CN100410969C - Medical radiation image detail enhancing method - Google Patents
Medical radiation image detail enhancing method Download PDFInfo
- Publication number
- CN100410969C CN100410969C CNB2006100618796A CN200610061879A CN100410969C CN 100410969 C CN100410969 C CN 100410969C CN B2006100618796 A CNB2006100618796 A CN B2006100618796A CN 200610061879 A CN200610061879 A CN 200610061879A CN 100410969 C CN100410969 C CN 100410969C
- Authority
- CN
- China
- Prior art keywords
- image
- subimage
- pyramid
- gauss
- layer
- 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.)
- Active
Links
Images
Landscapes
- Image Processing (AREA)
- Apparatus For Radiation Diagnosis (AREA)
Abstract
The present invention disclose a method which can strengthen the detail of medical radial image, it includes the follow steps: A. the primitive images are gradual decomposed as Gauss decomposition to Gauss pyramid and Laplace pyramid. Gauss pyramid is formed with multilayer Gauss sub-images Gj of the primitive image. Laplace pyramid is formed with multilayer Laplace sub-image Lj of the primitive images; B. The Laplace coefficient of the Laplace sub-image Lj in each layer of the Laplace pyramid is adjusted to strengthen the details of the Laplace sub-image Lj ; C. the reinforced Laplace sub-image L' j plus corresponding brother Gauss sub-image G' j until it arrives at the bottom of Laplace pyramid to accomplish the reconstruction of the image. The image which is processed with this method will neither to be brighter nor darker, both the contrast and the hierarchy are clear. It solved the equalizing problem of the tissue in radial image, especially the density equalization of radial image of the air, fat, soft tissue and skeleton.
Description
Technical field
The present invention relates to the digital image processing techniques field, especially a kind of detail enhancing method of medical radiation image.
Background technology
Radiodiagnosis is used for the existing over one hundred year history of clinical medicine, although some other advanced Imaging Techniques as the diagnosis to a part of disease such as CT and MRI, have demonstrated bigger superiority, they can not replace X-ray examination.The inspection at some positions, intestines and stomach for example, osteoarthrosis and cardiovascular still mainly uses X-ray examination.The X line also has characteristics such as imaging is clear, economic, easy, and therefore, at home and abroad, radiodiagnosis remains and uses the most extensive in the diagnostic imaging and fundamental method.
Radiation image is the projection summation after X-ray beam penetrates a certain position different densities and thickness institutional framework, is that this penetrates each layer projection mutual superposition image together on the path.Overlapping result makes the projection of some institutional framework in the body well be shown because of storage gain, and the projection of other institutional frameworks in the body maybe can not be shown than difficulty because of weakening to offset.
Adopt chemical method to handle X ray radiation image film, the processing time is long.If film exposure is not enough or transillumination angle mistake, must carry out all programs again.Though the spatial resolution of film is better, but film is linear bad and contrast range is narrow, adds the limitation that human eye has, resolving ability can not surpass 100 grey level, can not detect and obtain more accurate data from a film density that haves a wide reach.If use image intensifier, its range of application are subjected to its restriction of protecting bulky and the ken again, and the appearance distortion of the edge of image, have only the image of center useful for some application.In addition, the contrast of image intensifier and spatial resolution can not be mentioned in the same breath with other technology.No matter film or image intensifier, file and distribution all inconvenient.
Adopt computerized X ray technology, make x-ray imaging that huge variation take place as CR (Computed Radiography, computer X-ray photography), DR (Digital Radiography, digital X radiography).The contrast precision of CR is 12 or 4096 gray scales, and is similar with film.Although its spatial resolution does not also surpass film, readding sheet for most of medical science uses enough, the precision of CR be 5 lines right/millimeter (i.e. 100 μ m), because its contrast range is very big, CR can be applied to all digital X-ray technology basically, just can obtain the full depth scope of most reference objects by an inlet.By computing machine, can browse any interested position of full depth scope.But the existing shortcoming of CR is: influence the quality of photo after memory board is aging, need special reader.DR extracts X ray by digital flat panel directly to transform into digital picture, compares film and CR, and the DR technology all has remarkable advantages improve contrast, raising spatial resolution, reduction noise jamming, raising intensity and reducing cost etc.
The factor that influences the radiation image quality has a lot, aging such as digitizing tablet, the resolution of display screen, technique for taking of technician or the like.Good medical radioactive image requires well arranged, and good spatial level sense is arranged.Object for observing requires clear-cut, and there were significant differences with background.The general medical image of gathering from digital medical equipment, because dynamic range is excessive, add complicated human tissue structure, cause image " dim ", trickle focus is hidden in the useless information, early stage such as cancer, cancerous issue is more tiny, but because the medical image image quality is not high, has strengthened the difficulty of diagnosis; pathological tissues develops into certain size and could find, has missed the best opportunity of treatment; And traditional window width and window level adjusting observation image is wasted time and energy, and has seen place tissue clearly, and other is organized and became bright or dark excessively, and can't see the detailed information of image clearly.
Existing digital picture enhancement techniques is divided into two kinds of spatial domain and frequency fields." spatial domain " more typically has greyscale transformation, histogram equalization, spatial filtering method or the like image pixel directly is treated to the basis.The shortcoming of these methods is that image is regarded as the simple combination of pixel one by one, does not have the relevance between the considered pixel.Because the singularity of medical image, each pixel causes these pixels to vary for the contribution that the observer understands image because residing position is different.Be easy to arouse attention such as the pixel that is positioned at the edge, the quantity of information that variation produced of these pixels will be far longer than the pixel that is positioned at smooth region so.Therefore, if in single global scope, handle image again, just can not well give prominence to detailed information.
" frequency domain " treatment technology is to revise the Fourier transform of image.The basic step of frequency filtering is illustrated in fig. 1 shown below, because most sharp-pointed details have identical frequency characteristic with noise in the image, the greatest problem that adopts frequency field to handle has no idea to distinguish useful information and garbage exactly.The most frequently used wavelet transformation of frequency domain transform is based on the pyramid algorith of Mallat at present, though advantage is outstanding, it has time shift susceptibility, and this time shift susceptibility is caused by down-sampling.Therefore, wavelet transformation can produce " ringing effect ", brings artificial pseudo-shadow, and this is unallowed in medical imaging is handled.
Summary of the invention
Technical matters to be solved by this invention is the above-mentioned defective that exists at prior art, a kind of detail enhancing method of medical radiation image is provided, make image after treatment, contrast is clear, stereovision is obvious, solved and organized equalization problem in the irradiation image, solved the equalization problem of dynamic large tracts of land average density and detail edges, avoided " ringing effect " as far as possible.
The technical scheme that solution the technology of the present invention problem is adopted is: a kind of detail enhancing method of medical radiation image comprises the steps:
A, raw image is carried out Gauss step by step decompose, form multilayer Gauss's subimage G of raw image
jGaussian pyramid and multilayer Laplce image subsection L
jLaplacian pyramid;
B, to each straton image L of laplacian pyramid
jLaplace coefficient adjust, to strengthen each layer Laplce image subsection L
jDetails;
C, will strengthen the Laplce's image subsection L ' after the details
jThe at the same level Gauss subimage G ' corresponding with it
jAddition is until arriving this pyramidal bottom of pula to finish image reconstruction.
Described step B specifically may further comprise the steps:
B1, the gray-scale value of Laplce's subimage is mapped to [1,1], promptly for each the pixel P in Laplce's subimage
I, j, all divided by max pixel value max|P
I, j|, max|P
I, j| max|P by formula
I, j|=max{P
I, j0≤i＜ROW, 0≤j＜COLUMN calculates, and wherein, ROW, COLUMN are respectively the line number and the columns of Laplce's subimage; Obtain the pixel value P ' after the normalization thus
I, j:
B2, to P '
I, jCarry out the ^p conversion, degree of comparing equilibrium just obtains the pixel P after the equilibrium "
I, j
B3, with the pixel P after the equilibrium "
I, jMultiply by an overall intensity adjustment factor m,
P″′
i，j＝P″
i，j×max|P
i，j|×m，
To eliminate after the contrast equilibrium influence to integral image.
Described steps A specifically may further comprise the steps:
A1, with original image the 0th layer as gaussian pyramid, carry out 5*5 low pass Gaussian convolution nuclear low-pass filtering, form the 1st straton image through down-sampling, be benchmark image with the 1st straton image again, operation successively, the number of times of operation is pre-defined to be n, it is the number of plies of corresponding gaussian pyramid, the level of last layer subimage and vertical resolution are 1/2 of next straton image respectively, reach after the predefined level, and the subimage resolution of this moment is 1/2 of original image
n, n is the number of plies of gaussian pyramid;
A2, from the top of gaussian pyramid, i.e. this subimage of n floor height G
nBeginning is carried out up-sampling, Gauss's subimage G ' of formation to the Gauss's subimage that is positioned at this layer with 5*5 low pass Gaussian convolution nuclear
N-1With the Gauss's subimage G that was positioned at the n-1 layer originally
N-1Has identical resolution, G '
N-1And G
N-1Difference be exactly Laplce's subimage L of n-1 layer
N-1, and the like, form laplacian pyramid.
P″′
i，j＝P″
i，j×max|P
i，j|×m，
The decomposition number of plies of described gaussian pyramid and laplacian pyramid depends on the size of original image and the location information that needs strengthen, requirement according to medical image, for the bigger skeletal sites of density, decompose the number of plies and select the 5-7 layer, for the less soft tissue position of density, decompose the number of plies and select the 7-10 layer.
The beneficial effect that the present invention produced is: the image after the processing can be not bright partially or dark partially, contrast is clear, stereovision is obvious, has solved and has organized equalization problem in the irradiation image, especially to the equilibrium treatment of the radiation image density of air, fat, soft tissue, bone.
The present invention adopts Gauss-Laplce's gold tower method to use different reinforcing coefficient at different levels, has well solved the problem of how to give prominence to the image detail feature, has solved the equalization problem of dynamic large tracts of land average density and detail edges.
The invention provides the details parameter regulation instrument corresponding to each level, the doctor can regulate parameter according to observed position and personal habits oneself, up to obtaining best diagnosis effect.
Description of drawings
Fig. 1 is the basic step block scheme of the frequency filtering of prior art;
Fig. 2 is a block scheme of setting up image pyramid;
Fig. 3 is the structural representation of image pyramid;
Fig. 4 is the main flow chart of technical solution of the present invention;
Fig. 5 is gaussian pyramid and laplacian pyramid schematic diagram;
Fig. 6 is the pairing mathematic graph of ^p conversion;
Fig. 7, Fig. 8 are respectively through the effect contrast figure before and after the detail enhancing method of the present invention;
Fig. 9, Figure 10 are respectively through the effect contrast figure before and after the detail enhancing method of the present invention;
Figure 11 is that details of the present invention strengthens the parameter regulation surface chart.
Embodiment
The present invention is a kind of detail enhancing method of brand-new medical radiation image, at digitized radiation image, adopt the detailed information under the image pyramid structure extraction different resolution, adopt different adjusting parameters according to the residing level of details, both faithful to original image has improved the details quality again.
The present invention has adopted the Flame Image Process mode similar to the human visual system, because human eye has the characteristic of central authorities-on every side, the vision content of paying close attention under different resolution is inequality.Such as when the low resolution, the general shape that human eye can only resolution target, and when high resolving power, human eye can be seen lines, texture of target etc. clearly.Therefore, the present invention strengthens different details under different resolution.Use the different visual signatures of the method representation image of stratification.In a width of cloth gray level image, the all corresponding certain brightness of each pixel, for the human visual system, the brightness of pixel is not only a kind of signal excitation, because people's brain has the image understanding function, therefore, these signals can further describe into low-level image feature, as edge, line, also can be described as high-level characteristic, as texture, zone and object bounds, object self and relation between objects or the like.One width of cloth digital picture can be regarded as by different frequency range, and the subimage of different size combines.Each subimage has comprised some information specific.When adopting different observation multiples.Subimage shows different details.Each subimage is handled accordingly, just can be strengthened the information of correspondingly-sized.
When observing the medical radioactive image, what see usually is the texture zone similar to gray scale that is connected, and these zones are in conjunction with forming object.If the small-sized or contrast of object is not high, need to adopt high resolving power; If the size of object is very big or contrast is very strong, then only need lower resolution.If the size of object varies, contrast has to be had by force secretly, and analysis image just has advantage on different resolution so.
Adopt a kind of simple effective method of multiresolution technical Analysis image to adopt the image pyramid method exactly.So-called image pyramid is the image collection that a series of resolution of arranging with Pyramid progressively reduce, and the image of pyramidal low layer is an original image, and the high-level diagram picture is from low tomographic image.Fig. 2 has represented a step of setting up the image pyramid process, promptly how to produce the j+1 layer by the j layer, and produces the residual error of j layer.
In Fig. 2, the approximate output of j+1 layer is used for setting up the approximate value that image pyramid is positioned at the j+1 layer, adopts different wave filters, can obtain a plurality of different approximate value of original image.The residual error output of j layer is used to set up residual error pyramid, i.e. laplacian pyramid.For the subimage of j layer, at the information I of corresponding approximate pyramid j layer
jThe approximate evaluation I ' that obtains with subimage up-sampling based on the j+1 layer
jBetween be different, the difference between them has constituted the residual error of j layer.Approximate value and residual error pyramid all are to calculate in a kind of mode of iteration.Each iteration is made up of 3 steps:
1. the approximate value of the resolution of the minimizing of calculating input image.This can by to the input image carry out filtering and with 2 be step-length sample (being down-sampling) obtain.Carrying out filtering method has: neighborhood averaging, Gauss's low-pass filtering or directly carry out sub sampling in spatial domain and filter.Low-resolution image is used to analyze the whole content of big structure or image, and high-definition picture is used to analyze the characteristic of single body.If there is not wave filter, it is very fuzzy that the pyramid last layer can become, and sampled point is well not representative for the zone of being sampled yet.
2. the approximate value of the resolution that the input picture of previous step is reduced is that step-length is carried out interpolation (being up-sampling) and filtered with 2, has generated the predicted picture with resolution such as input.Owing between the output pixel of step 1, carry out interpolation arithmetic, the type decided of wave filter the degree of approximation between the input of predicted value and step 1.If ignore the insertion wave filter, predicted value is exactly the interpolation form of step 1 output, and the mosaic effect of copy pixel can be clearly.
3. predicted value in the calculation procedure 2 and the difference between the input in the step 1 identify this difference with j level residual error, and this difference will be used for the reconstruction of original image.Do not having under the situation of quantitative differences, the prediction residual pyramid can be used for generating corresponding approximate pyramid, comprises original image.This process is as shown in Figure 3:
J (the image P of the layer of 1＜j＜J)
jGenerate the upper level subgraph through filtering, down-sampling, promptly be positioned at the image P of pyramid j+1 layer
J-1, and P
J-1The next stage image P ' that generates during again through filtering, up-sampling
jThe time, since the effect of wave filter, P
jAnd P '
jBetween there are differences, and this difference only can be extracted and be showed in the transfer process of different resolution.
Technical scheme of the present invention can be divided into 3 stages, and promptly gaussian pyramid decomposes, the reconstruction of laplacian pyramid coefficient adjustment and gaussian pyramid, and idiographic flow can be consulted Fig. 4.
At Gauss's catabolic phase, be benchmark with the original image, also be original image is positioned at the 0th layer, carry out 5*5 low pass Gaussian convolution nuclear low-pass filtering, form the 1st straton image through down-sampling.The level of subimage and vertical resolution respectively are 1/2 of benchmark image; Be benchmark image with the 1st straton image again, operation successively.The number of times of operation can be pre-defined, i.e. the number of plies of corresponding gaussian pyramid.Reach after the predefined level, the subimage resolution of this moment is 1/2 of original image
n, n is the number of plies of gaussian pyramid.Then, from the top of gaussian pyramid, i.e. n straton image G
nBeginning is carried out up-sampling, the subimage G ' of formation to the subimage that is positioned at this layer with 5*5 low pass Gaussian convolution nuclear
N-1With the subimage G that was positioned at the n-1 layer originally
N-1Has identical resolution.G '
N-1And G
N-1Difference be exactly Laplce's subimage L of n-1 layer
N-1And the like, can obtain laplacian pyramid, principle as shown in Figure 5, the subimage that laplacian pyramid is positioned at the j level be Gauss's subimage at the same level with j+1 level Gauss subimage up-sampling after subimage poor of Gauss's low-pass filtering formation again.According to the characteristic of gaussian filtering, Laplce's subimage has reflected the detailed information of image.
In the laplace coefficient adjusting stage, because the profile of image mainly is distributed in low frequency part, noise mainly is distributed in HFS, and most of useful information is distributed in intermediate-frequency section, so the improvement of picture quality need suppress low frequency, high-frequency energy, outstanding intermediate frequency.By research relatively, the ^p conversion just in time can be satisfied this demand, and the function representation form of ^p conversion is as follows:
F (x)=x
pX ∈ [1,1] wherein, 0＜p≤1 (1)
So the mathematic graph of ^p conversion correspondence as shown in Figure 6.
As can be seen from the figure, the ^p transform method strengthens the amplitude maximum to coefficient less in the subimage, thereby has significantly improved the visual degree of corresponding details.
The present invention is mapped to [1,1] by removing max methods with the gray-scale value of Laplce's subimage, promptly for each the pixel P in Laplce's subimage
I, j, all divided by max pixel value max|P
I, j|, max|P
I, j| calculate according to following formula:
max|P
i，j|＝max{P
i，j} 0≤i＜m，0≤j＜n (2)
Wherein, m, n are respectively the line number and the columns of Laplce's subimage.
Obtain the pixel value P ' after the normalization thus
I, j
P ' then
I, jTo carrying out the ^p conversion, degree of comparing equilibrium just obtains the pixel P after the equilibrium "
I, j
If original image is crossed bright or dark excessively, produce a polarization even more serious after the ^p conversion so.In order to prevent this phenomenon, we have added an overall intensity adjustment factor m in the inverse transformation process, to improve or to reduce the overall contrast of image.Promptly
P″′
i，j＝P″
i，j×max|P
i，j|×m (4)
The effect of multiply by adjustment factor m is in order to eliminate after the contrast equilibrium the influence of integral image, to be equivalent to " macro adjustments and controls ".
Analyze from mathematics, the p value of ^p conversion is more little, and curve is crooked more, and promptly the amplitude of Tiao Jieing is big more.Therefore, for Laplce's subimage of different levels, the value of the p that is adopted is also different.And the optimum p value scheme that different types of medical radioactive image is taked when strengthening is also different.The present invention has provided the optimum p value scheme of common medical radioactive image according to the situation of practical application, and is as shown in table 1:
The image category in table 1 practical application and the parameter group table of comparisons
At phase of regeneration, with the Laplce's subimage L ' after strengthening
jBe added back in the Gauss's subimage with one deck.Concrete steps are:
1, as can be seen, the laplacian pyramid top i.e. the n-1 layer of corresponding gaussian pyramid, for convenient expression, might as well count L from Gauss's catabolic phase
N-1L
N-1After the laplace coefficient adjustment, be output as L '
N-1
2, will be through Laplce's subimage after strengthening and the gaussian pyramid subimage addition of layer together, promptly
G′
n-1＝G
n-1+L′
n-1 (5)
Like this, Gauss's subimage details originally is enhanced, and not very big change of other parts of image.
3, progressively downward from the n-1 layer of gaussian pyramid, repeat 1-2 step operation, up to the bottom of gaussian pyramid, i.e. original image.Such process of reconstruction strengthens amplitude by coefficient adjustment by progressively improving the image detail method for quality on each level, be added to the Gaussian image bottom at last, and promptly during original image, the detailed information of image can show fully.
In the present invention, the decomposition number of plies of Gauss-laplacian pyramid can be regulated.The number of plies of decomposing depends on the size of original image and the location information that needs strengthen.According to the requirement of medical image,, strengthen the number of plies and can select the 5-7 layer for the bigger positions such as bone of density; For the less positions such as soft tissue of density, strengthen the number of plies and can select the 7-10 layer.Because method is to the bottom, so more near the subimage at pyramid top, its index variation is big more to the influence of final output image from the computing layer by layer of pyramid top.Therefore, in non-linear adjusting, for the subimage near the pyramid top, the coefficient of ^P conversion is more near 1, promptly weak the enhancing; And more near the subimage of pyramid bottom, the coefficient of ^P conversion can be very little, promptly strengthens by force.
In addition, adopt the method for the thought generation image pyramid of multiresolution to have a lot, also can use the method for orthogonal wavelet transformation or Double Density Wavelet Transform.Because realize that Double Density Wavelet Transform need adopt the over-sampling wave filter, rather than the threshold sampling of traditional wavelet, therefore wavelet coefficient is redundant in the conversion of dual density wavelet decomposition, and has approximate TIME SHIFT INVARIANCE.In addition, the information that the conversion of dual density wavelet decomposition obtains can overcome the Ma Saike phenomenon of orthogonal wavelet preferably, helps improving image edge information.Because orthogonal wavelet transformation and Double Density Wavelet Transform have adopted over-sampling, have TIME SHIFT INVARIANCE, thus can not cause pseudo-shadow after strengthening yet, but computation complexity will be higher than method used in the present invention.
Practice result proves that the effect that adopts multiresolution to strengthen the generation of details method is better than common image enchancing method, can not bring artificial vestige, faithful to primitive medicine image.Its effect can be consulted Fig. 7,8 and Fig. 9,10 effect contrast figure.
The doctor can preset the reinforcing coefficient scheme, also can own finely tune at different level (i.e. the amplitude of accommodation among the figure), sees also Figure 11, and the effect after details strengthens shows in the drawings in real time, till the doctor is satisfied.The doctor can directly select " finishing ", also can store the form of this group adjusting parameter with scheme hereof, for directly calling next time.
Should be understood that; the above only is a preferred implementation of the present invention, for those skilled in the art, and under the prerequisite that does not break away from the principle of the invention; can also make some improvements and modifications, these improvements and modifications also should be considered as protection scope of the present invention.
Claims (3)
1. the detail enhancing method of a medical radiation image comprises the steps:
A, raw image is carried out Gauss step by step decompose, form gaussian pyramid and the multilayer Laplce image subsection L of multilayer Gauss's subimage Gj of raw image
jLaplacian pyramid;
B, to each straton image L of laplacian pyramid
jLaplace coefficient adjust, to strengthen each layer Laplce image subsection L
jDetails;
C, will strengthen the Laplce's image subsection L ' after the details
jThe at the same level Gauss subimage G ' corresponding with it
jAddition until the bottom that arrives laplacian pyramid to finish image reconstruction;
It is characterized in that described step B specifically may further comprise the steps:
B1, the gray-scale value of Laplce's subimage is mapped to [1,1], promptly for each the pixel P in Laplce's subimage
I, j, all divided by max pixel value max|P
I, j|, max|P
I, j| by formula
Max|P
I, j|=max{P
I, j0≤i＜ROW, 0≤j＜COLUMN calculates,
Wherein, ROW, COLUMN are respectively the line number and the columns of Laplce's subimage;
Obtain the pixel value P ' after the normalization thus
I, j:
B2, to P '
I, jCarry out the ^p conversion, degree of comparing equilibrium just obtains the pixel P after the equilibrium "
I, j
B3, with the pixel P after the equilibrium "
I, jMultiply by an overall intensity adjustment factor m,
P′″
i，j＝P″
i，j×max|P
i，j|×m，
To eliminate after the contrast equilibrium influence to integral image.
2. the detail enhancing method of medical radiation image according to claim 1 is characterized in that described steps A specifically may further comprise the steps:
A1, with original image the 0th layer as gaussian pyramid, carry out 5*5 low pass Gaussian convolution nuclear low-pass filtering, form the 1st straton image through down-sampling, be benchmark image with the 1st straton image again, operation successively, the number of times of operation is pre-defined to be n, it is the number of plies of corresponding gaussian pyramid, the level of last layer subimage and vertical resolution are 1/2 of next straton image respectively, reach after the predefined level, and the subimage resolution of this moment is 1/2 of original image
n, n is the number of plies of gaussian pyramid;
A2, from the top of gaussian pyramid, i.e. this subimage of n floor height G
nBeginning is carried out up-sampling, Gauss's subimage G ' of formation to the Gauss's subimage that is positioned at this layer with 5*5 low pass Gaussian convolution nuclear
N-1With the Gauss's subimage G that was positioned at the n-1 layer originally
N-1Has identical resolution, G '
N-1And G
N-1Difference be exactly Laplce's subimage L of n-1 layer
N-1, and the like, form laplacian pyramid.
3. the detail enhancing method of medical radiation image according to claim 1 and 2, it is characterized in that: the decomposition number of plies of described gaussian pyramid and laplacian pyramid depends on the size of original image and the location information that needs strengthen, requirement according to medical image, for the bigger skeletal sites of density, decompose the number of plies and select the 5-7 layer, for the less soft tissue position of density, decompose the number of plies and select the 7-10 layer.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2006100618796A CN100410969C (en) | 2006-07-26 | 2006-07-26 | Medical radiation image detail enhancing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2006100618796A CN100410969C (en) | 2006-07-26 | 2006-07-26 | Medical radiation image detail enhancing method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1889125A CN1889125A (en) | 2007-01-03 |
| CN100410969C true CN100410969C (en) | 2008-08-13 |
Family
ID=37578400
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2006100618796A Active CN100410969C (en) | 2006-07-26 | 2006-07-26 | Medical radiation image detail enhancing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100410969C (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101430787B (en) * | 2008-11-14 | 2012-07-04 | 西安交通大学 | Digital sternum heart calcification image intensification method |
Families Citing this family (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100461218C (en) * | 2007-03-29 | 2009-02-11 | 杭州电子科技大学 | Method for enhancing medical image with multi-scale self-adaptive contrast change |
| CN101366638B (en) * | 2007-08-17 | 2011-01-26 | 上海西门子医疗器械有限公司 | Method for improving image quality |
| US8406507B2 (en) * | 2009-01-14 | 2013-03-26 | A9.Com, Inc. | Method and system for representing image patches |
| CN101853517B (en) * | 2010-05-26 | 2011-11-16 | 西安交通大学 | Real image oil painting automatic generation method based on stroke limit and texture |
| CN102136138B (en) * | 2011-03-28 | 2013-01-16 | 深圳市蓝韵实业有限公司 | Processing method for denoising and enhancing medical image |
| CN102722871B (en) * | 2012-05-24 | 2015-02-11 | 中山大学 | Quick and effective image enhancing method |
| CN104715453A (en) * | 2013-12-12 | 2015-06-17 | 苏州长风航空电子有限公司 | Image enhancement method by adopting regional processing mode and circuit |
| CN104182939B (en) * | 2014-08-18 | 2017-02-15 | 成都金盘电子科大多媒体技术有限公司 | Medical image detail enhancement method |
| WO2016080311A1 (en) * | 2014-11-21 | 2016-05-26 | 株式会社日立製作所 | X-ray ct device, projection data upsampling method, and image reconstruction method |
| CN104680495B (en) * | 2015-03-17 | 2017-12-22 | 武汉倍尔生生物科技有限公司 | The self-adaptive solution method of ultrasonoscopy |
| CN106570831A (en) * | 2016-10-09 | 2017-04-19 | 中国航空工业集团公司洛阳电光设备研究所 | Gray image contrast equalization enhancement method |
| WO2018086000A1 (en) * | 2016-11-09 | 2018-05-17 | Edan Instruments, Inc. | Systems and methods for ultrasound imaging |
| CN109801279B (en) * | 2019-01-21 | 2021-02-02 | 京东方科技集团股份有限公司 | Method and device for detecting target in image, electronic equipment and storage medium |
| CN110021031B (en) * | 2019-03-29 | 2023-03-10 | 中广核贝谷科技有限公司 | X-ray image enhancement method based on image pyramid |
| CN111462004B (en) * | 2020-03-30 | 2023-03-21 | 推想医疗科技股份有限公司 | Image enhancement method and device, computer equipment and storage medium |
| CN111968045B (en) * | 2020-07-20 | 2023-09-05 | 华南理工大学 | Method for removing stripe noise of X-ray image |
| CN112053279A (en) * | 2020-09-03 | 2020-12-08 | 五邑大学 | Dimension reduction method and device for hyperspectral image and storage medium |
| CN114549377B (en) * | 2022-01-11 | 2024-02-02 | 上海应用技术大学 | Medical image fusion method |
| CN114841903B (en) * | 2022-07-05 | 2022-09-09 | 黄海造船有限公司 | Ship body surface corrosion degree evaluation method based on image processing |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1405734A (en) * | 2002-10-28 | 2003-03-26 | 武汉大学 | Method for reinforcing edge of medical picture |
| US6704591B2 (en) * | 2002-04-23 | 2004-03-09 | Fuji Photo Film Co., Ltd. | Method, apparatus and program for restoring phase information |
| US20040228511A1 (en) * | 2003-05-14 | 2004-11-18 | Jean Lienard | Method and apparatus for setting the contrast and brightness of radiographic images |
| CN1729481A (en) * | 2002-12-18 | 2006-02-01 | 皇家飞利浦电子股份有限公司 | Method of processing an input image by means of multi-resolution |
-
2006
- 2006-07-26 CN CNB2006100618796A patent/CN100410969C/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6704591B2 (en) * | 2002-04-23 | 2004-03-09 | Fuji Photo Film Co., Ltd. | Method, apparatus and program for restoring phase information |
| CN1405734A (en) * | 2002-10-28 | 2003-03-26 | 武汉大学 | Method for reinforcing edge of medical picture |
| CN1729481A (en) * | 2002-12-18 | 2006-02-01 | 皇家飞利浦电子股份有限公司 | Method of processing an input image by means of multi-resolution |
| US20040228511A1 (en) * | 2003-05-14 | 2004-11-18 | Jean Lienard | Method and apparatus for setting the contrast and brightness of radiographic images |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101430787B (en) * | 2008-11-14 | 2012-07-04 | 西安交通大学 | Digital sternum heart calcification image intensification method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1889125A (en) | 2007-01-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100410969C (en) | Medical radiation image detail enhancing method | |
| Laine et al. | Mammographic feature enhancement by multiscale analysis | |
| Chen et al. | Thoracic low-dose CT image processing using an artifact suppressed large-scale nonlocal means | |
| Raj et al. | Computer aided detection of brain tumor in magnetic resonance images | |
| CN100461218C (en) | Method for enhancing medical image with multi-scale self-adaptive contrast change | |
| DE60208970T2 (en) | Method for improving the contrast of digital portal images | |
| US6956975B2 (en) | Method for improving breast cancer diagnosis using mountain-view and contrast-enhancement presentation of mammography | |
| EP1344188B1 (en) | Method for optimizing visual display of enhanced digital images | |
| US5963676A (en) | Multiscale adaptive system for enhancement of an image in X-ray angiography | |
| Zhao et al. | A new approach for medical image enhancement based on luminance-level modulation and gradient modulation | |
| CN106683061B (en) | A method of medical image is enhanced based on modified multiple dimensioned retinex algorithm | |
| CN101201937A (en) | Digital image enhancement method and device based on wavelet restruction and decompose | |
| CN104091309B (en) | Balanced display method and system for flat-plate X-ray image | |
| CN107845079A (en) | 3D shearlet medicine CT video denoising methods based on compact schemes | |
| CN106530236A (en) | Medical image processing method and system | |
| CN102637292B (en) | Image processing method and device | |
| CN103226815A (en) | Low dose CT image filtering method | |
| Xia et al. | Deep residual neural network based image enhancement algorithm for low dose CT images | |
| Aosiman et al. | Medical image enhancement algorithm based on histogram equalization and dyadic wavelet transform | |
| Outtas et al. | Subjective and objective evaluations of feature selected multi output filter for speckle reduction on ultrasound images | |
| Tirupal et al. | On the use of UDWT and fuzzy sets for medical image fusion | |
| Wang et al. | Self-fusion Simplex Noise-based Diffusion Model for Self-supervised Low-dose Digital Radiography Denoising | |
| CN115578285B (en) | Mammary gland molybdenum target image detail enhancement method and system | |
| US20240169610A1 (en) | Label-free adaptive ct super-resolution reconstruction method, system and device based on generative network | |
| US20220335602A1 (en) | Method and system for image normalisation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| ASS | Succession or assignment of patent right |
Owner name: SHENZHEN LANYUN INDUSTRY CO.,LTD. Free format text: FORMER OWNER: JIAYITONG MEDICAL SCIENCE + TECHNOLOGY CO., LTD., SHENZHEN CITY Effective date: 20070323 |
|
| C41 | Transfer of patent application or patent right or utility model | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20070323 Address after: Futian District King Road Shenzhen city Guangdong province 518000 No. 81 country garden E building 6 floor Applicant after: Landwind Co., Ltd. Address before: Shenzhen City, Guangdong Province, 518034 King Road No. 81 Bi garden E building 6 floor Applicant before: Jiayitong Medical Science & Technology Co., Ltd., Shenzhen City |
|
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| ASS | Succession or assignment of patent right |
Owner name: SHENZHEN LANDWIND NETWORK CO., LTD. Free format text: FORMER OWNER: LANYUN INDUSTRY CO., LTD.;SHENZHEN CITY Effective date: 20120810 |
|
| C41 | Transfer of patent application or patent right or utility model | ||
| COR | Change of bibliographic data |
Free format text: CORRECT: ADDRESS; FROM: 518000 SHENZHEN, GUANGDONG PROVINCE TO: 518034 SHENZHEN, GUANGDONG PROVINCE |
|
| TR01 | Transfer of patent right |
Effective date of registration: 20120810 Address after: Futian District King Road Shenzhen city Guangdong province 518034 green garden building room 410 (only office) Patentee after: Shenzhen Landwind Network Co., Ltd. Address before: Futian District King Road Shenzhen city Guangdong province 518000 No. 81 country garden E building 6 floor Patentee before: Landwind Co., Ltd. |
|
| C56 | Change in the name or address of the patentee |
Owner name: LANWANG TECHNOLOGY CO., LTD. Free format text: FORMER NAME: SHENZHEN LANDWIND NETWORK CO., LTD. |
|
| CP03 | Change of name, title or address |
Address after: Nanshan District Xueyuan Road in Shenzhen city of Guangdong province 518000 No. 1001 Nanshan Chi Park A3 building 6 floor Patentee after: LANWON SCIENCE & TECHNOLOGY CO., LTD. Address before: Futian District King Road Shenzhen city Guangdong province 518034 green garden building room 410 (only office) Patentee before: Shenzhen Landwind Network Co., Ltd. |
|
| PE01 | Entry into force of the registration of the contract for pledge of patent right |
Denomination of invention: Medical radiation image detail enhancing method Effective date of registration: 20150515 Granted publication date: 20080813 Pledgee: Shenzhen high tech investment and financing Company limited by guarantee Pledgor: LANWON SCIENCE & TECHNOLOGY CO., LTD. Registration number: 2015990000374 |
|
| PLDC | Enforcement, change and cancellation of contracts on pledge of patent right or utility model | ||
| C56 | Change in the name or address of the patentee | ||
| CP01 | Change in the name or title of a patent holder |
Address after: Nanshan District Xueyuan Road in Shenzhen city of Guangdong province 518000 No. 1001 Nanshan Chi Park A3 building 6 floor Patentee after: LANWON TECHNOLOGY CO., LTD. Address before: Nanshan District Xueyuan Road in Shenzhen city of Guangdong province 518000 No. 1001 Nanshan Chi Park A3 building 6 floor Patentee before: LANWON SCIENCE & TECHNOLOGY CO., LTD. |
|
| PC01 | Cancellation of the registration of the contract for pledge of patent right |
Date of cancellation: 20160629 Granted publication date: 20080813 Pledgee: Shenzhen high tech investment and financing Company limited by guarantee Pledgor: LANWON TECHNOLOGY CO., LTD. Registration number: 2015990000374 |
|
| PLDC | Enforcement, change and cancellation of contracts on pledge of patent right or utility model | ||
| PM01 | Change of the registration of the contract for pledge of patent right |
Change date: 20160629 Registration number: 2015990000374 Pledgor after: LANWON TECHNOLOGY CO., LTD. Pledgor before: LANWON SCIENCE & TECHNOLOGY CO., LTD. |