CN107220979A - A kind of method in suitable rectangular background region in Quick positioning map picture - Google Patents
A kind of method in suitable rectangular background region in Quick positioning map picture Download PDFInfo
- Publication number
- CN107220979A CN107220979A CN201710346750.8A CN201710346750A CN107220979A CN 107220979 A CN107220979 A CN 107220979A CN 201710346750 A CN201710346750 A CN 201710346750A CN 107220979 A CN107220979 A CN 107220979A
- Authority
- CN
- China
- Prior art keywords
- image
- region
- white point
- window
- size
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- 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
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/155—Segmentation; Edge detection involving morphological operators
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/194—Segmentation; Edge detection involving foreground-background segmentation
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20016—Hierarchical, coarse-to-fine, multiscale or multiresolution image processing; Pyramid transform
Abstract
The present invention discloses suitable rectangular background region method in a kind of Quick positioning map picture, including：Step 1: reading in digital picture, and digital picture is converted into bianry image；Step 2: the method split using image, it is more than or equal in split window under the precondition of embedded images size, the segmentation for image circulated, take in split window comprising the minimum region of white point as new split window continuation segmentation, until white point number that obtained split window is included is minimum and can not be further continued for segmentation, the split window is now positioned；Step 3：The rectangular window of embedded images size is used in the split window navigated in step 2 as sliding window progress sliding window search, until the white point number of a certain position just regard the region as final region for 0, or without the region that white point number is 0, just using the minimum region of white point number as last region, suitable new image insertion region so just have found.
Description
Technical field
It is used to be embedded in the invention belongs to one piece in Computer Image Processing field, more particularly to a kind of energy Quick positioning map picture
The method in the suitable rectangular background region of new images.
Background technology
The development of science and technology promotes the extensive use of digital picture, and digital picture can be very easily obtained now,
And new image, such as Quick Response Code, so as to complete the modification to digital picture are embedded in the picture.However, wanting to make insertion
New image information can not cover the important information that image-carrier is included in itself, it is necessary to which a kind of method carrys out positioning carrier in itself
Background area, and in background area select one piece of suitable region, the new image information for being embedded in.
At present, the Position Research of background area is still in initial stage in digital picture, and most research work is several
Object detection in word image, and huge progress is achieved in this respect.Itti in 1998 proposes a kind of classical show
The method for writing figure (Saliency Map), main thought is exactly the visual information according to image bottom, and the vision for providing image shows
Write region.Also there is expert to propose a kind of segmentation nomography (SM) of object detection, SM be study how the thing to arbitrary shape
Body is positioned, and calibrates object to come using the border compacted.Traditional object positioning method is sliding window searching method
(Sliding Window Search), is uniformly to be slided on picture using the window of fixed size, passes through each position
When, the feature in now window is substituted into and tested in the grader for realizing training, with grader highest in whole process
That the window's position of response is then considered as the position of object.It is the detection method of objects in images above, it would be desirable to root
Go to position background area, although finally realizing the positioning of background area, the background found again according to the object area detected
Region be still dispersed distribution in original vector image, rather than one piece of suitable region for having determined that position wants reality
The positioning of existing correct position, in addition it is also necessary to continue to carry out localization process to scattered picture background region, improve the complexity of work
Degree, while greatly reducing the speed of our work and the efficiency and quality of service.
The content of the invention
It is an object of the invention to propose a kind of method in one piece of suitable rectangular background region in Quick positioning map picture, from
And can be by new image information, such as Quick Response Code, the correct position of embedded original vector picture, without covering the weight that original picture is included
Want information.Locating speed of the present invention is fast, and locating accuracy is high, and solving all the time can only be original with destroying in embedded new image
The problem of one being selected in image integrity.
To achieve the above object, the present invention is adopted the following technical scheme that：
Suitable rectangular background region method, comprises the following steps in a kind of Quick positioning map picture：
Step 1: reading in digital picture, and digital picture is converted into bianry image；
Step 2: the method split using image, big in split window (regarding original image as split window for the first time)
Under the precondition equal to embedded images size, the segmentation circulated to image takes minimum comprising white point in split window
Region continue to split as new split window, until white point number that obtained split window is included is minimum and can not be further continued for
Split (length and width that now, 2 times of the length and width of loading image are all respectively greater than split window), now position the split window；
Step 3：The rectangular window of embedded images size is used in the split window navigated in step 2 as slip
Window carries out sliding window search, until the white point number of a certain position is 0 just using the region as final region, or does not have
White point number is 0 region, just using the minimum region of white point number as last region, so just have found suitable new image embedding
Enter region.
Preferably, step one is specifically included：
Step 1.1, digital picture is converted into grayscale image；
Step 1.2, binary conversion treatment, are carried out rim detection to grayscale image, are such as carried out using Sobel edge detectors
Rim detection, bianry image BW is converted into by grayscale image I, and it is the value corresponding to white to detect its value of the position at edge
" 1 ", its value of other positions is the value " 0 " corresponding to black.
Preferably, fitting into embedded new images and size is more than or equal to and to be embedded in figure for positioning in image one piece
As the suitable rectangular area of size position the step of two include, first have to enter the bianry image obtained in image step one
Dilation operation in row morphology so that target increases in image, tiny cavity in blank map picture, and smoothly connect object
Border, expands outwardly border；Secondly, the method exactly split using image positions one piece of suitable rectangular background in image
Region, it comprises the following steps that 2.1-2.5：
Step 2.1. first determines whether that (wherein BW top left co-ordinate is for initial carrier image BW (being denoted as B) size [m n]
(x, y)) and the size [h w] of image to be embedded in, as long as meeting condition h<M and w<N, is carried out following steps；
If step 2.2. 2*h>M, 2*w<=n, be by divided in horizontal direction by the original image after being handled through step one
The part of B1, B2, B3 tri-, wherein,
B1=BW (x:x+m-1,y:y+n/2-1)；
B2=BW (x:x+m-1,y+n/2-1:y+n-1)；
B3=BW (x:x+m-1,y+n/4-1:y+3*n/4-1)；
The minimum region of white point in B1, B2, B3 is calculated, B is denoted as, if B is B1, x, y is constant；If B is B2, x is constant,
Y=n/2+y-1；If B is B3, x is constant, y=n/3+y-1；Meanwhile, m, n is changed into B size.As long as meeting h<M and w<N is just
Perform following steps；
If step 2.3. 2*h<M, 2*w<=n, by the original image after being handled through step one
It is divided into B1, B2, B3, B4, the parts of B5 five, wherein,
B1=BW (x:x+m/2-1,y:y+n/2-1)；
B2=BW (x:x+m/2-1,y+n/2-1:y+n-1)；
B3=BW (x+m/2-1:x+m-1,y:y+n/2-1)；
B4=BW (x+m/2-1:x+m-1:y+n/2-1:y+n-1)；
B5=BW (x+m/4-1:x+3*m/4-1,y+n/4-1:y+3*n/4-1)；
The minimum region of white point in B1, B2, B3, B4, B5 is calculated, B is denoted as, if B is B1, x, y is constant；If B is B2,
X is constant, y=n/2+y-1；If B is B3, y is constant, x=x+m/2-1；If B is B4, x=x+m/2-1, y=y+n/2-1；
If B is B5, x=x+m/3-1, y=y+n/3-1；Meanwhile, m, n is changed into B size.As long as meeting h<M and w<N, be carried out as
Lower step；
If step 2.4. 2*h<M, 2*w>N, the original image after being handled through step one is pressed
Vertical direction is divided into B1, B2, the parts of B3 tri-, wherein,
B1=BW (x:x+m/2-1,y:y+n-1)；
B2=BW (x+m/2-1:x+m-1,y:y+n-1)；
B3=BW (x+m/4-1:x+3*m/4-1,y:y+n-1)；
The minimum region B of white point in B1, B2, B3 is calculated, if B is B1, x, y is constant；If B is B2, y is constant, x=x+
m/2-1；If B is B3, y is constant, x=x+m/3-1；Meanwhile, m, n is changed into B size.As long as meeting h<M and w<N, is carried out
Following steps；
If step 2.5. 2*h>M, 2*w>N, then B is exactly the indivisible split window image finally navigated to.
Preferably, step 3 is specially：To the minimum indivisible split window image obtained in step 2, meter
The number of white point in the image for the middle that size in the image is [h w] (sizes of embedded images) is calculated, if the number is
0, then the window is just as the rectangular window being eventually found；If not being 0, then it is [h w] to calculate size in the image again
In upper, it is lower in, it is left, right in, the upper left corner, the upper right corner, the lower left corner, the number of white point in the subwindow of 8, the lower right corner, if
One white point number is 0, then just regard the window as the rectangular window being eventually found；If not being 0, then just use one
Individual size for [h w] rectangular area as sliding window, the minimum indivisible image obtained in step 2 is slided
It is dynamic, some regional location is obtained, the region white point number is 0 or minimum, regard the region as the suitable back of the body being eventually found
Scene area；Wherein, during middle is first calculated in the step and be upper, it is lower in, it is left, right in, the upper left corner, the upper right corner, the lower left corner,
The number of white point in the region of 9 subwindows in the lower right corner, is conducive to finding the suitable region in position in background, and can improve effect
Rate.
Brief description of the drawings
According to below by example and description to the specific embodiment of the invention referring to the drawings, will become apparent from out the present invention its
His feature and advantage, in the accompanying drawings：
Fig. 1 is the flow for being used to position the key step of the method in one piece of suitable rectangular background region in image of the present invention
Figure.
Fig. 2 is window segmentation figure algorithm flow chart；
Fig. 3 is the schematic diagram of image level partitioning scheme；
Fig. 4 is the schematic diagram of image vertical segmentation mode；
Fig. 5 is the schematic diagram of image level vertical segmentation mode；
Fig. 6 a are the schematic diagram of the white point number in the middle region for calculating subwindow；
Fig. 6 b are the schematic diagram of the white point number in the upper middle region for calculating subwindow；
Fig. 6 c are the schematic diagram of the white point number in the lower middle region for calculating subwindow；
Fig. 6 d are the schematic diagram of the white point number in the left region for calculating subwindow；
The schematic diagram of the white point number in region in the right sides of Fig. 6 e to calculate subwindow；
The schematic diagram of the white point number in region between the upper left corners of Fig. 6 f to calculate subwindow；
Fig. 6 g are the schematic diagram of the white point number for the upper right comer region for calculating subwindow；
Fig. 6 h are the schematic diagram of the white point number for the lower left corner region for calculating subwindow；
Fig. 6 i are the schematic diagram of the white point number for the lower right field for calculating subwindow；
Fig. 7 is the flow chart of sliding window algorithm；
Fig. 8 a are the original images of the Application Example of the present invention；
Fig. 8 b are the gray level images of the Application Example of the present invention；
Fig. 8 c are the bianry images of the Application Example of the present invention；
Fig. 8 d be the present invention Application Example expansion process after image；
Fig. 8 e are the schematic diagrames for finding minimum subwindow using sliding window as [200,200]；
Fig. 8 f are the schematic diagrames that region is found using sliding window as [200,200]；
Fig. 8 g are the schematic diagrames for finding minimum subwindow using sliding window as [220,220]；
Fig. 8 h are the schematic diagrames that region is found using sliding window as [220,220]；
Fig. 8 i are the schematic diagrames for finding minimum subwindow using sliding window as [250,250]；
Fig. 8 j are the schematic diagrames that region is found using sliding window as [250,250]；
Fig. 8 k are to show the region for finding the embedded original image of new image in 2 D code (size is respectively [200,200])
In final effect figure；
Fig. 8 l are showing the area for finding the embedded original image of new image in 2 D code (size is respectively [220,220])
Final effect figure in domain；
Fig. 8 m are showing the area for finding the embedded original image of new image in 2 D code (size is respectively [250,250])
Final effect figure in domain.
Embodiment
In order that the objects, technical solutions and advantages of the present invention become apparent from, below in conjunction with accompanying drawing with regard to embodiment, to this
Invention is described in further detail.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and
It is not used in the restriction present invention.
Fig. 1 is the flow for being used to position the key step of the method in one piece of suitable rectangular background region in image of the present invention
Figure.As illustrated, first reading in digital picture, and digital picture is converted into grayscale image, reuse edge detector by its
Bianry image is converted into, expansion process then is carried out to the image；Then one piece in the method split using image, positioning image
Fit into embedded new images and size is more than or equal to the position for the suitable rectangular area for wanting embedded images size；Finally exist
The method searched in the rectangular background region found above using sliding window, finds suitable new image insertion region.
Fig. 2 provides image window partitioning algorithm.The algorithm first step is to calculate original image BW (being denoted as B) size for [m
N], top left co-ordinate is (x, y), and embedded images size is [h w]；Second step, compares the length and width of original image and embedded images
(there is a primary condition here, that is, the size of the split window finally obtained have to be larger than the size equal to embedded images, therefore
All judgements to be carried out below should be based on this condition first), that is, judge h<m&&w<Whether n sets up, if h<m&&w<N is not
Set up, algorithm terminates, and the smallest partition window found is exactly B, conversely, carrying out the 3rd step, the 3rd step mainly includes following several
Judge：(1) if 2h>M, 2w<=n is set up, and with regard to carrying out the horizontal segmentation shown in Fig. 3, is divided into three parts, and every is partly the 1/2 of B
(length is still h, width for w 1/2), by calculating, obtain that minimum part of white point number in three parts, be denoted as B, lay equal stress on
The new size for calculating B is [m n] and its top left co-ordinate, then returns again to second step and continues to judge.(2) if 2h<=m, 2w>n,
The vertical segmentation shown in Fig. 4 is then carried out, is divided into three parts, 1/2 (length is the 1/2 of h, and width is constant) for being B per part passes through
Calculate, obtain that minimum part of white point number in three parts, be denoted as B, and it is [m n] and its upper left to recalculate B size
Angular coordinate, then returns again to second step and continues to judge.(3) if 2h<=m, 2w<=n, then carry out the vertical water shown in Fig. 5 and divide equally
Cut, point quinquepartite, each section is 1/4 (length and width are the 1/2 of artwork) of original image, by calculating, obtain five
That minimum part of white point number in point, is denoted as B, and it is [m n] and its top left co-ordinate, Ran Houzai to recalculate B size
Second step is returned to continue to judge.(4) if 2h>m,2w>N, then segmentation terminates, and B is exactly minimum split window.Three steps more than
Image partition method, obtain being embedded in the minimum subwindow of new image.
Fig. 6 a-Fig. 6 i are carried out before sliding window method, to the minimum subwindow B obtained in Fig. 2, first calculate the sub- window
Mouthful middle and it is upper in, it is lower in, it is left, right in, the upper left corner, the upper right corner, the lower left corner, (size is in the region of 9, the lower right corner
[h w]) white point number, if having in this 9 regions white point number be 0 region, just using the region as final region,
Without carrying out sliding window algorithm again, this is conducive to finding the appropriate region in position in background, and substantially increases efficiency.
Fig. 7 is sliding window algorithm, and (size is [m n] to the minimum subwindow obtained to dividing method；Top left co-ordinate is
(x, y)) for the sliding window of [h w] by left-to-right enter line slip from top to bottom with a size, find white point for 0 or white point most
Few sliding window region is used as final region.
Fig. 8 a to Fig. 8 m are the Application Examples of the present invention.
As shown in Fig. 8 a to Fig. 8 d, first read in digital picture (size [m n] be [950 775], coordinate (x, y) for (1,
1) digital picture then), is converted into grayscale image, edge detector is reused and converts thereof into bianry image, it is then right
The image carries out expansion process.
As shown in Fig. 8 e, Fig. 8 f, minimum subwindow first is found using image segmentation algorithm, is then calculated in the subwindow
Middle and it is upper in, it is lower in, it is left, right in, the upper left corner, the upper right corner, the lower left corner, (size [h w] in the region of 9, the lower right corner
For [200 200]) number of white point, it is found that white point number is 0 in the left intermediate region of the subwindow, therefore just make the region
For final area, now program just terminates.
As shown in Fig. 8 g, Fig. 8 h, minimum subwindow first is found using image segmentation algorithm, is then calculated in the subwindow
Middle and it is upper in, it is lower in, it is left, right in, the upper left corner, the upper right corner, the lower left corner, (size [h w] in the region of 9, the lower right corner
For [220 220]) number of white point, it is found that white point number is not 0 in 9 regions of the subwindow, therefore just in the sub- window
Enter line slip using sliding window (size [h w] is [220 220]) algorithm on mouth, have found white point number in the subwindow is
0 region, just using the region as final area, now program just terminates.
As shown in Fig. 8 i, Fig. 8 j, minimum subwindow first is found using image segmentation algorithm, is then calculated in the subwindow
Middle and it is upper in, it is lower in, it is left, right in, the upper left corner, the upper right corner, the lower left corner, (size [h w] in the region of 9, the lower right corner
For [250 250]) number of white point, it is found that white point number is not 0 in 9 regions of the subwindow, therefore just in the sub- window
Enter line slip using sliding window (size [h w] is [250 250]) algorithm on mouth, slip is finished, and finds do not have in the subwindow
There is the region that white point number is 0, therefore just using the minimum region of white point number as final area, now program just terminates.
As shown in Fig. 8 k, Fig. 8 l, Fig. 8 m, it is shown that by new image in 2 D code, (size is respectively [200 200], [220
220], [250 250]) it is embedded in the final effect figure in the region that original image is found.
The present invention also have to the similar application substantially of above-mentioned flow, no longer illustrate one by one herein.It is an advantage of the invention that logical
Position can preferably be obtained positioned at the minimum background area of middle white point number by crossing image segmentation algorithm, then be obtained by calculating
The middle of split window and it is upper in, it is lower in, it is left, right in, the upper left corner, the upper right corner, the lower left corner, in the region of 9, the lower right corner
The number of (size be [h w]) white point, if there is the region that white point number is 0 in this 9 regions, just using the region as final
Region, without carrying out sliding window algorithm again, this is conducive to finding the appropriate region in position in background, and substantially increases
Efficiency.Finally suitable rectangular background region can be obtained using sliding window algorithm.
The present invention can also have other various embodiments, in the case of without departing substantially from spirit of the invention and its essence, be familiar with
Those skilled in the art can make various corresponding changes and deformation according to the present invention, but these corresponding changes and deformation are all
The present invention should be belonged to and bear scope of the claims.
Claims (4)
1. suitable rectangular background region method in a kind of Quick positioning map picture, it is characterised in that comprise the following steps：
Step 1: reading in digital picture, and digital picture is converted into bianry image；
Step 2: the method split using image, is more than or equal under the precondition of embedded images size in split window, to figure
As the segmentation circulated, take in split window comprising the minimum region of white point as new split window continuation segmentation,
Until white point number that obtained split window is included is minimum and can not be further continued for segmentation, the split window is now positioned；
Step 3：The rectangular window of embedded images size is used in the split window navigated in step 2 as sliding window
Sliding window search is carried out, until the white point number of a certain position is 0 just using the region as final region, or does not have white point
Number is 0 region, just using the minimum region of white point number as last region, so just have found suitable new image insertion area
Domain.
2. suitable rectangular background region method in Quick positioning map picture as claimed in claim 1, it is characterised in that step one has
Body includes：
Step 1.1, digital picture is converted into grayscale image；
Step 1.2, binary conversion treatment, rim detection is carried out to grayscale image, such as edge is carried out using Sobel edge detectors
Detection, bianry image BW is converted into by grayscale image I, and it is the value " 1 " corresponding to white to detect its value of the position at edge,
Its value of other positions is the value " 0 " corresponding to black.
3. suitable rectangular background region method in Quick positioning map picture as claimed in claim 1, it is characterised in that step 2 bag
Include, first have to carry out the bianry image obtained in image step one the dilation operation in morphology so that target increases in image
Greatly, cavity tiny in blank map picture, and the border of object is smoothly connected, expand outwardly border；Secondly, it is exactly using image
The method of segmentation positions one piece of suitable rectangular background region in image, and it comprises the following steps that 2.1-2.5：
Step 2.1. first determine whether initial carrier image BW (being denoted as B) size [m n] (wherein BW top left co-ordinate for (x,
) and the size [h w] of image to be embedded in, y) as long as meeting condition h<M and w<N, is carried out following steps；
If step 2.2. 2*h>M, 2*w<=n, by the original image after being handled through step one by divided in horizontal direction be B1,
The part of B2, B3 tri-, wherein,
B1=BW (x:x+m-1,y:y+n/2-1)；
B2=BW (x:x+m-1,y+n/2-1:y+n-1)；
B3=BW (x:x+m-1,y+n/4-1:y+3*n/4-1)；
The minimum region of white point in B1, B2, B3 is calculated, B is denoted as, if B is B1, x, y is constant；If B is B2, x is constant, y=
n/2+y-1；If B is B3, x is constant, y=n/3+y-1；Meanwhile, m, n is changed into B size, as long as meeting h<M and w<N, just holds
Row following steps；
If step 2.3. 2*h<M, 2*w<=n, B1, B2, B3, B4, B5 are divided into by the original image after being handled through step one
Five parts, wherein,
B1=BW (x:x+m/2-1,y:y+n/2-1)；
B2=BW (x:x+m/2-1,y+n/2-1:y+n-1)；
B3=BW (x+m/2-1:x+m-1,y:y+n/2-1)；
B4=BW (x+m/2-1:x+m-1:y+n/2-1:y+n-1)；
B5=BW (x+m/4-1:x+3*m/4-1,y+n/4-1:y+3*n/4-1)；
The minimum region of white point in B1, B2, B3, B4, B5 is calculated, B is denoted as, if B is B1, x, y is constant；If B is B2, x is not
Become, y=n/2+y-1；If B is B3, y is constant, x=x+m/2-1；If B is B4, x=x+m/2-1, y=y+n/2-1；If B
B5, then x=x+m/3-1, y=y+n/3-1；Meanwhile, m, n is changed into B size.As long as meeting h<M and w<N, is carried out as follows
Step；
If step 2.4. 2*h<M, 2*w>N, B1, B2 are divided into by the original image after being handled through step one in a vertical direction,
The parts of B3 tri-, wherein,
B1=BW (x:x+m/2-1,y:y+n-1)；
B2=BW (x+m/2-1:x+m-1,y:y+n-1)；
B3=BW (x+m/4-1:x+3*m/4-1,y:y+n-1)；
The minimum region B of white point in B1, B2, B3 is calculated, if B is B1, x, y is constant；If B is B2, y is constant, x=x+m/2-
1；If B is B3, y is constant, x=x+m/3-1；Meanwhile, m, n is changed into B size, as long as meeting h<M and w<N, is carried out as follows
Step；
If step 2.5. 2*h>M, 2*w>N, then B is exactly the indivisible split window image finally navigated to.
4. suitable rectangular background region method in Quick positioning map picture as claimed in claim 1, it is characterised in that step 3 has
Body is：To the minimum indivisible split window image obtained in step 2, calculate in the image size be [h w] just
The number of white point in middle image, if the number is 0, the window is just as the rectangular window being eventually found；If no
For 0, then calculate again size in the image for [h w] it is upper in, it is lower in, it is left, right in, the upper left corner, the upper right corner, the lower left corner,
The number of white point in the subwindow of 8, the lower right corner, if a white point number is 0, then just using the window as finally looking for
The rectangular window arrived；If not being 0, then the rectangular area that a size is [h w] is just used as sliding window, to step
The minimum indivisible image obtained in rapid two enters line slip, obtains some regional location, the region white point number be 0 or
At least, it regard the region as the suitable background area being eventually found.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710346750.8A CN107220979B (en) | 2017-05-17 | 2017-05-17 | Method for quickly positioning appropriate rectangular background area in image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710346750.8A CN107220979B (en) | 2017-05-17 | 2017-05-17 | Method for quickly positioning appropriate rectangular background area in image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107220979A true CN107220979A (en) | 2017-09-29 |
| CN107220979B CN107220979B (en) | 2020-09-25 |
Family
ID=59944951
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710346750.8A Active CN107220979B (en) | 2017-05-17 | 2017-05-17 | Method for quickly positioning appropriate rectangular background area in image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107220979B (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2001352443A (en) * | 2000-06-07 | 2001-12-21 | Sharp Corp | Image-forming device and method, and storage medium used for the same |
| CN101453575A (en) * | 2007-12-05 | 2009-06-10 | 中国科学院计算技术研究所 | Video subtitle information extracting method |
| US7604172B2 (en) * | 2004-10-27 | 2009-10-20 | Denso Corporation | Image signal output device and a method of generating a coded image signal |
| CN102005026A (en) * | 2010-11-22 | 2011-04-06 | 天津职业技术师范大学 | Method for realizing robust digital watermarking by applying two-dimensional code |
| JP5071853B2 (en) * | 2007-12-21 | 2012-11-14 | 国立大学法人 鹿児島大学 | Image processing apparatus, image processing method, program, and recording medium |
| CN103106631A (en) * | 2011-11-11 | 2013-05-15 | 阿里巴巴集团控股有限公司 | Generation method and generation device of image verification code |
| KR20130129592A (en) * | 2012-05-21 | 2013-11-29 | 주식회사 유스타 | System for taking photo sticker and the method thereof |
| CN104574271A (en) * | 2015-01-20 | 2015-04-29 | 复旦大学 | Method for embedding advertisement icon into digital image |
| CN106447590A (en) * | 2016-10-11 | 2017-02-22 | 华南理工大学 | Watermark adding and extraction method in digital image |
-
2017
- 2017-05-17 CN CN201710346750.8A patent/CN107220979B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2001352443A (en) * | 2000-06-07 | 2001-12-21 | Sharp Corp | Image-forming device and method, and storage medium used for the same |
| US7604172B2 (en) * | 2004-10-27 | 2009-10-20 | Denso Corporation | Image signal output device and a method of generating a coded image signal |
| CN101453575A (en) * | 2007-12-05 | 2009-06-10 | 中国科学院计算技术研究所 | Video subtitle information extracting method |
| JP5071853B2 (en) * | 2007-12-21 | 2012-11-14 | 国立大学法人 鹿児島大学 | Image processing apparatus, image processing method, program, and recording medium |
| CN102005026A (en) * | 2010-11-22 | 2011-04-06 | 天津职业技术师范大学 | Method for realizing robust digital watermarking by applying two-dimensional code |
| CN103106631A (en) * | 2011-11-11 | 2013-05-15 | 阿里巴巴集团控股有限公司 | Generation method and generation device of image verification code |
| KR20130129592A (en) * | 2012-05-21 | 2013-11-29 | 주식회사 유스타 | System for taking photo sticker and the method thereof |
| CN104574271A (en) * | 2015-01-20 | 2015-04-29 | 复旦大学 | Method for embedding advertisement icon into digital image |
| CN106447590A (en) * | 2016-10-11 | 2017-02-22 | 华南理工大学 | Watermark adding and extraction method in digital image |
Non-Patent Citations (3)
| Title |
|---|
| CHRISTOPH H. LAMPERT ET AL: "Efficient Subwindow Search: A Branch and Bound Framework for Object Localization", 《IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE》 * |
| THALES SEHN KORTING ET AL: "A resegmentation approach for detecting rectangular objects in high-resolution imagery", 《IEEE GEOSCIENCE AND REMOTE SENSING LETTERS》 * |
| 韦仁忠: "数字图像中物体的定位与检测研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107220979B (en) | 2020-09-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11595737B2 (en) | Method for embedding advertisement in video and computer device | |
| CN114972329B (en) | Image enhancement method and system of surface defect detector based on image processing | |
| EP3620981A1 (en) | Object detection method, device, apparatus and computer-readable storage medium | |
| CN108121991B (en) | Deep learning ship target detection method based on edge candidate region extraction | |
| US8265350B2 (en) | Method and apparatus for detecting and processing specific pattern from image | |
| JP2018200685A (en) | Forming of data set for fully supervised learning | |
| CN106529573A (en) | Real-time object detection method based on combination of three-dimensional point cloud segmentation and local feature matching | |
| CN108009529B (en) | Forest fire smoke video target detection method based on characteristic root and hydrodynamics | |
| CN110334709B (en) | License plate detection method based on end-to-end multi-task deep learning | |
| CN112132164B (en) | Target detection method, system, computer device and storage medium | |
| CN112800955A (en) | Remote sensing image rotating target detection method and system based on weighted bidirectional feature pyramid | |
| CN111177811A (en) | Automatic fire point location layout method applied to cloud platform | |
| US7542589B2 (en) | Road position detection | |
| Yuan et al. | Superpixels with content-adaptive criteria | |
| CN114386504A (en) | Engineering drawing character recognition method | |
| JP6713422B2 (en) | Learning device, event detection device, learning method, event detection method, program | |
| CN106469432A (en) | Object extraction method and object extraction equipment | |
| CN112084103B (en) | Interface test method, device, equipment and medium | |
| CN111798481B (en) | Image sequence segmentation method and device | |
| CN107220979A (en) | A kind of method in suitable rectangular background region in Quick positioning map picture | |
| CN116091784A (en) | Target tracking method, device and storage medium | |
| JP2002117409A (en) | Image processing method and device thereof | |
| JP3638845B2 (en) | Image processing apparatus and method | |
| CN114549833A (en) | Instance partitioning method and device, electronic equipment and storage medium | |
| JP2016081472A (en) | Image processing device, and image processing method and program |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |