CN101383006A - Image pre-processing method of oval correction and circle normalization - Google Patents
Image pre-processing method of oval correction and circle normalization Download PDFInfo
- Publication number
- CN101383006A CN101383006A CNA2007100769112A CN200710076911A CN101383006A CN 101383006 A CN101383006 A CN 101383006A CN A2007100769112 A CNA2007100769112 A CN A2007100769112A CN 200710076911 A CN200710076911 A CN 200710076911A CN 101383006 A CN101383006 A CN 101383006A
- Authority
- CN
- China
- Prior art keywords
- image
- normalization
- oval
- circle
- processing method
- 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
Images
Landscapes
- Apparatus For Radiation Diagnosis (AREA)
Abstract
The invention discloses an image pre-processing method by oval correction and circle normalization. The method comprises the steps of identifying a targeted image from original images, carrying out oval fitting to boundary points of the targeted image, correction verifying and correction processing, and circle normalization. The invention corrects the inclination by an oval method, judges whether an image is inclined or not according to the situation of the oval fitting, also provides the method for correcting the image inclination and carries out the geometric normalization to the corrected image by the method of circle normalization so as to obtain good conformity in images and increase the conformity in image analysis. In addition, the method has significance to image analysis, such as CT image analysis, human face image identification and the like.
Description
Technical field
The present invention relates to the image preconditioning technique, the image pre-processing method of particularly a kind of oval correction and circle normalization.
Background technology
The image pre-service is the primary stage of graphical analysis.The image pre-service can comprise a lot of contents, for example to the correction of the normalization of the filtering and noise reduction of original image, gray scale, image level rotation, scale size even to removal of glasses and partial occlusion thing or the like.The pre-service quality will directly have influence on the result that successive image is analyzed.Therefore in the serial of methods and work of graphical analysis, carry out pretreated link and seem very important.
With regard to brain CT medical image, because the information that the distinct device that image acquisition is used, the different patients of identical device and different operating personnel collect all can have certain difference, how to eliminate these differences, make to have consistance between image.With regard to facial image, facial image with official portrait by equipment such as scanner or digital camera, video camera input computing machine, because the influence of switching device and surrounding environment, make image be subjected to noise, produce distortion, come out, cause real edge not to be detected and usually noise is used as endpoint detections when carrying out edge detection process, therefore, must remove noise.Because the influence of factors such as people's face collection environment or uneven illumination are even; cause the gray difference between facial image very big through regular meeting; at this moment just need carry out gray scale normalization in advance handles; the tonal range of different facial images is stretched to identical zone, just can discerns accordingly and matching treatment then.And for facial image, the feature of extraction must be carried out geometrical normalization earlier.
Summary of the invention
The image pre-processing method that the purpose of this invention is to provide a kind of oval correction and circle normalization, after adopting common way denoising and identifying the target area, with this image pre-processing method image is corrected with circle normalization and to be handled, can make to have good consistance between image, thereby can increase the consistance of graphical analysis.
For achieving the above object, oval correction of the present invention and circle normalization image pre-processing method may further comprise the steps:
The first step: the frontier point of recognition target image from original image;
Second step: the frontier point to target image carries out ellipse fitting;
The 3rd goes on foot: correct the step of checking and correction process, this step comprises whether the diaxon of differentiating ellipse is overlapping fully by translation with the rectangular coordinate system diaxon, goes to next step as if overlapping fully; Otherwise, carry out correction process, comprise and calculate the minimum anglec of rotation and direction, be rotated according to this angle and direction, realize the rectification of image, make the diaxon of target image and rectangular coordinate system overlapping, and remove the oval perimeter of rotation back image;
The 4th step: the step of circle normalization, this step comprise the major axis of oval image and minor axis are equal to specified length, make the oval image transformation of different major and minor axis become the circular image of same diameter.
In the first step, adopt overall histogram selected threshold, as choose methods such as adopting Otsu, computing machine identifies the border of target image automatically from original image; In second step, computing machine carries out the ellipse fitting of least square method according to the border of the target image that identifies.
In the 4th step, long axis of ellipse and minor axis are changed into identical specified length, realize the conversion of oval all normalization circle; Described specified length is according to concrete processing image, artificial stipulates radius of a circle length behind this type of image normalization.
The present invention adopts oval method to correct degree of tilt, judge according to the ellipse fitting situation whether image tilts, and provided the method that correcting image tilts, and adopt the method for circle normalization that the image after correcting is carried out geometrical normalization, thereby make to have good consistance between image, can increase the consistance of graphical analysis.It is significant as CT graphical analysis etc. to graphical analysis.
Description of drawings
Fig. 1 is the process flow diagram of image pre-processing method of the present invention;
Fig. 2 is two groups of brain CT Flame Image Process contrast figure, wherein first row and the third line are the stacking diagrams of original image and red line fitted ellipse, second row and fourth line be respectively correspondence first go and the third line original image by the pretreated result images of the technology of the present invention (artificially the normalization length of regulation CT image is 256);
Figure 3 shows that brain CT image processing effect analysis chart, wherein, Fig. 3 a is single people's the result of brain CT cross sectional image after the inventive method is handled; Fig. 3 b is the stacking diagram of close or identical brain CT cross sectional image after the inventive method is handled of 10 different people; Fig. 3 c is the stacking diagram's (result images normalization length is 256) who does not carry out pretreated 10 the close or identical brain CT of different people cross sectional image.
Fig. 4 is the contrast figure of lineup's face image before and after the inventive method is handled, wherein first row is an original image, second row is the stacking diagram with the red line fitted ellipse, the third line is by the pretreated result images of the technology of the present invention (result images normalization length is 92, also is the length of artificial regulation).
Embodiment
With reference to Fig. 1, the image pre-processing method of this oval correction and circle normalization may further comprise the steps:
1, recognition objective and carry out ellipse fitting, concrete grammar is: adopt overall histogram selected threshold, as choose methods such as adopting Otsu, automatically identify the border of target image; And carry out the ellipse fitting of least square method according to this border;
2, oval correction checking and correction process, concrete grammar is: whether the diaxon of judging the oval image after the match judges by the translation realization is overlapping whether this image needs to correct by rotation with rectangular coordinate system; If overlapping fully, then need not correct, go to next step; If not exclusively overlapping then the needs corrects, calculate the minimum anglec of rotation and direction, according to this angle and direction image rotating, realize the rectification of image, make the diaxon of target image and rectangular coordinate system overlapping, and remove the oval perimeter of rotation back image;
3, circle normalization, concrete grammar is: long axis of ellipse and minor axis are changed into identical specified length, realize oval normalization conversion to the same radius circle; Described specified length is according to concrete processing image, the artificial radius length of stipulating garden behind this type of image normalization.
The present invention adopts oval method to correct degree of tilt, judge according to the ellipse fitting situation whether image tilts, and provided the method that correcting image tilts, and adopt the method for circle normalization that the image after correcting is carried out geometrical normalization, thereby make to have good consistance between image, can increase the consistance of graphical analysis.
Adopt the inventive method effectively to eliminate to use the difference between the image that distinct device or the different patients of identical device or different operating personnel collect, make and have consistance between image, analysis for the CT medical image is significant, illustrates below.
Figure 2 shows that two groups of brain CT cross sectional image processing contrast figure, wherein first the row and the third line be the stacking diagram of original image and red line fitted ellipse, second the row and fourth line be respectively the correspondence first the row and the third line original image by the pretreated result images of the technology of the present invention (result images normalization length is 256).Can obviously find out: compare with first row, four images of second row have consistance preferably; Compare with the third line, four images of fourth line have consistance preferably.
Figure 3 shows that brain CT image processing effect analysis chart, wherein, Fig. 3 a is single people's the result of brain CT cross sectional image after the inventive method is handled; Fig. 3 b is the stacking diagram of close or identical brain CT cross sectional image after the inventive method is handled of 10 different people; Fig. 3 c is the stacking diagram's (result images normalization length is 256) who does not carry out pretreated 10 the close or identical brain CT of different people cross sectional image.Clearly can see: Fig. 3 b is more clear, can see the most information in the preceding image of stack clearly; And Fig. 3 c is very fuzzy, almost can't see any information in the preceding image of stack, and as seen the image after the inventive method is handled has good consistance.
Fig. 4 is the contrast figure of lineup's face image before and after the inventive method is handled, wherein first row is an original image, second row is that the third line is by the pretreated result images of the technology of the present invention (result images normalization length is 92) with the stacking diagram of red line fitted ellipse.Can find out significantly that five images in the third line have consistance preferably.
Claims (3)
1, the image pre-processing method of a kind of oval correction and circle normalization is characterized in that may further comprise the steps:
The first step: the frontier point of recognition target image from original image;
Second step: the frontier point to target image carries out ellipse fitting;
The 3rd step: the step of correcting checking and correction process, this step comprises whether the diaxon of differentiating ellipse is overlapping fully by translation with the rectangular coordinate system diaxon, go to next step as if overlapping fully, otherwise, then calculate the minimum anglec of rotation and direction, purpose is to realize that the diaxon and the rectangular coordinate system of target image are overlapping, is rotated according to this angle and direction, realize the rectification of image, and remove the oval perimeter of rotation back image;
The 4th step: the step of circle normalization: the major axis and the minor axis of oval image are equal to specified length, make the circular image of oval image transformation; Described specified length is the difference according to research object, the radius length of predefined normalization circular image.
2, image pre-processing method according to claim 1 is characterized in that: in the first step, adopt overall histogram selected threshold, computing machine identifies the border of target image automatically from original image; In second step, computing machine carries out the ellipse fitting of least square method according to the border of the target image that identifies.
3, image pre-processing method according to claim 1 is characterized in that: in the 4th step, long axis of ellipse and minor axis are changed into identical specified length, realize the conversion of oval all normalization circle; Described specified length is according to concrete processing image, artificial stipulates radius of a circle length behind this type of image normalization.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2007100769112A CN101383006A (en) | 2007-09-05 | 2007-09-05 | Image pre-processing method of oval correction and circle normalization |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2007100769112A CN101383006A (en) | 2007-09-05 | 2007-09-05 | Image pre-processing method of oval correction and circle normalization |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101383006A true CN101383006A (en) | 2009-03-11 |
Family
ID=40462839
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNA2007100769112A Pending CN101383006A (en) | 2007-09-05 | 2007-09-05 | Image pre-processing method of oval correction and circle normalization |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101383006A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101825448A (en) * | 2010-03-30 | 2010-09-08 | 中国计量学院 | Method for determining included angle between lens plane of thermal infrared imager and plane to be measured |
| CN102043951A (en) * | 2010-12-31 | 2011-05-04 | 大连理工大学 | Joint finger segmentation system |
| CN106846395A (en) * | 2016-12-31 | 2017-06-13 | 中国农业科学院农业环境与可持续发展研究所 | Targeted graphical area computation method and system in photo |
| CN109785296A (en) * | 2018-12-25 | 2019-05-21 | 西安电子科技大学 | A kind of spherical assessment of indices method of three-dimensional based on CTA image |
-
2007
- 2007-09-05 CN CNA2007100769112A patent/CN101383006A/en active Pending
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101825448A (en) * | 2010-03-30 | 2010-09-08 | 中国计量学院 | Method for determining included angle between lens plane of thermal infrared imager and plane to be measured |
| CN102043951A (en) * | 2010-12-31 | 2011-05-04 | 大连理工大学 | Joint finger segmentation system |
| CN102043951B (en) * | 2010-12-31 | 2012-07-25 | 大连理工大学 | Joint finger segmentation method |
| CN106846395A (en) * | 2016-12-31 | 2017-06-13 | 中国农业科学院农业环境与可持续发展研究所 | Targeted graphical area computation method and system in photo |
| CN106846395B (en) * | 2016-12-31 | 2019-12-27 | 中国农业科学院农业环境与可持续发展研究所 | Method and system for calculating area of target graph in photo |
| CN109785296A (en) * | 2018-12-25 | 2019-05-21 | 西安电子科技大学 | A kind of spherical assessment of indices method of three-dimensional based on CTA image |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR100374708B1 (en) | Non-contact type human iris recognition method by correction of rotated iris image | |
| CN106096569B (en) | A kind of finger vein identification method | |
| CN106339702A (en) | Multi-feature fusion based face identification method | |
| US20050196044A1 (en) | Method of extracting candidate human region within image, system for extracting candidate human region, program for extracting candidate human region, method of discerning top and bottom of human image, system for discerning top and bottom, and program for discerning top and bottom | |
| CN109002799B (en) | Face recognition method | |
| EP1296279A3 (en) | Method and computer program product for locating facial features | |
| CN108171674B (en) | Vision correction method for projector image with any visual angle | |
| CN110909631B (en) | Finger vein image ROI extraction and enhancement method | |
| CN101383006A (en) | Image pre-processing method of oval correction and circle normalization | |
| Mohammed et al. | Accurate pupil features extraction based on new projection function | |
| JP2000105819A (en) | Face image area detecting device | |
| CN101364257B (en) | Human face recognizing method for recognizing image origin | |
| CN108960100A (en) | A kind of recognition methods of the sugarcane sugarcane section based on image procossing | |
| CN106650735B (en) | A kind of LED character automatic positioning recognition methods | |
| CN113034526B (en) | Grabbing method, grabbing device and robot | |
| Li et al. | An efficient face normalization algorithm based on eyes detection | |
| EP2287782A3 (en) | System, Method And Program Product For Camera-Based Object Analysis | |
| JP7044504B2 (en) | Image processing device, image processing method and image processing program | |
| CN106909880A (en) | Facial image preprocess method in recognition of face | |
| CN113642570A (en) | Method for recognizing license plate of mine car in dark environment | |
| CN105721738B (en) | A kind of chromoscan file and picture preprocess method | |
| JP2001022929A (en) | Method and device for detecting colony microorganism | |
| Aydi et al. | A fast and accurate eyelids and eyelashes detection approach for iris segmentation | |
| CN105488803A (en) | Human eye pupil image judgment method | |
| Wen et al. | Lightning arrester monitor pointer meter and digits reading recognition based on image processing |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Open date: 20090311 |