CN105095581B - A kind of casting shrinking hole defect image generation method - Google Patents
A kind of casting shrinking hole defect image generation method Download PDFInfo
- Publication number
- CN105095581B CN105095581B CN201510437966.6A CN201510437966A CN105095581B CN 105095581 B CN105095581 B CN 105095581B CN 201510437966 A CN201510437966 A CN 201510437966A CN 105095581 B CN105095581 B CN 105095581B
- Authority
- CN
- China
- Prior art keywords
- defect image
- value
- defect
- image
- dimensional array
- 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
Landscapes
- Image Analysis (AREA)
- Image Processing (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
A kind of casting shrinking hole defect image generation method based on diamond square algorithms disclosed by the invention, including defect image generation, defect image conversion process, fusion product structural context three phases, the defect image generation phase includes the structure of two-dimensional array, the generation and storage of altitude data, the normalized of altitude data；The defect image conversion process stage includes the histogram equalization processing of defect image, thresholding processing, gaussian filtering, gray scale stretching；The fusion product structural context stage includes the normalized of defect image, the dot product of background image and defect image.There is the method for the present invention good randomness and simulation quality, the defect of generation can be used for teaching, increase sample and test defect detection algorithm during generating casting shrinking hole defect image.
Description
Technical field
The present invention relates to Computer Image Processing field, more particularly to a kind of casting shrinking hole defect image generation method.
Background technology
Teaching either in casting flaw training process, or the automatic detection defective system of design and test defect
Detection algorithm is required for a large amount of defect sample.It is ideal that defect sample is acquired from actual production, but foundry engieering is horizontal
Raising substantially reduced defect rate or even the defect of some types is more rare so that it is very tired largely to obtain similar sample
It is difficult.
The method of computer generated image mainly has CAD methods based on parameter and is based on non-parametric image superposition at present
Method.
Defect image method of formation and the defect addition method based on CAD model have become two kinds of main sides for generating casting flaw
Method.Ray trace and X-ray attenuation are the bases that CAD model generates 3-D casting flaws.Its process is imagined from x-ray source
Emit one group of ray towards detector, there is intersection point on the surface of X-ray detection workpiece and the surface of defect of ray and imagination, lead to
It crosses ray attenuation formula and calculates the ray energy intensity for reaching each pixel of detector, generate defect image, friendship is drafted in transformation
The position of point is to get to different images.This method deficiency is：1) it due to needing the workpiece that hand-designed is imagined first, determines
The spatial form of defect, thus manual drawing process significantly limits the diversity and randomness that defect is presented naturally；2) three
The expression of dimension data and ray attenuation calculate, and operand is very big, are unfavorable for a large amount of defect and generate.
The defect addition method is another casting flaw technology.The difference of it and CAD methods is that defect uses during creating
Casting defect is added on true ray image by 2-D image techniques.It had not both needed complicated three-dimensional software packet, did not needed yet
Survey cast model.However, the result of study of early stage shows the method used difference, the defect type of generation is totally different, some sides
Method can not make the defect of generation close to the feature of real defect, and practical purpose is not achieved.
Invention content
The shortcomings that it is an object of the invention to overcome the prior art and deficiency provide a kind of casting shrinking hole defect image generation
Method, this method is based on diamond-square algorithms, for generating shrinkage cavity rarefaction defect image common in casting.Wherein,
Diamond-square algorithms are also square-square close classification, be by Method of Random Mid-point Displacement for square-shaped planar and
The terrain surface specifications of generation.
The purpose of the present invention is realized by the following technical solution：
A kind of casting shrinking hole defect image generation method, including defect image generation, defect image conversion process, fusion production
Product structural context three phases, specially：
S1. defect image generation phase：
S101. one (2 is builtn+1)×(2n+ 1) two-dimensional array, the two-dimensional array are empty two-dimensional array, the value model of n
It encloses for (1,10)；
S102. diamond-square algorithms are utilized to generate (2n+1)×(2n+ 1) altitude data of size, and corresponded to
Ground is stored in the two-dimensional array that step S101 is established；
S103. the altitude data in two-dimensional array is normalized；
S104. the altitude data after normalized is multiplied with the maximum gradation value of 8 gray-value images i.e. 255, and
As the defect image generated；
S2. the defect image conversion process stage：
S201. histogram equalization is carried out to the defect image obtained through step S104 processing；
S202. corresponding thresholding is carried out to the defect image obtained through step S201 processing；
S203. gaussian filtering is carried out to the image obtained through step S202 processing；
S204. gray scale stretching is carried out to the image obtained through step S203 processing；
S205. step S201~S204 is repeated, the defect image after conversion process is obtained；
S3. the fusion product structural context stage：
S301. the image with defect in the true casting in interception part；
S302. after textures synthesis is handled, background image I is used as after the image of step S301 is removed defect；
S303. its maximum gradation value I is detected to background image I calculatingmax, minimum gradation value Imin, calculate H=255/
ImaxValue；
S304. it using the H values in step S303 as the normalization upper limit of defect image, and is lacked to what is obtained by step S205
Sunken image is normalized to obtain defect image D；
S305. background image I and defect image D are subjected to dot product, obtain final defect image M=I*D.
The step S102 includes the following steps：
S102-1. (the 2 of step S101 generationsn+1)×(2n+ 1) two-dimensional array is (2n+1)×(2n+ 1) a pixel is being just
It is rectangular, the i.e. square iterative fractal schematic diagram using the square of 5*5 as diamond-square algorithms, with range (- 1,1)
Random value initializes the height value of square four angle points A, B, C, D, is respectively set as hA、hB、hC、hD, and according to hA、hB、hC、
hDRandom value of the average value of summation plus a range in (- 1,1)Determine the height value of square center point M, then in
The height value of heart point M is
S102-2. square is divided into 4 son squares, and according to hA、hB、hC、hDWith hMThe average value of summation adds
One range determines the height value of this 4 son square boundaries midpoint E, F, G, H in the random value of (- 1,1), then its height value point
It is not
S102-3. step S102-2 is repeated, recurrence divides each son square, the height until determining all elevational points of square
Journey value, and it is stored correspondingly in the two-dimensional array that step S101 is established.
The step S103 includes the following steps：
S103-1. the minimum value of altitude data in two-dimensional array is found out, and takes its absolute value, is set as hmin；
S103-2., the altitude data of all the points in two-dimensional array is added to the absolute value that step S103-1 is obtained respectively, i.e.,
h'A=hA+hmin、h'B=hB+hmin..., and as new altitude data, that is, h'A、h'B... it is stored in two-dimensional array；
S103-3. the maximum value h of altitude data in the two-dimensional array that step S103-2 is obtained is found outmax；
S103-4. the maximum value obtained altitude data all in two-dimensional array divided by step S103-3 is completed to height
The normalized of number of passes evidence.
The step S202 includes the following steps：
S202-1. according to Ostu maximum variance between clusters, the segmentation of the defect image obtained through step S201 processing is found out
Threshold value；
S202-2. the point that all gray values in defect image are less than segmentation threshold is selected, and the gray value of these points is set
It is set to the average value of defect image gray value；
S202-3. the gray value of the defect image obtained through step S202-2 processing is mapped to one by one between (0,255)
Gray value on.
The step S204 following steps：
S204-1. the size and wave band number of the defect image obtained through step S203 processing are obtained；
S204-2. gray scale stretching is carried out successively to each wave band of defect image with gray scale stretching formula；
S204-3. gray scale stretching result is preserved.
The step S304 following steps：
S304-1. the mode of defect image gray value is obtained, i.e., the frequency of occurrences is highest a certain in image grey level histogram
Gray value is set as gm；
S304-2. all gray values in defect image are selected and are less than gmPoint, and by these point gray values be set as gm；
S304-3. according to normalization upper limit H and normalization lower limit 1, to the defect map obtained through step S304-2 processing
As being normalized, the defect image obtained after normalized
D=1+ ((I-Imin)/(Imax-Imin))*(H-1)。
Compared with prior art, the present invention having the following advantages that and advantageous effect：
1, by changing (the 2 of structure in the range of (1,10)n+1)×(2n+ 1) size of n can give birth in two-dimensional array
At various sizes of defect image.
2, due to hA、hB、hCAnd hDIt is random number, therefore the defect image shape generated every time has randomness, meets scarce
The uncertain feature of sunken shape, to reach sense of reality effect.
3, it is efficient to generate defect sample by the present invention.
Description of the drawings
Fig. 1 is a kind of flow chart of casting shrinking hole defect image generation method of the present invention；
Fig. 2 is the diamond-square algorithm flow charts of the present invention；
Fig. 3 is the diamond-square algorithm iteration point shape schematic diagrames of the present invention；
Fig. 4 a, 4b, 4c are respectively the first, second, third sample for surveying defect；
Fig. 5 a, 5b, 5c are respectively that the first, second, third sample of actual measurement defect is flawless after textures synthesis is handled
Product background image；
Fig. 6 a, 6b, 6c are respectively the first, second, third defect image generated based on diamond-square algorithms；
Fig. 7 a, 7b, 7c are respectively the first, second, third emulation defect sample, wherein the first emulation defect sample is actual measurement
The first sample of defect flawless product background image after textures synthesis is handled is given birth to based on diamond-square algorithms
At the first defect image synthesis emulation defect sample, second emulation defect sample be actual measurement defect the second sample through texture
Flawless product background image is synthesized with based on the second defect image that diamond-square algorithms generate after synthesis processing
Emulation defect sample, third emulation defect sample be actual measurement defect the flawless production after textures synthesis is handled of third sample
The emulation defect sample that product background image is synthesized with the third defect image generated based on diamond-square algorithms.
Specific implementation mode
Present invention will now be described in further detail with reference to the embodiments and the accompanying drawings, but embodiments of the present invention are unlimited
In this.
As shown in Figure 1, a kind of casting shrinking hole defect image generation method based on diamond-square algorithms, including lack
Image generation, defect image conversion process, fusion product structural context three phases are fallen into, the defect image generation phase includes
Following steps：
S1. one (2 is builtn+1)×(2n+ 1) two-dimensional array, the two-dimensional array are empty two-dimensional array, the value range of n
For (1,10)；
S2. diamond-square algorithms are utilized to generate (2n+1)×(2n+ 1) altitude data of size, and accordingly by it
It is stored in the two-dimensional array that step S1 is established；
S3. the altitude data in two-dimensional array is normalized；
S4. the altitude data after normalized is multiplied with the maximum gradation value of 8 gray-value images i.e. 255, and will
It is as the defect image generated；
The defect image conversion process stage includes the following steps：
S5. histogram equalization is carried out to the defect image obtained through step S4 processing；
S6. corresponding thresholding is carried out to the defect image obtained through step S5 processing；
S7. gaussian filtering is carried out to the image obtained through step S6 processing；
S8. gray scale stretching is carried out to the image obtained through step S7 processing；
S9. step S5~S8 is repeated, the defect image after conversion process is obtained；
The fusion product structural context stage includes the following steps：
S10. the image with defect in the true casting in interception part；
S11. after textures synthesis is handled, background image I is used as after the image of step S10 is removed defect；
S12. its maximum gradation value I is detected to background image I calculatingmax, minimum gradation value Imin, calculate H=255/Imax
Value；
S13. using the H values in step S12 as the normalization upper limit of defect image, and the defect map to being obtained by step S9
As being normalized；
S14. background image I and defect image D are subjected to dot product, obtain final defect image M=I*D.
Specifically, as shown in Fig. 2, the step S2 includes the following steps：
S2-1. (the 2 of step S1 generationsn+1)×(2n+ 1) two-dimensional array is (2n+1)×(2n+ 1) square of a pixel,
In the present invention, as shown in figure 3, the i.e. square iterative fractal signal using the square of 5*5 as diamond-square algorithms
Figure is respectively set as h with the height value of random value initialization square four angle points A, B, C, D of range (- 1,1)A、hB、hC、
hD, and according to hA、hB、hC、hDRandom value of the average value of summation plus a range in (- 1,1)To determine the square center
The height value of point M, then the height value of central point M be
S2-2. square is divided into 4 son squares, and according to hA、hB、hC、hDWith hMThe average value of summation adds one
Range determines the height value of this 4 son square boundaries midpoint E, F, G, H in the random value of (- 1,1), then its height value is distinguished
For
S2-3. step S2-2 is repeated, recurrence divides each son square, the elevation until determining all elevational points of square
Value, and it is stored correspondingly in the two-dimensional array that step S1 is established.
The step S3 includes the following steps：
S3-1. the minimum value of altitude data in two-dimensional array is found out, and takes its absolute value, is set as hmin；
S3-2., the altitude data of all the points in two-dimensional array is added to the absolute value that step S3-1 is obtained, i.e. h' respectivelyA=
hA+hmin、h'B=hB+hmin..., and as new altitude data, that is, h'A、h'B... it is stored in two-dimensional array；
S3-3. the maximum value h of altitude data in the two-dimensional array that step S3-2 is obtained is found outmax；
S3-4. the maximum value obtained altitude data all in two-dimensional array divided by step S3-3 is completed to high number of passes
According to normalized.
The step S6 includes the following steps：
S6-1. according to Ostu maximum variance between clusters, the segmentation threshold of the defect image obtained through step S6 processing is found out；
S6-2. the point that all gray values in defect image are less than segmentation threshold is selected, and the gray value of these points is arranged
For the average value of defect image gray value；
S6-3. the ash gray value of the defect image obtained through step S6-2 processing being mapped to one by one between (0,255)
In angle value.
The step S8 following steps：
S8-1. the size and wave band number of the defect image obtained through step S7 processing are obtained；
S8-2. gray scale stretching is carried out successively to each wave band of defect image with gray scale stretching formula；
S8-3. gray scale stretching result is preserved.
The step S13 following steps：
S13-1. the mode of defect image gray value, i.e., the highest a certain ash of the frequency of occurrences in image grey level histogram are obtained
Angle value is set as gm；
S13-2. all gray values in defect image are selected and are less than gmPoint, and by these point gray values be set as gm；
S13-3. according to normalization upper limit H and normalization lower limit 1, to the defect image obtained through step S13-2 processing
It is normalized, the defect image obtained after normalized
D=1+ ((I-Imin)/(Imax-Imin))*(H-1)。
Fig. 4 a, Fig. 4 b, the real defect sample that Fig. 4 c are actual acquisition；Fig. 5 a, Fig. 5 b, Fig. 5 c are by above-mentioned real defect
Sample carries out textures synthesis processing, makes flawless background image, this step is to preferably compare actual measurement defect
Sample and the effect after emulation defect image and real background image co-registration；Fig. 6 a, Fig. 6 b, Fig. 6 c are shown based on diamond-
The defect image that square algorithms generate, the defect image shape generated every time as seen from the figure have randomness, meet defect
The uncertain feature of shape；Fig. 7 a, Fig. 7 b, Fig. 7 c are shown defect image and the conjunction of casting background image in the embodiment of the present invention
At defect sample, evaluated and tested by subjective vision, the effect of synthesis have compared with surveying defect sample in terms of the sense of reality compared with
High recognition rate.
The above embodiment is a preferred embodiment of the present invention, but embodiments of the present invention are not by above-described embodiment
Limitation, it is other it is any without departing from the spirit and principles of the present invention made by changes, modifications, substitutions, combinations, simplifications,
Equivalent substitute mode is should be, is included within the scope of the present invention.
Claims (5)
1. a kind of casting shrinking hole defect image generation method, which is characterized in that generated including defect image, at defect image transformation
Reason, fusion product structural context three phases, specially：
S1. defect image generation phase：
S101. one (2 is builtn+1)×(2n+ 1) two-dimensional array, the two-dimensional array are empty two-dimensional array, and the value range of n is
(1,10)；
S102. diamond-square algorithms are utilized to generate (2n+1)×(2n+ 1) altitude data of size, and it is accordingly deposited
It stores up in the two-dimensional array that step S101 is established；
Specifically include following steps：
S102-1. (the 2 of step S101 generationsn+1)×(2n+ 1) two-dimensional array is (2n+1)×(2n+ 1) square of a pixel,
The i.e. square iterative fractal schematic diagram using the square of 5*5 as diamond-square algorithms, it is random with range (- 1,1)
The height value of value initialization square four angle points A, B, C, D, are respectively set as hA、hB、hC、hD, and according to hA、hB、hC、hDAlways
Random value of the average value of sum plus a range in (- 1,1)Determine the height value of square center point M, then central point
The height value of M is
S102-2. square is divided into 4 son squares, and according to the elevation of the height value of angle point A, B, C, D and central point M
The random value of value and range (- 1,1) determines the height value of this 4 son square boundaries midpoint E, F, G, H respectively, then its height value
Respectively
S102-3. step S102-2 is repeated, recurrence divides each son square, the elevation until determining all elevational points of square
Value, and it is stored correspondingly in the two-dimensional array that step S101 is established；
S103. the altitude data in two-dimensional array is normalized；
S104. the altitude data after normalized is multiplied with the maximum gradation value of 8 gray-value images i.e. 255, and by its
As the defect image generated；
S2. the defect image conversion process stage：
S201. histogram equalization is carried out to the defect image obtained through step S104 processing；
S202. corresponding thresholding is carried out to the defect image obtained through step S201 processing；
S203. gaussian filtering is carried out to the image obtained through step S202 processing；
S204. gray scale stretching is carried out to the image obtained through step S203 processing；
S205. step S201~S204 is repeated, the defect image after conversion process is obtained；
S3. the fusion product structural context stage：
S301. the image with defect in the true casting in interception part；
S302. after textures synthesis is handled, background image I is used as after the image of step S301 is removed defect；
S303. its maximum gradation value I is detected to background image I calculatingmax, minimum gradation value Imin, calculate H=255/Imax's
Value；
S304. using the H values in step S303 as the normalization upper limit of defect image, and the defect map to being obtained by step S205
As being normalized to obtain defect image D；
S305. background image I and defect image D are subjected to dot product, obtain final defect image M=I*D.
2. casting shrinking hole defect image generation method according to claim 1, which is characterized in that the step S103 includes
Following steps：
S103-1. the minimum value of altitude data in two-dimensional array is found out, and takes its absolute value, is set as hmin；
S103-2., the altitude data of all the points in two-dimensional array is added to the absolute value that step S103-1 is obtained, i.e. h' respectivelyA=
hA+hmin、h'B=hB+hmin..., and as new altitude data, that is, h'A、h'B... it is stored in two-dimensional array；
S103-3. the maximum value h of altitude data in the two-dimensional array that step S103-2 is obtained is found outmax；
S103-4. the maximum value obtained altitude data all in two-dimensional array divided by step S103-3 is completed to high number of passes
According to normalized.
3. casting shrinking hole defect image generation method according to claim 1, which is characterized in that the step S202 includes
Following steps：
S202-1. according to Ostu maximum variance between clusters, the segmentation threshold of the defect image obtained through step S201 processing is found out；
S202-2. the point that all gray values in defect image are less than segmentation threshold is selected, and sets the gray value of these points to
The average value of defect image gray value；
S202-3. the ash gray value of the defect image obtained through step S202-2 processing being mapped to one by one between (0,255)
In angle value.
4. casting shrinking hole defect image generation method according to claim 1, which is characterized in that the step S204 or less
Step：
S204-1. the size and wave band number of the defect image obtained through step S203 processing are obtained；
S204-2. gray scale stretching is carried out successively to each wave band of defect image with gray scale stretching formula；
S204-3. gray scale stretching result is preserved.
5. casting shrinking hole defect image generation method according to claim 1, which is characterized in that the step S304 or less
Step：
S304-1. the mode of defect image gray value, i.e., the highest a certain gray scale of the frequency of occurrences in image grey level histogram are obtained
Value, is set as gm；
S304-2. all gray values in defect image are selected and are less than gmPoint, and by these point gray values be set as gm；
S304-3. according to normalization upper limit H and normalization lower limit 1, to the defect image that is obtained through step S304-2 processing into
Row normalized, the defect image obtained after normalized
D=1+ ((I-Imin)/(Imax-Imin))*(H-1)。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510437966.6A CN105095581B (en) | 2015-07-22 | 2015-07-22 | A kind of casting shrinking hole defect image generation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510437966.6A CN105095581B (en) | 2015-07-22 | 2015-07-22 | A kind of casting shrinking hole defect image generation method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105095581A CN105095581A (en) | 2015-11-25 |
| CN105095581B true CN105095581B (en) | 2018-09-14 |
Family
ID=54576006
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510437966.6A Active CN105095581B (en) | 2015-07-22 | 2015-07-22 | A kind of casting shrinking hole defect image generation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105095581B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106127780B (en) * | 2016-06-28 | 2019-01-18 | 华南理工大学 | A kind of curved surface defect automatic testing method and its device |
| CN106530292B (en) * | 2016-10-28 | 2019-07-05 | 鞍钢未来钢铁研究院有限公司 | A kind of steel strip surface defect image Fast Identification Method based on line scan camera |
| CN107358636B (en) * | 2017-06-16 | 2020-04-28 | 华南理工大学 | Loose defect image generation method based on texture synthesis |
| CN109360231B (en) * | 2018-10-25 | 2022-01-07 | 哈尔滨工程大学 | Sea ice remote sensing image simulation method for generating confrontation network based on fractal depth convolution |
| CN110263640A (en) * | 2019-05-16 | 2019-09-20 | 东软集团股份有限公司 | Slab shrinkage cavity recognition methods, device, storage medium and electronic equipment |
| CN114266802B (en) * | 2022-03-01 | 2022-07-08 | 聚时科技(江苏)有限公司 | Method for generating irregular hole morphological defect image |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7076093B2 (en) * | 2002-09-16 | 2006-07-11 | Lee Shih-Jong J | Structure-guided image inspection |
| CN1936963A (en) * | 2006-09-14 | 2007-03-28 | 清华大学 | Virtual oil-layer natural simulation system and method |
| CN101021491A (en) * | 2006-12-28 | 2007-08-22 | 华南理工大学 | Automatic decting method and device for wheel hub casting fault based on image understanding |
| CN104091327A (en) * | 2014-06-19 | 2014-10-08 | 华南理工大学 | Method and system for generating dendritic shrinkage porosity defect simulation image of casting |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101706206B1 (en) * | 2013-05-13 | 2017-02-15 | 한국전자통신연구원 | Method for describing terrain, method for creating terrain primitives, and apparatus using the methods |
-
2015
- 2015-07-22 CN CN201510437966.6A patent/CN105095581B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7076093B2 (en) * | 2002-09-16 | 2006-07-11 | Lee Shih-Jong J | Structure-guided image inspection |
| CN1936963A (en) * | 2006-09-14 | 2007-03-28 | 清华大学 | Virtual oil-layer natural simulation system and method |
| CN101021491A (en) * | 2006-12-28 | 2007-08-22 | 华南理工大学 | Automatic decting method and device for wheel hub casting fault based on image understanding |
| CN104091327A (en) * | 2014-06-19 | 2014-10-08 | 华南理工大学 | Method and system for generating dendritic shrinkage porosity defect simulation image of casting |
Non-Patent Citations (7)
| Title |
|---|
| A Threshold Selection Method from Gray-Level Histograms;NOBUYUKI OTSU;《IEEE Transactions on Systems,Man and Cybernetics》;19790131;第9卷(第1期);第62-66页 * |
| Application of a new image segmentation method to detection of defects in castings;Yinggan Tang等;《The International Journal of Advanced Manufacturing Technology》;20080917;第43卷(第5-6期);第431-439页 * |
| Casting Defect Analysis using Design of Experiments(DoE) and Computer Aided Casting Simulation Technique;Uday A.Dabade等;《Forty Sixth CIRP Conference on Manufacturing Systems 2013》;20131231;第616-621页 * |
| Diamond-Square细分算法的三维地形交互生成系统的实现;吴凤娟等;《计算机系统应用》;20080515(第05期);第66-68+76页 * |
| 基于diamond-square算法的数字地形模型构建与三维可视化研究;刘庆元等;《测绘工程》;20140205;第23卷(第02期);第1-4+9页 * |
| 基于灰度值算法与Diamond-Square算法的三维地形生成;程义等;《电脑开发与应用》;20100905;第23卷(第9期);第33-35+38页 * |
| 基于语义描述的铸造缺陷图像生成研究;汪玉琳;《中国优秀硕士学位论文全文数据库 信息科技辑》;20150115(第01期);第I138-886页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105095581A (en) | 2015-11-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105095581B (en) | A kind of casting shrinking hole defect image generation method | |
| CN111524135B (en) | Method and system for detecting defects of tiny hardware fittings of power transmission line based on image enhancement | |
| CN104299260B (en) | Contact network three-dimensional reconstruction method based on SIFT and LBP point cloud registration | |
| CN110415167B (en) | Rough surface crack generation method and test system based on digital image technology | |
| CN113450307A (en) | Product edge defect detection method | |
| CN107358636B (en) | Loose defect image generation method based on texture synthesis | |
| CN104299250B (en) | Front face image synthetic method and system based on prior model | |
| CN107194872A (en) | Remote sensed image super-resolution reconstruction method based on perception of content deep learning network | |
| CN107767453B (en) | Building LIDAR point cloud reconstruction optimization method based on rule constraint | |
| CN110866531A (en) | Building feature extraction method and system based on three-dimensional modeling and storage medium | |
| CN106504277B (en) | A kind of improved ICP point cloud autoegistration method | |
| CN111754618B (en) | Object-oriented live-action three-dimensional model multi-level interpretation method and system | |
| CN108305327A (en) | A kind of image rendering method | |
| CN111553858A (en) | Image restoration method and system based on generation countermeasure network and application thereof | |
| CN104091327A (en) | Method and system for generating dendritic shrinkage porosity defect simulation image of casting | |
| CN109064556B (en) | Landform high-precision simulation modeling system facing ISR | |
| CN111028335A (en) | Point cloud data block surface patch reconstruction method based on deep learning | |
| CN113240790A (en) | Steel rail defect image generation method based on 3D model and point cloud processing | |
| CN106815879A (en) | A kind of quick texture synthesis method based on LBP features | |
| CN114529689B (en) | Ceramic cup defect sample amplification method and system based on antagonistic neural network | |
| CN116402973A (en) | Oblique photography model optimization method and system based on LOD reconstruction | |
| CN114332411A (en) | Method for generating three-dimensional graph real-time grid | |
| CN112002019B (en) | Method for simulating character shadow based on MR mixed reality | |
| CN114119874A (en) | Single image reconstruction high-definition 3D face texture method based on GAN | |
| CN113902887A (en) | Three-dimensional visual edge generation method, system, computer and readable storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20190530 Address after: 518107 Unit B3-4A1, Guangming Science Park, China Merchants Bureau, 3009 Guangming Street Sightseeing Road, Guangming New District, Shenzhen City, Guangdong Province Patentee after: Deep sea precision technology (Shenzhen) Co., Ltd. Address before: 510640 No. five, 381 mountain road, Guangzhou, Guangdong, Tianhe District Patentee before: South China University of Technology |