CN102038562A - Method for manufacturing artificial implant - Google Patents

Abstract

The invention discloses a method for manufacturing an artificial implant, which comprises the following steps of: acquiring a preoperative three-dimensional volume image and a postoperative three-dimensional volume image of an operation site, and separating the preoperative object image and the postoperative object image which have objects respectively accordingly; performing registration on the preoperative object image and the postoperative object image to form a registered image; obtaining a differential image of the defective position of each object according to the registered image, and establishing three-dimensional mold point coordinates of the defective position according to the differential image; and manufacturing the artificial implant of which the appearance is the same as that of the defective position according to the three-dimensional mold point coordinates.

Description

Make the method for artificial implantation

Technical field

The present invention relates to a kind of method of making artificial implantation, particularly set up three-dimensional mould point coordinates, and three-dimensional in view of the above mould point coordinates is made the method for artificial implantation according to perioperative image.

Background technology

At present, in clinical neurosurgery, patient is causing under the too high situation of intracranial pressure because of brain tumor, hydrocephalus or other cause of disease, often reduce intracranial pressure by craniotomy, and the skull defective that craniotomy caused, then repair via the skull reconstruction operations, under the prerequisite of the head shape before reducing postoperative infection as far as possible and recovering patient's art, reconstructing damaged skull part should be as far as possible and the operative incision driving fit, and the external form of this skull (radian is with big or small) should conform to the former capitiform of patient.

Common skull restorative procedure is to accept to utilize the principle of mirror image under the situation of one-sided craniotomy patient, by the capitiform of this patient's offside intact skulls, reconstruct the damaged skull that cuts side, but this mirror image skull reconstruction method meets with three main problems:

(1) individual human skull is not complete left-right symmetric, and the mode that mirror image is rebuild can't restore real capitiform before the art, causes the change of patient's outward appearance and uncomfortable;

(2) utilize the mode of mirror image can't reach effective driving fit of rebuilding skull and operative incision, even still can't accomplish complete driving fit via technician's manual adjustment;

(3) this method can't provide effective reconstruction because the restriction that mirror image is drawn materials is only applicable to one-sided craniotomy for the patient who accepts the bilateral craniotomy.

Another kind of situation is, when patient need carry out the surgical excision operation because of injured or pathological changes, a certain position (as nose cartilaginous tissue or breast) excision with health, and the formation defect, make patient's outward appearance change to some extent and imperfect, at this moment, need rebuild artificial implantation in defect, in the hope of reduction patient's outward appearance by plastic surgery operations.But the contour shape of artificial implantation is not made according to the original complexion of defect usually, and therefore, patient's outward appearance still changes to some extent, and artificial implantation still has problems such as effectively driving fit, outward appearance change.

Summary of the invention

In view of above-mentioned problems of the prior art, the invention provides a kind of method of making artificial implantation, change to solve the outward appearance that causes after the mirror image artificial bone-head repairs, driving fit and can't provide the patient of bilateral craniotomy to rebuild the first-class problem of artificial bone fully.

The technical solution used in the present invention is, a kind of method of making artificial implantation is proposed, this method obtains the preceding three-D volumes image of art of operative site, and isolate in view of the above have object art before the object image, obtain the postoperative three-D volumes image of operative site, and isolate postoperative object image in view of the above with object, it is that object has at least one defect that postoperative object image is compared the preceding object image of Rhizoma Atractylodis Macrocephalae, object image before the art and postoperative object image are carried out para-position, produce the para-position image, obtain the difference image of the defect of corresponding object according to the para-position image, set up the three-dimensional mould point coordinates of defect according to the difference image, make artificial implantation with the identical moulding of defect according to three-dimensional mould point coordinates again.

The method of the making artificial implantation that the present invention proposes has following advantage:

1, after artificial implantation is implanted in patient's defect with reconstruction operations, the face shaping similarity height before face shaping behind patient's the defect reconstruction operations and the art;

2, the contour shape of artificial implantation is close with defect, makes the adaptation height of artificial implantation and defect;

3, artificial implantation can be used in the patient of bilateral craniotomy or many defect.

Description of drawings

The present invention will be described in detail below in conjunction with drawings and Examples, wherein:

Fig. 1 is the method flow sketch map of the artificial implantation made of the present invention;

Fig. 2 is the sagittate section sketch map of the preceding three-D volumes image of art in one embodiment of the invention;

Fig. 3 is the sagittate section sketch map of postoperative three-D volumes image in the embodiment of the invention;

Fig. 4 is the sagittate section sketch map of the preceding object image of art in the embodiment of the invention;

Fig. 5 is the sagittate section sketch map of postoperative object image in the embodiment of the invention;

Fig. 6 is the sagittate section sketch map of para-position image in the embodiment of the invention;

Fig. 7 is the horizontal section sketch map of para-position image in the embodiment of the invention;

Fig. 8 is a difference image sketch map in the embodiment of the invention;

Fig. 9 is a three-dimensional mould point coordinates sketch map in the embodiment of the invention;

Figure 10 is the three-dimensional reconstruction sketch map after repairing in the embodiment of the invention;

Figure 11 is the three-dimensional reconstruction sketch map of the low resolution of the preceding three-D volumes image of art in the embodiment of the invention;

Figure 12 is that the high-resolution three-dimension of postoperative three-D volumes image in the embodiment of the invention is rebuild sketch map;

Figure 13 is a flow chart of obtaining the preceding object image of art in the embodiment of the invention before the art in the three-D volumes image;

Figure 14 is the sketch map of the preceding binaryzation image of art in the embodiment of the invention;

Figure 15 is the sketch map of closed outline in the embodiment of the invention;

Figure 16 is the flow chart that produces postoperative object image in the embodiment of the invention;

Figure 17 is the flow chart that produces the para-position image in the embodiment of the invention;

Figure 18 is the sketch map that postoperative object image repeatedly is placed on object image before the art in the embodiment of the invention;

Figure 19 is the flow chart that produces the difference image in the embodiment of the invention from the para-position image;

Figure 20 is the sketch map of preliminary image in the embodiment of the invention;

Figure 21 is the secondary image sketch map of further embodiment of this invention;

Figure 22 is the present invention's three image sketch maps of another embodiment again;

Figure 23 is the present invention's difference image sketch map of another embodiment again;

Figure 24 be the present invention again another embodiment produce the flow chart of three-dimensional mould point coordinates;

Figure 25 is that bilateral skull reconstruction operations of the present invention is repaired three-dimensional reconstruction sketch map later.

The specific embodiment

Fig. 1 is a kind of schematic flow sheet of making the method for artificial implantation that the present invention proposes.This method is made artificial implantation according to the following step:

(S10) obtain the preceding three-D volumes image 1 (as shown in Figure 2) of art, three-D volumes image 1 shows the original complexion of object before damaged of operative site before the art;

(S20) obtain postoperative three-D volumes image 2 (as shown in Figure 3), the postoperative complexion of the object that postoperative three-D volumes image 2 shows operative sites after damaged, this postoperative complexion is compared with original complexion, has at least one defect 20 on the object;

(S30) obtain object image 3 (as shown in Figure 4) before the art the three-D volumes image 1 before art, object image 3 is object three-D volumes image of (complete) before damaged before this art;

(S40) obtain postoperative object image 4 (as shown in Figure 5) from postoperative three-D volumes image 2, the three-D volumes image that this postoperative object image 4 has defect 20 for object after damaged;

(S50) object image before the art 3 and postoperative object image 4 are carried out para-position, and produce 5 (as shown in Figure 6 and Figure 7) of para-position image;

(S60) the difference image 6 (as shown in Figure 8) of object image 3 and the defect of postoperative object image 4 corresponding objects before the removal surgery from para-position image 5;

(S70) set up the three-dimensional mould point coordinates 7 (as shown in Figure 9) of defect according to difference image 6; And

(S80) make artificial implantation with defect 20 identical moulding according to three-dimensional mould point coordinates 7.

Figure 10 is the three-dimensional reconstruction sketch map of one embodiment of the invention after repairing.Among the figure, the defect at postoperative three-D volumes image 4 positions has been inserted the artificial implantation of rebuilding according to three-dimensional mould point coordinates 7.Because artificial implantation is made according to the three-dimensional mould point coordinates 7 of defect 20, therefore, the similarity of itself and defect 20 is very high, after the implantation, defect 20 is splendid with the adaptation of artificial implantation 8, and the outward appearance of operative site and preoperative outward appearance are almost without any difference.

In one embodiment of the invention, three-D volumes image 1 is the computerized tomography image of patient's low resolution of (during diagnosis) before operation before the art, its matrix size is that 512*512 (pixel (Pixel)) * 32 (tangent plane number), pixel size are 0.47mm*0.47mm*5mm (tangent plane thickness), sees also Figure 11.Postoperative three-D volumes image 2 be patient at postoperative high resolution computer fault image, its matrix size is that 512*512 (pixel) * 184 (tangent plane number), pixel size are 0.49mm*0.49mm*0.63mm, sees also Figure 12.Because the tangent plane number of three-D volumes image 1 is 32 before the art, and the thickness of every tangent plane thicker (tangent plane thickness is 5mm), therefore, three-D volumes image 1 (as shown in Figure 2) is easy to generate partial volume effect (partial volume effect) before the art of low resolution, i.e. stepped part among the figure.

In other words, if the resolution of three-D volumes image 1 is lower than the resolution of postoperative three-D volumes image 2 before the art, and the tangent plane thickness of three-D volumes image 1 is bigger than the tangent plane thickness of postoperative three-D volumes image 2 before the art, be that the picture specification of three-D volumes image 1 and postoperative three-D volumes image 2 is inconsistent before the art, can't directly obtain object image 3 before the art with postoperative object image 4 same positions from three-D volumes image 1 before the art.Therefore, in this embodiment, obtain the three-D volumes image 1 before art that the step of object image 3 also comprises (seeing also Figure 13) before the art:

(S301) adjust the brightness and the contrast of three-D volumes image 1 before the art, make it be fit to differentiate the object in the three-D volumes image 1 before art (be example with the skull among this embodiment, but be not limited to this);

(S302) three-D volumes image 1 before the art is resampled, use the specification of adjusting the preceding three-D volumes image 1 of art, make that the specification of three-D volumes image 1 is identical with the specification of postoperative three-D volumes image 2, thereby resample image before the formation art;

(S303) image of resampling before the art is used threshold method (thresholding), resolution target thing and background, and the video conversion Cheng Shuqian binaryzation image 11 (as shown in figure 14) of resampling before with art with the binaryzation method, before the art in the binaryzation image 11 numerical value equal 1 and be object (as: skull), equal 0 and be considered as background; And

(S304) use active profile method to remove influenced by partial volume effect in the binaryzation image 11 before the art the false shadow that is produced, use and obtain object image 3 (as shown in Figure 4) before the art.

Because the profile before the art in the binaryzation image 11 is not to be closed, therefore, in this embodiment, before the active profile method of use is removed the false shadow of binaryzation image 11 before the art, elder generation is with the profile of the object (as: first half of binaryzation image 11 before the sagittate section of head and the art of crown tangent plane) of the some in the binaryzation image 11 before the art respectively, constitute closed outline (as shown in figure 15) with image method, again active profile method is carried out in the inboard and the outside of closed outline, in order to object image 3 before the generation art.

In this embodiment, the object of binaryzation image 11 (as: skull) is inboard before to art when carrying out active profile method, earlier binaryzation image before the art 11 is carried out the image expansion (Image Dilation) of two pixels; When active profile method is carried out in the object of binaryzation image 11 (as: skull) outside before to art, earlier binaryzation image before the art 11 is carried out the image erosion (Image Erosion) of two pixels, to reach the curved profile that coats binaryzation image 11 before the art but not the evagination profile, produce object image 3 (as shown in Figure 4) before the art.

In sum, three-D volumes image 1 can be obtained object image 3 before the art with postoperative object image 4 same sizes according to the processing of above steps before the art.

In the above embodiment of the present invention, the parameter setting of active profile method has iterations, step pitch, elasticity (elasticity) and flintiness (rigidity) etc., wherein iterations be that 200 times, step pitch are 0.1, elasticity is the spatial spreading degree square and flintiness be the biquadratic of spatial spreading degree.

In another embodiment of the present invention, see also Figure 16, the flow process that produces postoperative object image 4 comprises the following steps:

(S401) brightness and the contrast of adjustment postoperative three-D volumes image 2 make its suitable object of differentiating in postoperative three-D volumes image 2 (as: skull); And

(S402) postoperative three-D volumes image 2 is used threshold method (thresholding), with resolution target thing and background, and convert postoperative three-D volumes image 2 to postoperative binaryzation image with the binaryzation method, numerical value equals 1 and is object in the postoperative binaryzation image, equal 0 and be considered as background, and with postoperative binaryzation image as postoperative object image 4.

In the various embodiments described above of the present invention, three-D volumes image 1 is the computerized tomography image before the art, and the brightness value of computerized tomography image is with computerized tomography number [Computed Tomography Value, be called for short: the CT value, its unit is Hounsfield unit (HU)] expression, and the CT value of various tissue substances is in certain scope, use and distinguish different tissue substances, for defect is bone, its CT value is between 1400HU-3000HU, therefore, in aforesaid each embodiment of the present invention, threshold method differentiation skull part can be 1400HU-3000HU with the CT value of background parts.

In another embodiment of the present invention, see also Figure 17, the flow process that produces the para-position image comprises the following steps:

(S501) postoperative object image 4 is according to the initial conversion parameter of video conversion model, repeatedly is placed on before the art on the object image 3 (as shown in figure 18);

(S502) whether object image 3 aligns with postoperative object image 4 before use cost function (Cost Function) the assessment art, concrete steps are: the gap value of the object of object image 3 before object after the first logistic in the object image 4 and the art, to add up after the gap value square again, calculate residual sum of squares (RSS) (Sum of Squared Difference, be called for short: SSD), if this residual sum of squares (RSS) is less than a default value, then carry out step (S505), otherwise carry out step (S503);

(S503) utilize the optimization alternative manner to adjust conversion parameter, produce in view of the above and upgrade conversion parameter;

Whether (S504) postoperative object image 4 is with the renewal conversion parameter of video conversion model, repeatedly is placed on before the art on the object image 3, align according to step (S502) assessment again; And,

(S505) make postoperative object image 4 through final conversion parameter conversion, produce para-position image 5 (shown in Fig. 6 and 7).

In the above-described embodiments, the video conversion model is affine conversion (affine transform) model, and the conversion parameter of this affine transformation model comprises translation, rotation, convergent-divergent and crooked etc.The optimization alternative manner then uses Powell method (Powell method), selects suitable linear convergence direction in regular turn in each iteration, with the residual sum of squares (RSS) that function was calculated that progressively reduces cost, makes residual sum of squares (RSS) less than default value.

In the above-described embodiments, before use cost function calculation residual sum of squares (RSS), comparison is with the postoperative object image after the video conversion model conversion earlier, see outside the scope that whether drops on the preceding object image 3 of art, if, then when assessing the cost function, add a point penalty value, otherwise, direct use cost function calculation residual sum of squares (RSS).Wherein, suppose that the postoperative object image 4 and the visual field of the preceding object image 3 of art are fairly similars, therefore, the object in the postoperative object image 4 all should be able to find corresponding object on the object image 3 before art, at this moment, the point penalty value can a given bigger numerical value.Otherwise, if tolerate when some situations that can't find corresponding skull part take place the then given less numerical value of point penalty value.

In another embodiment of the present invention, see also Figure 19, the flow process that produces the difference image from the para-position image comprises the following steps:

(S601) calculate in the para-position image 5, the difference of complete object and damaged object produces preliminary image 51 (as shown in figure 20);

(S602) first half to preliminary image 51 uses first structure matrix to carry out image erosion (erosion) and expansion (dilation), and use second structure matrix to carry out image to the Lower Half of preliminary image 51 to corrode and expansion, use and remove trickle difference part in the preliminary image, produce 52 (as shown in figure 21) of secondary image, wherein first structure matrix is more than the row matrix columns of second structure matrix;

(S603) use three-dimensional zone (region growing) method of growing up to remove the part that is not connected of non-genus defect in the secondary image 52, to produce three images 53 (as shown in figure 22); And

(S604) manually to wipe the part that is not connected of non-genus defect in three images 53 of method removal, promptly produce difference image 6 (as shown in figure 23).

In the above-described embodiments, first structure matrix is 5 * 5 matrixes, shown in its matrix thes contents are as follows:

$\left[\begin{array}{ccccc}0& 0& 1& 0& 0\\ 0& 1& 1& 1& 0\\ 1& 1& 1& 1& 1\\ 0& 1& 1& 1& 0\\ 0& 0& 1& 0& 0\end{array}\right]$

Second structure matrix is 3 * 3 matrixes, shown in its matrix thes contents are as follows:

$\left[\begin{array}{ccc}0& 1& 0\\ 1& 1& 1\\ 0& 1& 0\end{array}\right]$

In yet another embodiment of the present invention, see also Figure 24, the flow process that produces three-dimensional mould point coordinates comprises the following steps:

(S701) use the square method (Marching cube) of advancing difference image 6 to be set up the initial three-dimensional mould point coordinates of the defect of corresponding object;

(S702) use surperficial Laplce's method (surface Laplacian) to repair (smoothing processing) initial three-dimensional mould point coordinates,, produce comparatively level and smooth three-dimensional mould point coordinates 7 (Fig. 9) so that the surface of initial three-dimensional mould point coordinates is more level and smooth.

In the various embodiments described above of the present invention, image before each art, the image that reaches the postoperative image and be correlated with is all 3-dimensional image, each 3-dimensional image can be along crown, sagittal and horizontal direction cutting, to form coronal section, sagittate section and horizontal section, for ease of each handling process of the present invention is described, in the present invention mostly with the image of one of them tangent plane as representative, and, the all images of each embodiment are example with one-sided craniotomy operation all in the present invention, but do not limit to the present invention and only can be applied in one-sided craniotomy operation, the present invention also can use at bilateral skull reconstruction operations, as shown in figure 25, among the figure, the left side is the three-dimensional mould point coordinates 8 of the bilateral defect of skull, and the right side is to insert the simulating image of rebuilding this artificial implantation according to three-dimensional mould point coordinates 8.Certainly also can make other artificial implantation, such as: the artificial cartilage of nose, the breast prosthesis of chest etc.

The above is only for illustrating, but not is restriction.Any spirit of the present invention and category of not breaking away from carries out to such an extent that equivalent modifications or change all should be contained in the scope of patent protection of the attached application in back.

Claims (22)

1. a method of making artificial implantation is characterized in that, described method comprises the row step:

Obtain the preceding three-D volumes image of art, the three-D volumes image has the original complexion of object before damaged of operative site before this art;

Obtain the postoperative three-D volumes image of this operative site, the postoperative complexion of this object that this postoperative three-D volumes image has this operative site after damaged, this postoperative complexion is to have at least one defect on this object compared to original complexion;

The three-D volumes image is obtained object image before the art before the art, and the object image is the three-D volumes image of this object before damaged before this art;

Obtain postoperative object image from postoperative three-D volumes image, this postoperative object image three-D volumes image that to be this object have this defect after damaged;

Object image before the art and postoperative object image are carried out para-position, and produce the para-position image according to this;

From this para-position image before the removal surgery in object image and the postoperative object image to the difference image of defect that should object;

Set up the three-dimensional mould point coordinates of this defect according to this difference image;

And, make artificial implantation with the identical moulding of this defect according to this three-dimensional mould point coordinates.

2. the method for claim 1 is characterized in that, obtains the three-D volumes image before art that the step of object image also comprises before the art:

Adjust the brightness and the contrast of the preceding three-D volumes image of art, make its suitable object of differentiating in the preceding three-D volumes image of this art;

Three-D volumes image before the art is resampled, use the specification of adjusting the preceding three-D volumes image of this art, make that the specification of the preceding three-D volumes image of this art is identical with the specification of this postoperative three-D volumes image, thereby resample image before the formation art;

Image uses threshold method to resampling before this art, differentiate this object and background, and with resample before with this art binaryzation image before the art that converts to of image of binaryzation method, before this art in the binaryzation image numerical value equal 1 and be object, equal 0 and be considered as background;

And, use active profile method to remove influenced by partial volume effect in the binaryzation image before this art the false shadow that is produced, use and obtain object image before this art.

3. method as claimed in claim 2, it is characterized in that, before the active profile method of use is removed the false shadow of binaryzation image before the art, earlier the profile of a part of object in the binaryzation image before this art is constituted closed outline with image method respectively, again active profile method is carried out in the inboard and the outside of the closed outline of binaryzation image before this art, in order to produce object image before this art.

4. method as claimed in claim 3 is characterized in that, when active profile method is carried out in the inboard of the closed outline of binaryzation image before to this art, binaryzation image before the art is carried out the image expansion of two pixels.

5. method as claimed in claim 3 is characterized in that, carry out should be active during the profile method in the outside of the closed outline of binaryzation image before to art, binaryzation image before the art carried out the image erosion of two pixels.

6. method as claimed in claim 2 is characterized in that the parameter setting of active profile method comprises iterations, step pitch, elasticity and flintiness.

7. method as claimed in claim 6 is characterized in that, described iterations is 200 times.

8. method as claimed in claim 6 is characterized in that, described step pitch is 0.1.

9. method as claimed in claim 6 is characterized in that, described elasticity be the spatial spreading degree square.

10. method as claimed in claim 6 is characterized in that, described flintiness is the biquadratic of spatial spreading degree.

11. the method for claim 1 is characterized in that, the step that produces postoperative object image from postoperative three-D volumes image also comprises:

Adjust the brightness and the contrast of postoperative three-D volumes image, make its suitable object of differentiating in this postoperative three-D volumes image;

And, postoperative three-D volumes image is used threshold method, to differentiate this object and background, and this postoperative three-D volumes video conversion is become postoperative binaryzation image with the binaryzation method, numerical value equals 1 and is this object in this postoperative binaryzation image, equal 0 and be considered as this background, and with this postoperative binaryzation image as this postoperative object image.

12. the method for claim 1 is characterized in that, the step that the object image produces the para-position image before postoperative object image and the art also comprises:

Postoperative object image is according to the initial conversion parameter of video conversion model, repeatedly is placed on before the art on the object image;

And, whether the object image aligns with this postoperative object image before this art of use cost function evaluation, concrete steps are: the gap value of the object of object image before object after the first logistic in the object image and the art, to add up after the gap value square again, calculate residual sum of squares (RSS), if this residual sum of squares (RSS) less than a default value, then produces this para-position image according to this.

13. method as claimed in claim 12 is characterized in that,

When this residual sum of squares (RSS) greater than a default value, then utilize the optimization alternative manner to adjust this conversion parameter, produce to upgrade conversion parameter according to this;

And, with the renewal conversion parameter of this postoperative object image, repeatedly be placed on before this art on the object image with this video conversion model, produce this para-position image according to this.

14. method as claimed in claim 12 is characterized in that, this video conversion model is affine transformation model, and that the conversion parameter of this affine transformation model comprises is mobile, rotation, convergent-divergent and crooked.

15. method as claimed in claim 13, it is characterized in that this optimization alternative manner is a Powell method, is to select the capable convergence of suitable line direction in regular turn in each iteration, this residual sum of squares (RSS) that is calculated progressively to reduce this cost function makes this residual sum of squares (RSS) less than this default value.

16. method as claimed in claim 13, it is characterized in that, when this residual sum of squares (RSS) of use cost function calculation, comparison is with the postoperative object image after this video conversion model conversion earlier, as drop on outside the scope of object image before this art, then when calculating this cost function, add the point penalty value.

17. the method for claim 1 is characterized in that, the step that this para-position image produces the difference image also comprises:

Calculate the difference of object complete in the para-position image and damaged object, to produce preliminary image;

This preliminary image is carried out image corrode and expansion, use and remove trickle difference part in the preliminary image, produce the secondary image;

Use three-dimensional regional flop-in method to remove the part that is not connected of non-genus defect in this secondary image, to produce image respectively three times;

And, remove the part that is not connected of non-this defect of genus in these three images with the artificial method of wiping, promptly produce this difference image.

18. method as claimed in claim 17, it is characterized in that, when each preliminary image being carried out the image erosion with expansion, the first half to this preliminary image uses first structure matrix, carrying out image corrodes and expansion, the Lower Half of this preliminary image is used second structure matrix, carry out image and corrode and expansion.

19. method as claimed in claim 18 is characterized in that, described first structure matrix is 5 * 5 matrixes, shown in its matrix thes contents are as follows:

$\left[\begin{array}{ccccc}0& 0& 1& 0& 0\\ 0& 1& 1& 1& 0\\ 1& 1& 1& 1& 1\\ 0& 1& 1& 1& 0\\ 0& 0& 1& 0& 0\end{array}\right].$

20. method as claimed in claim 18 is characterized in that, described second structure matrix is 3 * 3 matrixes, shown in its matrix thes contents are as follows:

$\left[\begin{array}{ccc}0& 1& 0\\ 1& 1& 1\\ 0& 1& 0\end{array}\right].$

21. the method for claim 1 is characterized in that, described difference image is with the square method of advancing, and sets up initial three-dimensional mould point coordinates that should defect.

22. method as claimed in claim 21 is characterized in that, described initial three-dimensional mould point coordinates carries out the surface smoothing processing with surperficial Laplce's method, produces this level and smooth three-dimensional mould point coordinates.

Images