CN107978004A - Sinking shaft mural painting archaeology drawing Fast Generation based on heuristic route - Google Patents
Sinking shaft mural painting archaeology drawing Fast Generation based on heuristic route Download PDFInfo
- Publication number
- CN107978004A CN107978004A CN201711120926.4A CN201711120926A CN107978004A CN 107978004 A CN107978004 A CN 107978004A CN 201711120926 A CN201711120926 A CN 201711120926A CN 107978004 A CN107978004 A CN 107978004A
- Authority
- CN
- China
- Prior art keywords
- point
- route
- grad
- pixel
- gradient
- 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.)
- Pending
Links
- 238000010422 painting Methods 0.000 title claims abstract description 25
- 238000000034 method Methods 0.000 claims abstract description 28
- 238000012545 processing Methods 0.000 claims abstract description 13
- 238000000205 computational method Methods 0.000 claims description 2
- 230000001629 suppression Effects 0.000 claims description 2
- 238000013461 design Methods 0.000 abstract description 3
- 238000005516 engineering process Methods 0.000 description 4
- 238000003708 edge detection Methods 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 238000013527 convolutional neural network Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 235000013399 edible fruits Nutrition 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000033228 biological regulation Effects 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 230000006378 damage Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000007598 dipping method Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 239000003973 paint Substances 0.000 description 1
- 230000008092 positive effect Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T11/00—2D [Two Dimensional] image generation
- G06T11/001—Texturing; Colouring; Generation of texture or colour
-
- G06T5/70—
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
Abstract
The present invention relates to a kind of sinking shaft mural painting archaeology drawing Fast Generation based on heuristic route, its technical characterstic is：For the mural painting image I of input, noise is removed using Gaussian filter, obtains the result I after denoising0；I is calculated using the finite difference of single order local derviation0Gradient magnitude figure G and edge orientation map D, and to gradient magnitude carry out adjusting thresholds, generate adjusting thresholds after image be denoted as G1；According to gradient magnitude figure G1, edge orientation map D and setting anchor point threshold value extract anchor point；Anchor point is attached using heuristic routing algorithm to obtain edge graph E.Present invention design is reasonable, can describe edge feature exactly, so as to generate clean, continuous single pixel side, meet archaeology drawing requirements.The present invention automatically processes generation by computer completely, and without further contour connection process, processing speed is quickly.
Description
Technical field
The invention belongs to technical field of image processing, especially a kind of sinking shaft mural painting archaeology based on heuristic route is drawn
Fast Generation.
Background technology
It is that cartography is applied to archaeological work and a special kind of skill of research that archaeology, which is drawn, it is in the form of lines to archaeology
Traces, remains carry out science record and statement, are an essential parts in archaeological work.Tradition archaeology Plot Work is main
Taking hand dipping, this mode can easily cause historical relic secondary destruction, and the result of Hand drawing also varies with each individual with drawing, and
There are the shortcomings that speed is slow, efficiency is low, not accurate enough.
With the development of computer and graphics technology, archaeology drawing digitlization and the hair automated is greatly facilitated
Exhibition, researcher have attempted to use the equipment dedicated for archaeology shooting and measurement, and research to automatically generate archaeology and draw
Technology and software.Archaeology drawing digital is related to edge detection into key technology.The algorithm of edge detection has many kinds, passes
System is had a wide range of applications such as Canny operators, but these detectors only take into account local change dramatically, particularly face
The change dramatically of color, brightness etc..2013, researcher learnt joint color, brightness, gradient etc. with the method for data-driven
Feature does edge detection, popular method such as gPb, StructuredEdge, but such algorithm speed and conventional method ratio
Speed is slower.Method based on deep learning etc. is tasted using convolutional neural networks CNN, is explored using embedded many high levels, more
The information of scale, still, it generally requires substantial amounts of training data and parameter regulation process.
Archaeology, which is drawn, its unique technical requirements.Archaeology drawing be using descriptive geometry as its theoretical foundation, it is exquisite
It is specification, pursuit is the Science Effice for truely and accurately reflecting things original appearance, and without ideological content, it is single pixel as a result to require
Expression.In addition sinking shaft mural painting is formed at grottoes more, and quantity is more, and area is big, therefore quickly generation just has practical guidance meaning
Justice.
By retrieving, the domestic patent of invention related with archaeology drawing is concentrated mainly on the drawing of archaeology implements and classifies at present
And mural painting Image Acquisition." a kind of view-based access control model relies on the historical relic line chart method for drafting of Curvature Estimate to patent document
(CN201611223161.2) ", its archaeology drawing generation for being mainly used to solve three-dimensional implements, is not suitable for mural painting historical relic, with
This patent content is simultaneously uncorrelated.
The content of the invention
It is overcome the deficiencies in the prior art mesh of the invention, proposes that a kind of design rationally, fast and accurately is based on opening
The sinking shaft mural painting archaeology drawing Fast Generation of hairdo route.
The present invention solves its technical problem and takes following technical scheme to realize：
A kind of sinking shaft mural painting archaeology drawing Fast Generation based on heuristic route, it is characterised in that including following step
Suddenly：
Step 1, the mural painting image I for input, remove noise using Gaussian filter, obtain the result after denoising
I0；
Step 2, using the finite difference of single order local derviation calculate I0Gradient magnitude figure G and edge orientation map D, and to ladder
Spend amplitude and carry out adjusting thresholds, generate the image after adjusting thresholds and be denoted as G1；
Step 3, according to gradient magnitude figure G, edge orientation map D and the anchor point threshold value of setting extract anchor point；
Step 4, be attached anchor point to obtain edge graph E using heuristic routing algorithm.
2nd, the sinking shaft mural painting archaeology drawing Fast Generation according to claim 1 based on heuristic route, its
It is characterized in that：The Gaussian filter of the step 1 uses 5 × 5 Gaussian kernel, its parameter σ is arranged to 1.
3rd, the sinking shaft mural painting archaeology drawing Fast Generation according to claim 1 based on heuristic route, its
It is characterized in that：The concrete processing procedure of the step 2 is：
(1) Grad of each pixel is calculated using Sobel operators to obtain the gradient magnitude of I0, gradient magnitude figure
The computational methods of G are as follows：
Wherein, GxAnd GyThe gradient of pixel (x, y) both horizontally and vertically is represented respectively；Grad G
[x, y]=| Gx|+|Gy| to replace, thus form gradient amplitude figure G；
(2), the horizontal gradient G of compared pixels point is passed throughxWith vertical gradient GySize calculate edge orientation map；If | Gx
|≥|Gy|, then it is assumed that a vertical side can pass through the pixel, and the edge direction of the pixel is level, i.e. and D [x, y]=
h；Otherwise, then it is assumed that a horizontal side will pass through the pixel, D [x, y]=v；Thus edge orientation map D is formed；
(3) non-maxima suppression is carried out to gradient magnitude：Adjusting thresholds are carried out to Grad to eliminate some weak pixels, threshold
Image after value adjustment is denoted as G1。
4th, the sinking shaft mural painting archaeology drawing Fast Generation according to claim 1 based on heuristic route, its
It is characterized in that：The concrete processing procedure of the step 3 is：
(1) for pending pixel (x, y), if its edge direction is horizontal direction, i.e. D [x, y]=h, then press
Following formula judge the whether big Grad for descending abutment points thereon of the Grad of the point：
G[x,y]-G[x,y-1]≥AnchorThresh and
G[x,y]-G[x,y+1]≥AnchorThresh
If meeting conditions above, which is anchor point；
(2) if the edge direction at pending pixel (x, y) place is vertical direction, i.e. D [x, y]=v, then by following
Formula judges whether the Grad of the point is more than the Grad of its left and right abutment points：
G[x,y]-G[x-1,y]≥AnchorThresh and
G[x,y]-G[x+1,y]≥AnchorThresh
If meeting conditions above, which is anchor point；
AnchorThresh in above-mentioned formula is the threshold value of setting.
5th, the sinking shaft mural painting archaeology drawing Fast Generation according to claim 1 based on heuristic route, its
It is characterized in that：The concrete processing procedure of the step 4 is：
(1) if the direction D [x, y] of pixel (x, y) is horizontal direction, and gradient G [x, y] is more than 0, E [x, y] and does not have
Labeled as edge, then the point is marked as marginal point, i.e. E [x, y]=EDGE, then route to the left, to the right respectively；It route to the left
When compare three points (x-1, y-1), (x-1, y), the gradient magnitudes of (x-1, y+1) of its left adjoining, choose wherein that Grad is most
Big pixel is connected with point (x, y), and continues iteration route since the point；It route to the right, is three neighbours on the right side of comparison
Point (x+1, y-1), (x+1, y), the size of (x+1, y+1), choose the wherein point of gradient maximum and are connected with (x, y) point, and continue
Iteration is route since the point newly chosen；
(2) when the direction D [x, y] of pixel is vertical direction, then it is route upwards, downwards respectively from (x, y) place, respectively
The point for obtaining Grad maximum in three points of upper and lower adjoining is attached, and route as next step iteration
Starting point；
(3) until iterating at non-boundary point, i.e., Grad is 0, or all boundary points have stepped through to finish and then may be used
Terminate iteration.
The advantages and positive effects of the present invention are：
Present invention design is reasonable, then it is calculated by carrying out gaussian filtering process to sinking shaft image, removing noise processed
Gradient magnitude and edge direction simultaneously search out anchor point accordingly, and finally carrying out anchor point using heuristic routing algorithm connects to be formed intentionally
The edge contour of justice, can describe edge feature exactly, so as to generate clean, continuous single pixel side, meet archaeology and paint
Figure demand.The present invention automatically processes generation by computer completely, and without further contour connection process, processing speed is quickly.
Brief description of the drawings
Fig. 1 is the process chart of the present invention；
Fig. 2 is the Intelligent routing search strategy schematic diagram of the present invention；
Fig. 3 is the handling result schematic diagram of the present invention；
Fig. 4 is the comparative result figure that the present invention is directed to sinking shaft image with Canny, gPb, HED method；
Fig. 5 is time-consuming comparison figure of the present invention with the processing of Canny, gPb, HED method with piece image.
Embodiment
The embodiment of the present invention is further described below in conjunction with attached drawing.
A kind of sinking shaft mural painting archaeology drawing Fast Generation (EGSR) based on heuristic route, as shown in Figure 1, including
Following steps：
Step 1, the mural painting image I for input, remove noise using Gaussian filter, obtain the result after denoising
I0。
In this step 1 is arranged to using the Gaussian kernel of one 5 × 5, parameter σ.
Step 2, using the finite difference of single order local derviation calculate I0Gradient magnitude figure G and edge orientation map D, and to ladder
Spend amplitude and carry out adjusting thresholds, generate the image after adjusting thresholds and be denoted as G1。
The concrete processing procedure of this step is as follows：
(1) Grad of each pixel is calculated using Sobel operators to obtain I0Gradient magnitude, the calculating side of G
Method is as follows：
Wherein GxAnd GyThe gradient of pixel (x, y) both horizontally and vertically is represented respectively.In order to obtain faster
Computational efficiency, Grad can use G [x, y]=| Gx|+|Gy| to replace, thus form gradient amplitude figure G.
(2) edge orientation map is calculated, it is the horizontal gradient G by compared pixels pointxWith vertical gradient GySize count
Calculate.If | Gx|≥|Gy|, then it is assumed that a vertical side can pass through the pixel, and the edge direction of the pixel is level, i.e.,
D [x, y]=h；Otherwise, then it is assumed that a horizontal side will pass through the pixel, D [x, y]=v.Thus edge direction is formed
Scheme D.
(3) adjusting thresholds are carried out to gradient magnitude to eliminate some weak pixels, the image after adjusting thresholds is denoted as G1。
Step 3, according to gradient magnitude figure G1Anchor point is extracted with edge orientation map D, and the anchor point threshold value of setting
Anchor。
In this step, (x, y) is made to represent pending pixel, AnchorThresh is the threshold value of setting.Anchor point carries
Take process as follows：
(1) for pixel (x, y), if its edge direction is horizontal direction, i.e. D [x, y]=h, then judge the point
The whether big Grad for descending abutment points thereon of Grad.Judgment formula is as follows：
If meeting conditions above, which is anchor point.
(2) edge direction such as fruit dot (x, y) place is vertical direction, i.e. D [x, y]=v, then judging the Grad of the point is
The no Grad more than its left and right abutment points.Judgment formula is：
G[x,y]-G[x-1,y]≥AnchorThresh and
G[x,y]-G[x+1,y]≥AnchorThresh (3)
If meeting conditions above, which is anchor point.
Step 4, be attached anchor point to obtain edge graph E using heuristic routing algorithm.
By taking gradient magnitude figure and edge orientation map that (a) in Fig. 2 is provided as an example, illustrate the specific treated of this step
Journey：
(1) if the direction D [x, y] of pixel (x, y) is horizontal direction, and gradient G [x, y] is more than 0, E [x, y] and does not have
Labeled as edge, then the point is marked as marginal point, i.e. E [x, y]=EDGE, then route to the left, to the right, see in Fig. 2 respectively
(b).Compare three points (x-1, y-1), (x-1, y), the gradient magnitudes of (x-1, y+1) of its left adjoining, choosing when routeing to the left
Take the wherein pixel of Grad maximum to be connected with point (x, y), and continue iteration route since the point.And route to the right, it is
Compare right side three adjoint points (x+1, y-1), (x+1, y), the sizes of (x+1, y+1), choose wherein the point of gradient maximum and (x,
Y) point connection, and continue the iteration since the point newly chosen and route.
(2) when the direction D [x, y] of pixel is vertical direction, then it is route upwards, downwards respectively from (x, y) place, sees figure
(c) in 2.Method for routing is similar with to from left to right route, obtains Grad maximum in three points of upper and lower adjoining respectively
Put to be attached, and as the starting point of next step iteration route.
(3) until iterating at non-boundary point, i.e., Grad is 0, or all boundary points have stepped through to finish and then may be used
Terminate iteration.
In order to verify the effect of the present invention, put down using the Visual C++2010 under Windows7 systems as experiment simulation
Platform.Mural painting sinking shaft image is selected to amount to 100 width images as test set.Horizontal/vertical resolution is 300dpi, and pixel number is
512×512.As shown in Figures 3 and 4, (a) is to test mural painting sinking shaft image in Fig. 3, and (b) is the processing knot of the present invention in Fig. 3
Fruit, (a) in Fig. 4 are the handling result using Canny methods, and (b) in Fig. 4 is the handling result using gPb methods, Fig. 4
In (c) be using HED methods handling result, it can be seen that test image is handled using this patent institute extracting method,
Good treatment effect is obtained, as shown in Figures 3 and 4.The average treatment speed of the present invention is 14ms, and processing speed is substantially real
When, it disclosure satisfy that requirement, as shown in Figure 5.
It is emphasized that embodiment of the present invention is illustrative, rather than it is limited, therefore present invention bag
The embodiment being not limited to described in embodiment is included, it is every by those skilled in the art's technique according to the invention scheme
The other embodiment drawn, also belongs to the scope of protection of the invention.
Claims (5)
1. a kind of sinking shaft mural painting archaeology drawing Fast Generation based on heuristic route, it is characterised in that including following step
Suddenly：
Step 1, the mural painting image I for input, remove noise using Gaussian filter, obtain the result I after denoising0；
Step 2, using the finite difference of single order local derviation calculate I0Gradient magnitude figure G and edge orientation map D, and to gradient width
Value carries out adjusting thresholds, generates the image after adjusting thresholds and is denoted as G1；
Step 3, according to gradient magnitude figure G1, edge orientation map D and setting anchor point threshold value extract anchor point；
Step 4, be attached anchor point to obtain edge graph E using heuristic routing algorithm.
2. the sinking shaft mural painting archaeology drawing Fast Generation according to claim 1 based on heuristic route, its feature
It is：The Gaussian filter of the step 1 uses 5 × 5 Gaussian kernel, its parameter σ is arranged to 1.
3. the sinking shaft mural painting archaeology drawing Fast Generation according to claim 1 based on heuristic route, its feature
It is：The concrete processing procedure of the step 2 is：
(1) Grad of each pixel is calculated using Sobel operators to obtain the gradient magnitude of I0, gradient magnitude figure G's
Computational methods are as follows：
<mrow>
<mi>G</mi>
<mo>&lsqb;</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>&rsqb;</mo>
<mo>=</mo>
<msqrt>
<mrow>
<msubsup>
<mi>G</mi>
<mi>x</mi>
<mn>2</mn>
</msubsup>
<mo>+</mo>
<msubsup>
<mi>G</mi>
<mi>y</mi>
<mn>2</mn>
</msubsup>
</mrow>
</msqrt>
</mrow>
Wherein, GxAnd GyThe gradient of pixel (x, y) both horizontally and vertically is represented respectively；Grad is with G [x, y]
=| Gx|+|Gy| to replace, thus form gradient amplitude figure G；
(2), the horizontal gradient G of compared pixels point is passed throughxWith vertical gradient GySize calculate edge orientation map；If | Gx|≥|
Gy|, then it is assumed that a vertical side can pass through the pixel, and the edge direction of the pixel is horizontal, i.e. D [x, y]=h；It is no
Then, then it is assumed that a horizontal side will pass through the pixel, D [x, y]=v；Thus edge orientation map D is formed；
(3) non-maxima suppression is carried out to gradient magnitude：Adjusting thresholds are carried out to Grad to eliminate some weak pixels, threshold value tune
Image after whole is denoted as G1。
4. the sinking shaft mural painting archaeology drawing Fast Generation according to claim 1 based on heuristic route, its feature
It is：The concrete processing procedure of the step 3 is：
(1) for pending pixel (x, y), if its edge direction is horizontal direction, i.e. D [x, y]=h, then by following
Formula judges the whether big Grad for descending abutment points thereon of the Grad of the point：
G[x,y]-G[x,y-1]≥AnchorThresh and
G[x,y]-G[x,y+1]≥AnchorThresh
If meeting conditions above, which is anchor point；
(2) if the edge direction at pending pixel (x, y) place is vertical direction, i.e. D [x, y]=v, then by following formula
Judge whether the Grad of the point is more than the Grad of its left and right abutment points：
G[x,y]-G[x-1,y]≥AnchorThresh and
G[x,y]-G[x+1,y]≥AnchorThresh
If meeting conditions above, which is anchor point；
AnchorThresh in above-mentioned formula is the threshold value of setting.
5. the sinking shaft mural painting archaeology drawing Fast Generation according to claim 1 based on heuristic route, its feature
It is：The concrete processing procedure of the step 4 is：
(1) if the direction D [x, y] of pixel (x, y) is horizontal direction, and gradient G [x, y] is not marked more than 0, E [x, y]
For edge, then the point is marked as marginal point, i.e. E [x, y]=EDGE, then route to the left, to the right respectively；When ratio is route to the left
Compared with three points (x-1, y-1) of its left adjoining, (x-1, y), (x-1, y+1) gradient magnitude, it is maximum to choose wherein Grad
Pixel is connected with point (x, y), and continues iteration route since the point；It route to the right, is three adjoint point (x on the right side of comparison
+ 1, y-1), (x+1, y), the size of (x+1, y+1), choose the wherein point of gradient maximum and (x, y) point and connect, and continuation is from new
The point of selection starts iteration route；
(2) when the direction D [x, y] of pixel is vertical direction, then it route upwards, downwards from (x, y) place, obtains respectively respectively
The point of Grad maximum is attached in three points of upper and lower adjoining, and as the starting point that next step iteration is route；
(3) until iterating at non-boundary point, i.e., Grad is 0, or all boundary points have stepped through and finish, and can terminate
Iteration.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711120926.4A CN107978004A (en) | 2017-11-14 | 2017-11-14 | Sinking shaft mural painting archaeology drawing Fast Generation based on heuristic route |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711120926.4A CN107978004A (en) | 2017-11-14 | 2017-11-14 | Sinking shaft mural painting archaeology drawing Fast Generation based on heuristic route |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107978004A true CN107978004A (en) | 2018-05-01 |
Family
ID=62013497
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711120926.4A Pending CN107978004A (en) | 2017-11-14 | 2017-11-14 | Sinking shaft mural painting archaeology drawing Fast Generation based on heuristic route |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107978004A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108921862A (en) * | 2018-05-28 | 2018-11-30 | 天津科技大学 | Mural painting line drawing figure generating algorithm based on convolutional neural networks |
| CN113554087A (en) * | 2021-07-20 | 2021-10-26 | 清华大学深圳国际研究生院 | Archaeological graph automatic generation method based on generation type confrontation network |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040190092A1 (en) * | 1999-09-17 | 2004-09-30 | Kia Silverbrook | Scanning device for coded data |
| JP2009026079A (en) * | 2007-07-19 | 2009-02-05 | Canon Inc | Image processing device, image processing method, and image processing program |
| CN106530317A (en) * | 2016-09-23 | 2017-03-22 | 南京凡豆信息科技有限公司 | Stick figure computer scoring and auxiliary coloring method |
| CN107256557A (en) * | 2017-05-03 | 2017-10-17 | 华南理工大学 | A kind of controllable subdivision curved surface image vector method of error |
-
2017
- 2017-11-14 CN CN201711120926.4A patent/CN107978004A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040190092A1 (en) * | 1999-09-17 | 2004-09-30 | Kia Silverbrook | Scanning device for coded data |
| JP2009026079A (en) * | 2007-07-19 | 2009-02-05 | Canon Inc | Image processing device, image processing method, and image processing program |
| CN106530317A (en) * | 2016-09-23 | 2017-03-22 | 南京凡豆信息科技有限公司 | Stick figure computer scoring and auxiliary coloring method |
| CN107256557A (en) * | 2017-05-03 | 2017-10-17 | 华南理工大学 | A kind of controllable subdivision curved surface image vector method of error |
Non-Patent Citations (1)
| Title |
|---|
| 许迪: "面向中国古代壁画的线描画生成方法", 《计算机工程》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108921862A (en) * | 2018-05-28 | 2018-11-30 | 天津科技大学 | Mural painting line drawing figure generating algorithm based on convolutional neural networks |
| CN113554087A (en) * | 2021-07-20 | 2021-10-26 | 清华大学深圳国际研究生院 | Archaeological graph automatic generation method based on generation type confrontation network |
| CN113554087B (en) * | 2021-07-20 | 2023-11-10 | 清华大学深圳国际研究生院 | Automatic archaeological line graph generation method based on generation type countermeasure network |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106780612B (en) | Object detecting method and device in a kind of image | |
| CN110807422B (en) | Natural scene text detection method based on deep learning | |
| JP7048225B2 (en) | Trained model for building area extraction | |
| CN104573731B (en) | Fast target detection method based on convolutional neural networks | |
| CN109903331B (en) | Convolutional neural network target detection method based on RGB-D camera | |
| CN108121991B (en) | Deep learning ship target detection method based on edge candidate region extraction | |
| CN108898610A (en) | A kind of object contour extraction method based on mask-RCNN | |
| CN108304873A (en) | Object detection method based on high-resolution optical satellite remote-sensing image and its system | |
| CN109409366A (en) | Distorted image correction method and device based on Corner Detection | |
| CN107038416B (en) | Pedestrian detection method based on binary image improved HOG characteristics | |
| CN101777129B (en) | Image matching method based on feature detection | |
| CN105678806B (en) | A kind of live pig action trail automatic tracking method differentiated based on Fisher | |
| CN107844795A (en) | Convolutional neural networks feature extracting method based on principal component analysis | |
| CN108614994A (en) | A kind of Human Head Region Image Segment extracting method and device based on deep learning | |
| CN106934455B (en) | Remote sensing image optics adapter structure choosing method and system based on CNN | |
| CN109376740A (en) | A kind of water gauge reading detection method based on video | |
| CN112597495B (en) | Malicious code detection method, system, equipment and storage medium | |
| CN104794455B (en) | A kind of Dongba pictograph recognition methods | |
| CN105046694A (en) | Quick point cloud registration method based on curved surface fitting coefficient features | |
| CN104504381B (en) | Non-rigid object detection method and its system | |
| CN111062331B (en) | Image mosaic detection method and device, electronic equipment and storage medium | |
| CN103218605A (en) | Quick eye locating method based on integral projection and edge detection | |
| CN106157323A (en) | The insulator division and extracting method that a kind of dynamic division threshold value and block search combine | |
| CN109886267A (en) | A kind of soft image conspicuousness detection method based on optimal feature selection | |
| CN105405138A (en) | Water surface target tracking method based on saliency detection |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180501 |