;
step two, uniformly slicing and preprocessing the three-dimensional puckery palate data to obtain a coordinate sequence of the puckery palate slice;
step three, extracting inflection point characteristics of the tangent line of each section of the pucker palate, wherein the coordinate positions, the inflection points and the corners of all the pucker palate tangent lines jointly form a feature vector of the pucker palate;
in the third step, a point with severe change in the palate fold tangent line is regarded as an inflection point of the palate fold tangent line, skeleton information of the palate fold tangent line is converted into region chain code information, and the region chain code information acquisition method comprises the following steps: processing the coordinate sequence one-dimensional array which is obtained in the step two and represents the position of the tangent line of the puccinia palatogenes, wherein L is a determined value, the unit of L is the number of pixels, and a pixel point a on the tangent line of the puccinia palatogenes isMRepresents the Mth pixel from the left end, pixelPoint aM+LA pixel representing a distance L from the pixel point, pixel point aMTo pixel point aM+LThe vector direction angle of (a) is defined as a pixel point (a)MTraversing the meaningful identification area of the whole palate wrinkle tangent line to obtain all pixel point region chain code direction angles, wherein the vector direction angles form region chain code direction angle information of the palate wrinkle tangent line; preliminarily determining the inflection point of a tangent line of the pucker palate by the following method: traversing all pixel points of a meaningful identification area of the palate wrinkle tangent line from a left end point, defining vector direction angles from the pixel points to the pixel points as corners of the pixel points, if the corners of the pixel points are more than a given threshold value by 10 degrees, considering the pixel points as inflection points of the palate wrinkle tangent line, and sequentially processing all the pixel points to obtain an inflection point sequence of the palate wrinkle tangent line, wherein the inflection points are sparsely distributed in an approximately linear region of the palate wrinkle tangent line and densely distributed in a bending region of the palate wrinkle tangent line, and when a plurality of continuous points in a certain section of tangent line are all inflection points, selecting representative inflection points from the inflection points as the inflection points of the section of tangent line; screening the preliminarily determined inflection points, and removing redundant inflection points on the tangent line of the palate wrinkles, wherein the removing principle is that only one representative inflection point is considered in a certain pixel length, and the removing method comprises the following steps: sequencing all inflection points on the crease palate tangent line according to the sizes of corners, inserting the inflection points from large to small in sequence, if a certain inflection point is inserted, the inserted inflection point is not provided within the distance of 5 pixels on the left and right of the inflection point, the insertion of the inflection point is permitted, otherwise, the insertion is not performed, and the next inflection point is continuously processed, so that an inflection point sequence with the distance between the inflection points not less than 5 pixels is finally obtained, and thus, the effective inflection point of the crease palate tangent line is obtained; acquiring adjacent inflection point characteristics and separated inflection point characteristics of a crease palate tangent line, and determining the adjacent inflection point characteristics: starting from the 2 nd inflection point from left to right, if the inflection point P is the current inflection point, calculating the distance L from the inflection point P to the adjacent inflection point P1 before the inflection point PP1Distance L of P from its next adjacent inflection point P2P2Direction angle theta from inflection point P1 to inflection point PP1Direction angle θ from inflection point P to inflection point P2P2，ΔθP＝θP1-θP2And the ratio of the distances between the front and rear of the inflection point P is LP1/LP2The distance and the direction angle are adjacent inflection point characteristics; the determination method of the characteristic of inflection point separation comprises the following steps: starting from the 3 rd inflection point from left to right, if the inflection point P is the current inflection point, calculating the distance L from the inflection point P to the previous inflection point which is separated by an inflection point Q1Q1P to the next distance L from inflection point Q2Q2Direction angle theta from inflection point Q1 to inflection point QQ1Direction angle theta from inflection point Q to inflection point Q2Q2，ΔθQ＝θQ1-θQ2And the distance ratio DeltaL' between the front and back of the inflection point P is equal to LQ1/LQ2These quantities are all characteristic of inflection points apart;
and step four, matching the palate wrinkle tangent line based on inflection point feature identification, and outputting a matching result.
2. The method for identifying the ruffles in the mouth by using the uniform slicing and inflection point characteristic extraction as claimed in claim 1, wherein in the first step, the volunteer lies flat and disinfects in the mouth, the volunteer opens the mouth to the maximum opening degree, the ruffles in the mouth cavity are scanned in a non-contact mode, and after the scanning is finished, a complete ruffle palate scanning sample is led into a computer for storage, so that three-dimensional ruffles digital data are obtained.
3. The method of claim 2, wherein the non-contact scanning is based on the following principle: the measuring points are more at the place with large curvature, and the data points of the uniform or smooth area are less;
the specific scanning sequence is as follows: firstly, from the right occlusal surface to the left occlusal surface of the upper jaw of the volunteer; second, from the left buccal surface to the right buccal surface of the upper jaw of the volunteer; and thirdly, from the right lingual side to the left lingual side of the maxillary maxilla of the volunteer.
4. The method for identifying ruffles palate recognition using uniform slicing and inflection point feature extraction as claimed in claim 1, wherein in step two, the three-dimensional ruffles palate data is uniformly sliced, and the obtained slice is preprocessed to remove background color, extract the target ruffles tangent and calibrate meaningful identification regions of the ruffles tangent.
5. The method for identifying ruffles palate recognition by uniform slicing and inflection point feature extraction as claimed in claim 1, wherein in step four, the features of the test sample are compared with the features in the sample feature library one by one to obtain a matching error, and if the matching error is smaller than the matching error, the matching is considered to be successful.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911130386.7A CN111222407B (en) | 2019-11-18 | 2019-11-18 | Palate wrinkle identification method adopting uniform slicing and inflection point characteristic extraction |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911130386.7A CN111222407B (en) | 2019-11-18 | 2019-11-18 | Palate wrinkle identification method adopting uniform slicing and inflection point characteristic extraction |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111222407A CN111222407A (en) | 2020-06-02 |
| CN111222407B true CN111222407B (en) | 2022-07-01 |
Family
ID=70825825
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911130386.7A Active CN111222407B (en) | 2019-11-18 | 2019-11-18 | Palate wrinkle identification method adopting uniform slicing and inflection point characteristic extraction |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111222407B (en) |
Family Cites Families (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5680481A (en) * | 1992-05-26 | 1997-10-21 | Ricoh Corporation | Facial feature extraction method and apparatus for a neural network acoustic and visual speech recognition system |
| AUPQ896000A0 (en) * | 2000-07-24 | 2000-08-17 | Seeing Machines Pty Ltd | Facial image processing system |
| RU2358319C2 (en) * | 2003-08-29 | 2009-06-10 | Самсунг Электроникс Ко., Лтд. | Method and device for photorealistic three dimensional simulation of face based on image |
| CN101359366B (en) * | 2008-07-28 | 2011-11-16 | 同济大学 | Pattern matching recognition system and implementing method thereof |
| CN103902958A (en) * | 2012-12-28 | 2014-07-02 | 重庆凯泽科技有限公司 | Method for face recognition |
| US9959455B2 (en) * | 2016-06-30 | 2018-05-01 | The United States Of America As Represented By The Secretary Of The Army | System and method for face recognition using three dimensions |
| CN107346425B (en) * | 2017-07-04 | 2020-09-29 | 四川大学 | Three-dimensional texture photographing system, calibration method and imaging method |
| CN208659317U (en) * | 2017-10-26 | 2019-03-29 | 山西医科大学口腔医院 | Palate wrinkle image collecting device |
| CN109784335B (en) * | 2019-01-25 | 2023-03-24 | 太原科技大学 | Method for calibrating boundary of region of interest of ruffle image based on least square fitting |
-
2019
- 2019-11-18 CN CN201911130386.7A patent/CN111222407B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN111222407A (en) | 2020-06-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| De Tobel et al. | An automated technique to stage lower third molar development on panoramic radiographs for age estimation: a pilot study. | |
| Tilotta et al. | Construction and analysis of a head CT-scan database for craniofacial reconstruction | |
| Syed et al. | Conversion of palatal rugae pattern to scanable Quick Response code in an Arabian population | |
| CN113052902B (en) | Tooth treatment monitoring method | |
| CN112017743B (en) | Automatic generation platform and application of disease risk evaluation report | |
| Coward | The stability of lip p ability of lip p ability of lip pattern chara ttern chara ttern characteristics cteristics over time | |
| CN111222407B (en) | Palate wrinkle identification method adopting uniform slicing and inflection point characteristic extraction | |
| Li et al. | Computer-aided disease diagnosis system in TCM based on facial image analysis | |
| CN115690045A (en) | Quantitative assessment method for bone increment before and after periodontitis treatment based on curved surface fault slice | |
| Lestrel et al. | Mandibular shape analysis in fossil hominins: Fourier descriptors in norma lateralis | |
| WO2023089347A1 (en) | Method and system for biometric identification | |
| Jasso-Cuéllar et al. | Anterior dental arch shape and human identification: Kieser et al. method applied to 2D-3D dental models in Mexican population | |
| Seiffert et al. | Morphometric variation in the hominoid orbital aperture: a case study with implications for the use of variable characters in Miocene catarrhine systematics | |
| Toma | Characterization of normal facial features and their association with genes | |
| O'Higgins | A morphometric study of cranial shape in the Hominoidea | |
| Li | Postnatal development of pelvic sexual dimorphism in four anthropoid primates | |
| Braun | Variations in the Shape of the Chin in South African Using Cone Bean Computed Tomography Scans | |
| Archer et al. | Medico-Legal Investigation of Atrocities Committed during the Solomon Islands" Ethnic Tensions" | |
| Gupta et al. | Role of orthodontics in forensic facial reconstruction for human identification | |
| Charles | A geometric morphometric analysis of the human ossa coxae for sex determination | |
| CN114343703A (en) | Gender inference method, system, device and storage medium based on teeth | |
| Majd et al. | A novel hybrid approach for cephalometric landmark detection | |
| Umbelino et al. | THE JUVENILE NEANDERTHAL MANDIBLE FROM COUPE-GORGE CAVE, FRANCE | |
| Ridel | An automated computer-assisted approximation of the nose in South Africans from CBCT (Cone Beam Computed Tomography) scans. | |
| Reda et al. | Description and taxonomic assessment of fossil Cercopithecidae from the Pliocene Galili Formation (Ethiopia) |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |