CN102800066A - Local extremum and pixel value gradient based improved image enhancement method - Google Patents
Local extremum and pixel value gradient based improved image enhancement method Download PDFInfo
- Publication number
- CN102800066A CN102800066A CN2012102776766A CN201210277676A CN102800066A CN 102800066 A CN102800066 A CN 102800066A CN 2012102776766 A CN2012102776766 A CN 2012102776766A CN 201210277676 A CN201210277676 A CN 201210277676A CN 102800066 A CN102800066 A CN 102800066A
- Authority
- CN
- China
- Prior art keywords
- image
- local extremum
- pixel
- point
- 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.)
- Granted
Links
Images
Abstract
The invention discloses an image gradient and local extremum based image edge-preserving filter algorithm. According to the gradient value of pixel values of an image and the distribution information of local extremum points, the position of an edge in the image is judged. Because according to the algorithm disclosed by the invention, the position of an edge of an image is judged through the gradient of pixel values and the local extremum points of pixel values, the edge of the image can be determined more accurately. In the process of filtering noises and detail textures in the image, a filter reduces the influences on the edge of the image in the process of filtering according to the previous judgment, therefore, the filter can separate the contour of an object in the image from an original image.
Description
Technical field
The present invention relates to digital image processing field, be specifically related to a kind of improved image enchancing method based on Local Extremum distribution and pixel value gradient.
Background technology
The edge keeps the basic tool of wave filter as Flame Image Process, because range of application is current base image handling implement very extensively.Its objective is according to actual needs,, but keep important information (the for example edge of object) the less important information in the image (for example noise, grain details etc.) filtering.Certainly, different according to the practical problems of being faced, even same width of cloth image, its important information and less important INFORMATION IS NOT immobilize.For example, when people hope to remove noise in the image and when obtaining high-quality picture, this moment, noise was just as a kind of less important information and elimination; On the contrary, thus when people needed the suffered interference type of the pattern hypothetical system of noise in the analysis image, the noise in this moment image must keep as important information again.
As far back as the 1950's at initial stage of image processing techniques; Because basic status of its predecessor---image filter---and image processing techniques receive extensive attention in a plurality of applications, so detect and biomedical engineering etc. can be seen the figure of image filter in Aero-Space, military guidance, robot vision, police and judicial, medical science.In the image filtering technology in early days, research emphasis be how will quality be lower in a certain respect image (for example noise, edge fog etc.) through handling, improve this aspect quality and form output image.Based on this background, academia uses the Fourier transform in the frequency-region signal treatment technology, the definition image filtering.It extends to two dimensional image signal field with the signal Processing thought of one dimension.
In recent years, thus in the practical application to image processing techniques no longer as only requiring in early days image processing techniques to improve the single aspect of a certain width of cloth input picture quality or simply image is carried out multiple tracks and handle and improve the lower image quality index of several relevances; It requires image processing techniques that a kind of comprehensive method is provided more, with a plurality of indexs of raising image, or even some conflicting indexs.In fact, the edge keep with these two targets of filtering be conflicting to a certain extent: the purpose of wave filter itself is noise and the grain details in the removal of images, but this understands the edge in the blurred picture to a certain extent.This is because from traditional image treatment technology angle; Noise, grain details and image border three are the HFSs in the image, so use traditional method (for example Gaussian wave filter and Butterworth wave filter) can not obtain desirable effect merely.The wave filter that boundary filter proposes for the requirement of tackling this contradiction just.
The inventive method keeps wave filter possibly produce the defective of halation etc. to existing edge, proposes a kind of improved filtering method, and the object edge profile in the image is separated with article surface vein better, and treatment effect is preferably arranged.
Summary of the invention
The object of the present invention is to provide a kind of improved image border to keep filtering method based on local extremum and pixel value gradient; Can be used for image texture and strengthen (for example strengthening the perhaps stylization processing of image with the low image of contrast) fuzzy; Digital Image Processing such as the color range reconstruction of high-dynamics image, the whole visual effect of raising image.
The present invention; A kind of image border based on partial gradient and pixel gradient value keeps filtering method; Be on the basis of bilateral filtering method and local extremum filtering method; Through the mode that above-mentioned two kinds of different filtering methods are merged, propose a kind of local extremum distributed intelligence and pixel gradient and carry out the method that the image border keeps filtering according to neighborhood of pixel points.Being described below of new method:
Note input picture I (x, y) size is (b-a) * (d-c), at the regional Ω of two dimensional surface: on [a, b] * [c, the d], output image be O (x, y),
(1) with output image O (x, y] be initialized as zero;
(2) the subimage I of a k * k of definition in the input picture upper left corner
BAnd establish the step-length that its level and vertical direction move and be respectively h and v;
(3) antithetical phrase image I
BFiltering method according to based on local extremum and pixel gradient carries out Flame Image Process, requires all pixels of whole subgraph piece are handled, and its result is outputed to the pairing output image O of subimage central point, and (x is in position y);
(4) sub-piece is moved horizontally piece to move horizontally step-length h, if sub-piece does not exceed image boundary, repeating step (4), otherwise get into next step;
(5) with vertical moving step-length v vertical moving subgraph piece, if sub-piece does not exceed image boundary, repeating step (4), otherwise get into next step;
(6) after above step is accomplished, will obtain output image O (x, y).
Described filtering method based on local extremum and pixel gradient, its method is described below:
(1) the Local Extremum number of statistics original image pixels point
Whether any is the influence of extreme point to judging in the image certain in order to reduce noise in the image, and this method adopts the extreme point in the following mode check image.As point (x
0, y
0) pixel value at most than with it being k-1 pixel value of putting hour in the k * k neighborhood at center, algorithm is thought (x
0, y
0) be the local maximum point of image I.Similarly, algorithm is judged the local minizing point of image in the same way.The aforementioned calculation mode is equivalent to think that the pixel value that texture the caused vibration of image has only k pixel at interval at most.Through the size of adjustment parameter k, final algorithm is removed the texture of various frequencies.
Through said method, the number I of statistical pixel point extreme point in k * k neighborhood
Extrma
(2) calculating is based on the filter template of local extremum and pixel gradient
The present invention is on original two-sided filter basis, through introducing the template of local extremum information to two-sided filter.Pixel domain weight in the template of two-sided filter is added the Local Extremum number I of corresponding pixel points
ExtrmaAs correction term.At this moment, filter template not only judges based on the difference of the pixel value between the image each point whether this point is positioned at the edge of object, judges according to the number of extreme point in the neighborhood of pixel points simultaneously, thereby avoids the texture erroneous judgement that Oscillation Amplitude is big to be the edge of object.
A kind of image border based on local extremum and pixel gradient that the present invention proposes keeps filtering method, has the characteristics of following two aspects:
(1) this method is on the basis of two-sided filter method and local extremum filtered method, through utilizing the template of Local Extremum distributed intelligence correction two-sided filter, proposes a kind of more accurate image border and keeps filtered method.
(2) method after the improvement is compared with the local extremum wave filter with two-sided filter, can produce effect preferably, and the complexity of algorithm is identical with two-sided filter.
Description of drawings
Fig. 1 is passenger's number system integral module block diagram of the present invention;
Embodiment
Keep filtered method to do detailed description below in conjunction with accompanying drawing to a kind of image border of the present invention based on local extremum and pixel gradient.
The present invention is improving one's methods of on two-sided filter method basis, proposing.Therefore, at first specifically describe the image border maintenance filtering method of traditional gauss low frequency filter method and two-sided filter.
Discussing for convenient, is example with the single channel image only here.Note input picture I size is (b-a) * (d-c), the Ω in the two dimensional surface zone: on [a, b] * [c, the d].To any 1 the s ∈ Ω in the image, defining the point set that its neighborhood point forms is Q
sThe pixel value of point s is I
s, and remember that this pixel value is I after the t time iteration
s tAccording to above-mentioned notation convention, the result of the inferior back of Gauss spatial domain wave filter iterative computation (t+1) gained can be expressed as:
Wherein function g (x) is the Gauss function, and k (s) is the normalized factor of coefficient, promptly has
Similarly, the result of the inferior back of two-sided filter iterative computation (t+1) gained can be expressed as:
Function g wherein
S(x) and g
R(x) all be the Gauss function, act on respectively on spatial domain and the pixel domain; K (s) is the coefficient normalized factor, promptly has
The present invention is through Local Extremum number I in the computed image neighborhood
Extrma, the weight of filtering template in the correction two-sided filter, particularly,
Wherein function T (s) is the normalized factor of coefficient, promptly has
The rudimentary algorithm process flow diagram please refer to Figure of description 1.
Above content is to combine concrete preferred implementation to the further explain that the present invention did, and can not assert that practical implementation of the present invention is confined to these explanations.For the those of ordinary skill of technical field under the present invention, under the prerequisite that does not break away from the present invention's design, can also make some simple deduction or replace, all should be regarded as belonging to protection scope of the present invention.
Claims (2)
1. improved image enchancing method based on local extremum and pixel value gradient; It is characterized in that: on the basis of bilateral filtering method and local extremum filtering method; Through the mode that above-mentioned two kinds of different filtering methods are merged; Propose a kind of local extremum distributed intelligence and pixel gradient and carry out the method that the image border keeps filtering, may further comprise the steps according to neighborhood of pixel points:
A. remember input picture I (x, y) size is (b-a) * (d-c), the Ω in two dimensional surface zone: on [a, b] * [c, the d], output image be O (x, y) following steps are adopted in the back;
B. (x y) is initialized as zero with output image O;
C. define the subimage I of a k * k in the input picture upper left corner
BAnd establish the step-length that its level and vertical direction move and be respectively h and v;
D. antithetical phrase image I
BFiltering method according to based on local extremum and pixel gradient carries out Flame Image Process, requires all pixels of whole subgraph piece are handled, and its result is outputed to the pairing output image O of subimage central point, and (x is in position y);
E. sub-piece is moved horizontally piece to move horizontally step-length h, if sub-piece does not exceed image boundary, repeating step e, otherwise get into next step;
F. with vertical moving step-length v vertical moving subgraph piece, do not exceed image boundary as if sub-piece, repeating step e, otherwise get into next step;
G. after above step is accomplished, will obtain output image O (x, y).
2. a kind of image border based on local extremum and pixel gradient value according to claim 1 keeps filtering method, and it is characterized in that: described filtering method step based on local extremum and pixel gradient is following:
A. add up the Local Extremum number of original image pixels point: as point (x
0, y
0) pixel value at most than with it being k-1 pixel value of putting hour in the k * k neighborhood at center, algorithm is thought (x
0, y
0) be the local maximum point of image I, similarly, algorithm is judged the local minizing point of image in the same way, the aforementioned calculation mode is equivalent to think that the pixel value that texture the caused vibration of image has only k pixel at interval at most.Through the size of adjustment parameter k, final algorithm is removed the texture of various frequencies, through said method, and the number I of statistical pixel point extreme point in k * k neighborhood
Extrma
B. calculate filter template: on original two-sided filter basis based on local extremum and pixel gradient; Through introducing the template of local extremum information, pixel domain weight in the template of two-sided filter is added the Local Extremum number I of corresponding pixel points to two-sided filter
ExtrmaAs correction term; At this moment; Filter template not only judges based on the difference of the pixel value between the image each point whether this point is positioned at the edge of object, judges according to the number of extreme point in the neighborhood of pixel points simultaneously, thereby avoids the texture erroneous judgement that Oscillation Amplitude is big to be the edge of object.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210277676.6A CN102800066B (en) | 2012-08-03 | 2012-08-03 | Local extremum and pixel value gradient based improved image enhancement method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210277676.6A CN102800066B (en) | 2012-08-03 | 2012-08-03 | Local extremum and pixel value gradient based improved image enhancement method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN102800066A true CN102800066A (en) | 2012-11-28 |
CN102800066B CN102800066B (en) | 2015-01-07 |
Family
ID=47199165
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201210277676.6A Expired - Fee Related CN102800066B (en) | 2012-08-03 | 2012-08-03 | Local extremum and pixel value gradient based improved image enhancement method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102800066B (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106203266A (en) * | 2016-06-28 | 2016-12-07 | 比亚迪股份有限公司 | The extracting method of image extreme point and device |
CN106408535A (en) * | 2016-09-18 | 2017-02-15 | 福州大学 | Image enhancement method based on sub-line driving gray-scale modulation display system |
CN114536346A (en) * | 2022-04-06 | 2022-05-27 | 西南交通大学 | Mechanical arm accurate path planning method based on man-machine cooperation and visual detection |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20030128888A1 (en) * | 2002-01-10 | 2003-07-10 | Sharp Laboratories Of America, Inc. | Nonlinear edge-enhancement filter |
CN101271516A (en) * | 2008-04-02 | 2008-09-24 | 范九伦 | Direction filtering reinforcement method of fingerprint image |
-
2012
- 2012-08-03 CN CN201210277676.6A patent/CN102800066B/en not_active Expired - Fee Related
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20030128888A1 (en) * | 2002-01-10 | 2003-07-10 | Sharp Laboratories Of America, Inc. | Nonlinear edge-enhancement filter |
CN101271516A (en) * | 2008-04-02 | 2008-09-24 | 范九伦 | Direction filtering reinforcement method of fingerprint image |
Non-Patent Citations (2)
Title |
---|
ANUSTUP CHOUDHURY等: "Perceptually Motivated Automatic Sharpness Enhancement Using Hierarchy of Non-Local Means", 《2011 IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION WORKSHOPS》, 13 November 2011 (2011-11-13), pages 730 - 737, XP032095322, DOI: doi:10.1109/ICCVW.2011.6130325 * |
高龙等: "基于局部极值的保边缘图像分解算法", 《计算机工程》, vol. 37, no. 18, 20 September 2011 (2011-09-20), pages 217 - 218 * |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106203266A (en) * | 2016-06-28 | 2016-12-07 | 比亚迪股份有限公司 | The extracting method of image extreme point and device |
CN106203266B (en) * | 2016-06-28 | 2017-07-21 | 比亚迪股份有限公司 | The extracting method and device of image extreme point |
CN106408535A (en) * | 2016-09-18 | 2017-02-15 | 福州大学 | Image enhancement method based on sub-line driving gray-scale modulation display system |
CN106408535B (en) * | 2016-09-18 | 2019-04-05 | 福州大学 | A kind of image enchancing method based on sub-line driving gray modulation display system |
CN114536346A (en) * | 2022-04-06 | 2022-05-27 | 西南交通大学 | Mechanical arm accurate path planning method based on man-machine cooperation and visual detection |
CN114536346B (en) * | 2022-04-06 | 2023-04-07 | 西南交通大学 | Mechanical arm accurate path planning method based on man-machine cooperation and visual detection |
Also Published As
Publication number | Publication date |
---|---|
CN102800066B (en) | 2015-01-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103116875B (en) | Self-adaptation bilateral filtering image de-noising method | |
CN107657606B (en) | Method and device for detecting brightness defect of display device | |
CN105608679B (en) | A kind of image de-noising method of fusion structure tensor and the full variation of non-local | |
CN105976382B (en) | A kind of TFT-LCD Mura defects detection methods based on defect area anticipation and level set | |
CN110796616B (en) | Turbulence degradation image recovery method based on norm constraint and self-adaptive weighted gradient | |
CN105894474A (en) | Non-linear image enhancement method, and edge detection method using the same | |
CN105184766B (en) | A kind of level set image segmentation method of frequency domain boundary energy model | |
CN104392418B (en) | Image smoothing innovatory algorithm based on anisotropy parameter | |
CN102306378B (en) | Image enhancement method | |
CN107742291A (en) | A kind of the defects of photovoltaic glass detection method and device | |
CN116468641A (en) | Infrared blood vessel image enhancement processing method | |
CN105741281A (en) | Image edge detection method based on neighbourhood dispersion | |
CN102800066A (en) | Local extremum and pixel value gradient based improved image enhancement method | |
CN103455985A (en) | Road crack enhancement method based on Hessian structural analysis | |
Yu et al. | An improved method for cloth pattern cutting based on holistically-nested edge detection | |
Talai et al. | A fast edge detection using fuzzy rules | |
CN113344810A (en) | Image enhancement method based on dynamic data distribution | |
CN103413332A (en) | Image segmentation method based on two-channel texture segmentation active contour model | |
CN116485787B (en) | Method for detecting appearance defects of data line molding outer die | |
Spira et al. | Efficient Beltrami flow using a short time kernel | |
CN115965552B (en) | Frequency-space-time domain joint denoising and recovering system for low signal-to-noise ratio image sequence | |
CN111784610A (en) | Clustering-based side window filter optimization method | |
Li et al. | Image de-noising of locust sections based on adaptive wavelet and partial differential equation method | |
El Hassani et al. | Efficient image denoising method based on mathematical morphology reconstruction and the Non-Local Means filter for the MRI of the head | |
CN108389204B (en) | Degraded image fuzzy kernel double-parameter direct estimation method for high-speed online detection |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20150107 Termination date: 20210803 |