CN107230194A - A kind of smooth filtering method based on object point set - Google Patents
A kind of smooth filtering method based on object point set Download PDFInfo
- Publication number
- CN107230194A CN107230194A CN201710508010.XA CN201710508010A CN107230194A CN 107230194 A CN107230194 A CN 107230194A CN 201710508010 A CN201710508010 A CN 201710508010A CN 107230194 A CN107230194 A CN 107230194A
- Authority
- CN
- China
- Prior art keywords
- point set
- border
- object point
- edge
- pixel
- 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
- 238000000034 method Methods 0.000 title claims abstract description 33
- 238000001914 filtration Methods 0.000 title claims abstract description 20
- 238000005260 corrosion Methods 0.000 claims abstract description 8
- 230000007797 corrosion Effects 0.000 claims abstract description 8
- 230000003628 erosive effect Effects 0.000 claims 1
- 230000000903 blocking effect Effects 0.000 abstract description 7
- 230000000694 effects Effects 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a kind of smooth filtering method based on object point set, including step：Obtain border 1；Obtain border 2；The disposal of gentle filter is carried out to the region point set between border 1 and border 2, the border 1 is edge point set of the object point set after 10 times corrode, and the border 2 is edge point set of the object point set after 10 times expand.The method of the invention obtains border 1 and border 2 using the method for corrosion and expansion, and obtain region point set using border 1 and border 2, the region point set reflects the profile information of whole object point set, the disposal of gentle filter is carried out to the region point set, can eliminate between the object on picture and object, the blocking effect between object and background, improve the total quality of picture.
Description
Technical field
The present invention relates to technical field of image processing, more particularly to a kind of smooth filtering method based on object point set.
Background technology
Newest object extraction technology Mask R-CNN, the technology carries out semantic segmentation to image, and it is of the same race that segmentation is obtained
The collection of the pixel of classification is collectively referred to as object-point collection.One pictures include several object point sets, if entered to some object point set
During row processing, such as the adjustment of contrast is carried out to some object point set, after the adjustment, object point set and object point set it
Between, between object point set and background then easily there is blocking effect, these blocking effects influence whether the quality of whole picture.Therefore,
If eliminated between object point set and object point set, the blocking effect between object point set and background is into being badly in need of what is solved in field
Problem.
The content of the invention
In order to solve the above problems, the invention provides a kind of smooth filtering method based on object point set, this method can
Efficiently solve between object point set and object point set, the blocking effect between object point set and background.
The present invention solve its technical problem solution be：A kind of smooth filtering method based on object point set, including
Step：Obtain border 1；Obtain border 2；The disposal of gentle filter, the side are carried out to the region point set between border 1 and border 2
Boundary 1 is edge point set of the object point set after 10 times corrode, and the border 2 is edge of the object point set after 10 times expand
Point set.
Further, the method for 10 corrosion includes step：
A1) subtract current edge point set with current object point set and obtain new object point set；
A2) new edge point set is acquired with the new object point set；
A3) repeat step a1)-a2) 10 times.
Further, the acquisition methods of the edge point set include：Centered on any one pixel that object-point is concentrated,
The point is referred to as central pixel point, judge the central pixel point left and right, upper and lower pixel and its whether be same category,
Marginal point is obtained according to judged result, is edge point set by the edge point set.
Further, the method for 10 expansions includes step：
A11) center of structural elements is put on any point of current edge point set, and structural elements center along institute
Edge point set movement is stated, the pixel for belonging to structural elements but being not belonging to object point set is found, and be new by the pixel point set
Edge point set；
A12) by current object point set and the edge point set and as new object point set；
A13) repeat step a11)-a12) 10 times.
Further, smothing filtering is carried out to the region point set between border 1 and border 2 using 3 × 3 Filtering Template.
The beneficial effects of the invention are as follows：This method obtains border 1 and border 2, and profit using the method for corrosion and expansion
Region point set is obtained with border 1 and border 2, the region point set reflects the profile information of whole object point set, to the region point
Collection carries out the disposal of gentle filter, can eliminate between the object on picture and object, the blocking effect between object and background, improves
The total quality of picture.
Brief description of the drawings
Technical scheme in order to illustrate the embodiments of the present invention more clearly, makes required in being described below to embodiment
Accompanying drawing is briefly described.Obviously, described accompanying drawing is a part of embodiment of the present invention, rather than is all implemented
Example, those skilled in the art on the premise of not paying creative work, can also obtain other designs according to these accompanying drawings
Scheme and accompanying drawing.
Fig. 1 is the flow chart of the smooth filtering method based on object point set；
Fig. 2 is the flow chart of 10 corrosion；
Fig. 3 is the flow chart of 10 expansions.
Embodiment
The technique effect of the design of the present invention, concrete structure and generation is carried out below with reference to embodiment and accompanying drawing clear
Chu, it is fully described by, to be completely understood by the purpose of the present invention, feature and effect.Obviously, described embodiment is this hair
Bright a part of embodiment, rather than whole embodiments, based on embodiments of the invention, those skilled in the art is not paying
The other embodiment obtained on the premise of creative work, belongs to the scope of protection of the invention.In addition, be previously mentioned in text
All connection/annexations, not singly refer to component and directly connect, and refer to be added deduct by adding according to specific implementation situation
Few couple auxiliary, to constitute more excellent draw bail.Each technical characteristic in the invention, in not conflicting conflict
Under the premise of can be with combination of interactions.
Embodiment 1, with reference to Fig. 1, the picture after Mask R-CNN technical finesses is handled as follows：One kind is based on
The smooth filtering method of object point set,
S00：Finding needs object to be processed and obtains its current object point set；
S01：Object point set obtains border 1 by 10 corrosion；
S02：Object point set obtains border 2 by 10 expansions；
S03：The disposal of gentle filter is carried out to the region point set between border 1 and border 2.
It should be noted that step S01 and S02 do not have sequencing and limited, this implementation is using step S01 as first.With reference to
Fig. 2, wherein, the method for 10 corrosion comprises the following steps：
S1：Current edge point set, which is subtracted, with current object point set obtains new object point set；
S2：New edge point set is acquired with the new object point set；
S3：It is repeated 10 times step S1-S2.
The acquisition methods of wherein edge point set include：It is assumed that object point set is ∑ (xi,yi), in label classification figure,
With ∑ (xi,yi) each pixel (xi,yi) centered on, pixel (xi-1,yi),(xi+1,yi) it is respectively pixel (xi,
yi) 2 points of left and right；(xi,yi+1),(xi,yi- 1) it is respectively point (xi,yi) 2 points up and down.Pixel (the xi,yiIf)
Meet below equation：
|C(xi-1,yi)-C(xi,yi)|+|C(xi,yi)-C(xi+1,yi)|≥1
Or | C (xi,yi+1)-C(xi,yi)|+|C(xi,yi)-C(xi,yi-1) | >=1 judges the pixel (xi,yi) be
Marginal point, the edge point set to marginal point is concentrated.The formula is construed to：Pixel (xi,yi) from left to right or from upper
Pixel label classification value changes under are more than or equal to 1.Then judge pixel (xi,yi) it is marginal point, wherein function C (x, y)
Refer to the label class label of pixel (x, y).
With reference to Fig. 3, the method for 10 expansions includes step：
S11：On any point that the center of structural elements is put into current edge point set, and structural elements center along institute
State edge point set to move clockwise, find the pixel for belonging to structural elements but being not belonging to object point set, and by the pixel point set
It is combined into new edge point set；It is formulated as：Wherein A is current
Edge point set, B is structural elements, and z is exactly the edge point set that expansion is once obtained.
S12：By current object point set and the edge point set and as new object point set；
S13：It is repeated 10 times step S11-S12.
Determine behind border 1 and border 2, just obtained the region point set between border 1 and border 2, the region point collection is met
Condition below：
If pixel (x, y) represents that region point set takes up an official post meaning a bit, pixel (x1, y1) represents any point on border 1,
Pixel (x2, y2) represents any point on border 2, then when meeting x1=x=x2, y1<y<Y2 or y2<y<y2.
The disposal of gentle filter is carried out to the region point set between border 1 and border 2, the present embodiment uses 3 × 3 filtering mould
Plate, the template is：
Smothing filtering process is as follows：
Assuming that being the gray scale value matrix of piece image below：
| 0 | 0 | 143 | 0 | 0 |
| 0 | 93 | 157 | 89 | 0 |
| 189 | 156 | 162 | 0 | 56 |
| 0 | 89 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 | 0 |
After gray value f (3,1)=189, f (3,2)=156, f (3,3)=162 is by template convolution
Region point set between border 1 and border 2 is converted into gray scale value matrix, according to the method described above to each pixel
Point carry out the disposal of gentle filter successively so as to obtain it is final it is smooth after region point set.
The method of the invention obtains border 1 and border 2 using the method for corrosion and expansion, and utilizes the He of border 1
Border 2 obtains region point set, and the region point set reflects the profile information of whole object point set, the region point set is carried out smooth
Filtering process, can be eliminated between the object on picture and object, the blocking effect between object and background, improve the entirety of picture
Quality.
The better embodiment to the present invention is illustrated above, but the invention is not limited to the implementation
Example, those skilled in the art can also make a variety of equivalent modifications or replace on the premise of without prejudice to spirit of the invention
Change, these equivalent modifications or replacement are all contained in the application claim limited range.
Claims (5)
1. a kind of smooth filtering method based on object point set, it is characterised in that：Including step：Obtain border 1；Obtain border 2；
The disposal of gentle filter is carried out to the region point set between border 1 and border 2, the border 1 is object point set by 10 corrosion
Edge point set afterwards, the border 2 is edge point set of the object point set after 10 times expand.
2. a kind of smooth filtering method based on object point set according to claim 1, it is characterised in that：10 corruption
The method of erosion includes step：
A1) subtract current edge point set with current object point set and obtain new object point set；
A2) new edge point set is acquired with the new object point set；
A3) repeat step a1)-a2) 10 times.
3. a kind of smooth filtering method based on object point set according to claim 2, it is characterised in that：The marginal point
The acquisition methods of collection include：Centered on any one pixel that object-point is concentrated, the point is referred to as central pixel point, judges institute
State central pixel point left and right, upper and lower pixel and its whether be same category, according to judged result obtain marginal point, will
The edge point set is edge point set.
4. a kind of smooth filtering method based on object point set according to claim 1, it is characterised in that：Described 10 times swollen
Swollen method includes step：
A11) center of structural elements is put on any point of current edge point set, and structural elements center along the side
Edge point set is moved, and finds the pixel for belonging to structural elements but being not belonging to object point set, and is new side by the pixel point set
Edge point set；
A12) by current object point set and the edge point set and as new object point set；
A13) repeat step a11)-a12) 10 times.
5. a kind of smooth filtering method based on object point set according to claim 1, it is characterised in that：Using 3 × 3
Filtering Template carries out smothing filtering to the region point set between border 1 and border 2.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710508010.XA CN107230194A (en) | 2017-06-28 | 2017-06-28 | A kind of smooth filtering method based on object point set |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710508010.XA CN107230194A (en) | 2017-06-28 | 2017-06-28 | A kind of smooth filtering method based on object point set |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107230194A true CN107230194A (en) | 2017-10-03 |
Family
ID=59935394
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710508010.XA Pending CN107230194A (en) | 2017-06-28 | 2017-06-28 | A kind of smooth filtering method based on object point set |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107230194A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111161275A (en) * | 2018-11-08 | 2020-05-15 | 腾讯科技（深圳）有限公司 | Method and device for segmenting target object in medical image and electronic equipment |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TW200742448A (en) * | 2006-04-26 | 2007-11-01 | Quanta Comp Inc | Image processing apparatus and method of the same |
| CN104835127A (en) * | 2015-05-19 | 2015-08-12 | 中国农业科学院农业信息研究所 | Adaptive smooth filtering method |
| CN105787911A (en) * | 2016-03-21 | 2016-07-20 | 中国林业科学研究院资源信息研究所 | Image erosion and expansion processing method based on topology fractal algorithm |
| CN105825482A (en) * | 2016-03-15 | 2016-08-03 | 四川用联信息技术有限公司 | Depth image restoration algorithm |
| CN106485673A (en) * | 2016-09-19 | 2017-03-08 | 电子科技大学 | A kind of filtering method to sea SAR image |
| CN106813569A (en) * | 2015-11-30 | 2017-06-09 | 中国科学院沈阳自动化研究所 | A kind of automobile tire 3-D positioning method based on line-structured light |
-
2017
- 2017-06-28 CN CN201710508010.XA patent/CN107230194A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TW200742448A (en) * | 2006-04-26 | 2007-11-01 | Quanta Comp Inc | Image processing apparatus and method of the same |
| CN104835127A (en) * | 2015-05-19 | 2015-08-12 | 中国农业科学院农业信息研究所 | Adaptive smooth filtering method |
| CN106813569A (en) * | 2015-11-30 | 2017-06-09 | 中国科学院沈阳自动化研究所 | A kind of automobile tire 3-D positioning method based on line-structured light |
| CN105825482A (en) * | 2016-03-15 | 2016-08-03 | 四川用联信息技术有限公司 | Depth image restoration algorithm |
| CN105787911A (en) * | 2016-03-21 | 2016-07-20 | 中国林业科学研究院资源信息研究所 | Image erosion and expansion processing method based on topology fractal algorithm |
| CN106485673A (en) * | 2016-09-19 | 2017-03-08 | 电子科技大学 | A kind of filtering method to sea SAR image |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111161275A (en) * | 2018-11-08 | 2020-05-15 | 腾讯科技（深圳）有限公司 | Method and device for segmenting target object in medical image and electronic equipment |
| CN111161275B (en) * | 2018-11-08 | 2022-12-23 | 腾讯科技（深圳）有限公司 | Method and device for segmenting target object in medical image and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2020107866A1 (en) | Text region obtaining method and apparatus, storage medium and terminal device | |
| CN106960208B (en) | Method and system for automatically segmenting and identifying instrument liquid crystal number | |
| CN112528863A (en) | Identification method and device of table structure, electronic equipment and storage medium | |
| US10062170B2 (en) | Apparatus and method for extracting object | |
| CN103942797B (en) | Scene image text detection method and system based on histogram and super-pixels | |
| CN107228860A (en) | A kind of gear defect detection method based on image rotation cyclophysis | |
| CN107679437B (en) | Bar code image recognition method based on Zbar | |
| US8433104B2 (en) | Image processing method for background removal | |
| CN108346148B (en) | High-density flexible IC substrate oxidation area detection system and method | |
| KR101893679B1 (en) | Card number recognition method using deep learnig | |
| CN104239909A (en) | Method and device for recognizing images | |
| CN115457063A (en) | Method, device and equipment for extracting edge of circular hole of PCB (printed Circuit Board) and storage medium | |
| CN104715221A (en) | Decoding method and system for ultra-low-contrast two-dimension code | |
| CN111860176B (en) | Non-metal inclusion full-view-field quantitative statistical distribution characterization method | |
| CN110996096A (en) | Tone mapping image quality evaluation method based on structural similarity difference | |
| CN109741273A (en) | A kind of mobile phone photograph low-quality images automatically process and methods of marking | |
| CN112348831A (en) | Shale SEM image segmentation method based on machine learning | |
| CN107230194A (en) | A kind of smooth filtering method based on object point set | |
| CN110751156A (en) | Method, system, device and medium for table line bulk interference removal | |
| CN110610163B (en) | Table extraction method and system based on ellipse fitting in natural scene | |
| CN115719326A (en) | PCB defect detection method and device | |
| CN116030472A (en) | Text coordinate determining method and device | |
| CN112651936B (en) | Steel plate surface defect image segmentation method and system based on image local entropy | |
| CN107492073B (en) | Digital image restoration method based on weight guide filtering | |
| CN112651368B (en) | DNA ploidy analysis method for large-size cell microscopic image |
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 |
Application publication date: 20171003 |
|
| RJ01 | Rejection of invention patent application after publication |