CN100461218C - Method for enhancing medical image with multi-scale self-adaptive contrast change - Google Patents
Method for enhancing medical image with multi-scale self-adaptive contrast change Download PDFInfo
- Publication number
- CN100461218C CN100461218C CNB2007100676936A CN200710067693A CN100461218C CN 100461218 C CN100461218 C CN 100461218C CN B2007100676936 A CNB2007100676936 A CN B2007100676936A CN 200710067693 A CN200710067693 A CN 200710067693A CN 100461218 C CN100461218 C CN 100461218C
- Authority
- CN
- China
- Prior art keywords
- image
- coefficient
- level
- detail
- contrast
- 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.)
- Expired - Fee Related
Links
Images
Abstract
A medical image intensifying method of multi-dimension adaptive contrast transform includes decomposing medical image to be image set being set with resolution to be decreased gradually and being arranged in pyramid from, regulating the layering coefficient obtained from decomposition to intensify contrast of each layer image and local region at detailed gradation image, resynthesizing each detailed gradation image with coefficient being regulated to be image of original image being intensified.
Description
Technical field
The invention belongs to image processing method, relate to a kind of image enhancement processing method, particularly a kind of medical image (is comprised the x ray image, the CT scan image) enhancement process method, be used to strengthen bone, internal organs, the medical image enhancement method of details such as soft tissue and focus and marginal information.
Background technology
X ray (being called for short the X line) has the ability that penetrates object, and sensitization, fluorescence and ionizing event and thermal effect are arranged, and also tissue of biological cells is had inhibition, damage even downright bad biological effect.X line medical imaging device is to penetrate each physiological tissue position of human body on the path to the accumulated value imaging of x ray absorption by the x ray.Each physiological tissue is superimposed in the human body, after the tissue owing to uptake is little and the X ray uptake is big of portion of tissue structure superposes, just can't show on the X ray medical image.
Developing rapidly owing to subjects such as physics, electronics and computing machines after the seventies in 19 century, X line machine is further developed on the one hand, technology such as anti-scattering resistant grating, contrast preparation and image amplifier have been invented, many on the other hand novel X line diagnostic imaging new technologies are arisen at the historic moment, as the auxiliary tomography (X line CT) of X computer on line, digital subtraction angiography (DSA), Computed Radiography (CR) and digital radiation photography (DR) etc.Computed Radiography and digitized processing and traditional X-ray radiophotography have the following advantages: (1) can use digitizing window/grade to handle and compensate automatically when under-exposure or overexposure, and depth of exposure no longer is vital.But this does not also mean that, reducing dosage does not have defective: when exposure is in a lower magnitude, noise will be more outstanding.(2) Computed Radiography and digitized processing can allow local contrast to strengthen.But the phosphorus particle in the system imaging process in the image plate makes the x ray have scattering, when passing the image plate deep, also exist scattering with the laser of laser scanner in the scanning process, thereby make image blurring, reduced image resolution ratio, influence picture quality, thereby influence the diagnosis accuracy.So medical image need carry out image enhancement processing, strengthens bone in the image, internal organs, details and marginal informations such as soft tissue and focus.
The treatment technology of medical image enhancing at present mainly comprises following three kinds:
1, the method for contrast enhancing
Contrast enhancement technique can carry out window/grade adjustment to image.But if near the big zone of grey scale change, fine feature still can not show, and so also requires further improvement the contrast of image.
2, the method for self-adapting histogram equilibrium
Smooth imagery zone use steep gradient curve and in correspondence has the zone of broad tonal range the soft gradient curve of application, the self-adapting histogram equilibrium technology is enhancing contrast ratio by this method just.Self-adaptation Nogata balancing technique can be on each position of image, and the shape of adjustment curve distributes to adapt to local gray level automatically.
3, the method for ambiguity correction
This special space frequency strip filtering technique is optionally strengthened the characteristics of image in the image particular range.Two kinds of technology are arranged equally, here: the edge strengthens and amplitude compression.The edge strengthens: if just emphasize finegrained structure and edge, usually with reference to " edge enhancing " technology.The characteristics of edge enhancing technique are that the diameter of filter center is little.Amplitude compression: during for example shoulder and lower limb detect, if because dominant gray scale overlay image, to cause feature little or medium image to become seeing not clear, and use spatial frequency filtering technology so and can obtain effect preferably with big central diameter.Strengthen grey scale change significantly, minimizing is caused the dynamic range of contrast improvement.The amplitude compression technology equally also can be regarded the compression of dynamic range as.
Existing image enhancement technique all is based on characteristic dimension, has certain defective.Operation all is certain space frequency strip of regulating by the decision of filter center diameter.Edge enhancement filter can be emphasized finegrained image detail.As a result, just see as the middle-sized soft images such as little details of dimness not clear, and may be very fuzzy.Selecting best central diameter for a kind of specific type of detection is not that part is easy to thing.For same image, under the different ratio (even very approaching), have many critical structures.If emphasize the image of certain size, the effect that obtains so is similar to image fine-feature under covering up another size.Another problem about the bar band filter is near high-contrast edges, and for example the edge of bone interface tissue or metal implant has wavy image to generate in the image.These wavy images have harmful effect to doctor's diagnosis; In some cases, these images can be used as certain pathology and prove that in the other situation, these wavy images may have been hidden some trickle pathologies on normal film.Many improvement technology about this basic local enhancement techniques are at how suppressing these wavy images to be born specially.
Summary of the invention
The objective of the invention is at the deficiencies in the prior art, propose a kind ofly can satisfy the subjective requirement of human eye vision preferably, strengthen effect and be better than traditional single two time scales approach, and operation time short multi-scale self-adaptive contrast variation medical image enhancement method.
In order to realize above purpose, the inventive method mainly may further comprise the steps:
1, utilizes the laplacian pyramid decomposition transform, medical image is resolved into the image collection that resolution that Pyramid arranges progressively reduces, be the details that different level of detail images is represented different scale respectively, a series of level of detail image coefficient value has been expressed the details on the yardstick separately.
2, the lamination factor that decomposition is drawn adjusts, and comprises
A) adopt contrast variation's method to strengthen the contrast of every tomographic image integral body, the contrast variation does mapping by nonlinear curve and handles, and the function representation of this curve is
Wherein x represents the coefficient on the level of detail image, and promptly exploded view looks like to obtain the coefficient of each layer edge image, and y is the coefficient of level of detail image after the corresponding conversion;
B) adopting the method for fringing coefficient adjustment and dynamic range compression to strengthen the contrast in level of detail image local zone, specifically is in the level of detail image border when faint, and the level of detail image coefficient multiply by amplification factor ae
kAdjust the coefficient of small scale, when the level of detail gradation of image was crossed over the tonal range of full images, the layered image coefficient multiply by compressibility factor al
k
Amplification factor ae
k, computing method:
Compressibility factor al
kComputing method:
For obtaining effect preferably: n
eGet 3, n
lGet 5, f
eGet 1.7, f
lGet 1.4.
3, utilize laplacian pyramid to decompose inverse transformation, with each level of detail image image after the composite artwork image intensifying again of adjusting behind the coefficient.
The present invention compares with existing many medical image algorithms, has following characteristics:
1, details is amplified
The common example selectivity that the multiple dimensioned method that being used for contrast increases has been abandoned based on characteristic dimension strengthens.Replace, multiple dimensioned method has been brought the notion that details is amplified or details strengthens.Characteristics of image, for example edge, institutional framework, spot or can be enhanced or desalinate than macrostructure.If be reinforced, just these characteristics of image have good contrast, if desalinated, the contrast of these characteristics of image will be lower.The standard of this judgement and the size of characteristics of image or diameter (in mm or pixel) are irrelevant.If strengthen faint tiny characteristics of image, can improve the identifiability of image effectively, but weaken in the image identifiability of clear part simultaneously; And the latter's reinforcement can not influence other parts.Here it is, and multiple dimensioned method contrast is strengthened the ultimate principle of technology: the equilibrium of contrast and characteristic dimension are irrelevant.
2, the difference relevant with the space frequency strip filtering technique
2 differences between this method and the space frequency strip filtering technique: (a) band filtering must be transferred to the peculiar scope of characteristic dimension, and this method does not have such requirement; (b) different with band filtering, this method partly is a cost with distinct image, strengthens faint tiny characteristics of image.
Core concept of the present invention is partly to be cost with distinct image, strengthens faint tiny characteristics of image.The multiple dimensioned contrast enhancement process that the present invention proposes is a kind of very effective medical image enhancement method, and the low contrast regions visibility of size can be greatly improved arbitrarily, and can not produce the grain effect.The inventive method can satisfy the subjective requirement of human eye vision preferably, strengthen effect and be better than traditional single two time scales approach, and operation time is short, and this method has broad application prospects on the enhancement process of medical image, can be used as the pre-service of graphical analysis.
Description of drawings
Fig. 1 is image pyramid synoptic diagram among the present invention;
Fig. 2 is laplacian pyramid decomposition transform and inverse transformation synoptic diagram among the present invention.
Embodiment
Describe the quick Enhancement Method of medical image that the present invention is based on the multi-scale self-adaptive contrast variation in detail below in conjunction with accompanying drawing.
The inventive method mainly comprises 3 steps: (1) laplacian pyramid decomposition transform; (2), comprise that the contrast variation strengthens the contrast of integral image and the contrast that fringing coefficient adjustment, dynamic range compression are improved regional area to the adjustment of the every layer coefficients of level of detail; (3) laplacian pyramid decomposes inverse transformation.
One by one each step is described below.
Step 1: laplacian pyramid decomposition transform
Image pyramid (as shown in Figure 1) with multiple dimensioned come interpretation of images a kind of effectively but the simple structure of notion is exactly an image pyramid.Image pyramid is used for machine vision and compression of images at first, and the pyramid of piece image is the image collection that the resolution of a series of Pyramids arrangements progressively reduces.Pyramidal bottom is that the high resolving power of pending image is represented, and the top is the approximate of low resolution.When move on pyramidal upper strata, size and resolution just reduce.Because the size of base level J is 2J * 2J or N * N (J=log2N), the size of intergrade j is 2j * 2j, wherein 0≤j≤J.Complete pyramid is made up of J+1 stage resolution ratio, and by 2J * 2J to 20 * 20, but most of pyramid has only the P+1 level, j=J-P wherein ..., J-2, J-1, J and 1≤P≤J.That is to say, limit them usually and only use the P level to reduce the size of original image approximate value.
The basic idea of multiple dimensioned enhancing is to become different level of detail images to represent the details of different scale and approximate respectively picture breakdown, then directly improves contrast on these level of detail images.The picture breakdown of multiple dimensioned pyramid structure algorithm is according to flow process as shown in Figure 2.Wherein, LP represents low-pass filtering.The LP low-pass filter adopts 5 * 5 specific low pass template filtering, and filter template is as follows:
As Fig. 2, original image is obtained ground floor detail pictures g by filtering sampling
1, with g
1Obtain g by filtering sampling
2, obtain g by iterative manner successively
3, g
4G
LWith i layer detail pictures g
iImage that obtains behind the filtering interpolation and last layer L-1 image carry out algebraically and subtract each other and can obtain i layer residual image b
i, method can get one group of residual image b that progression is L-1 like this
0, b
1, b
2, b
3B
L-1Earlier original image is made down-sampling, middle result remakes interpolation and gets back to original image size, and smoothing processing before doing sampling and after the interpolation, and the image that draws deducts from former figure by pixel, and the difference that at this moment obtains is exactly multiple dimensioned tower shape layer coefficients.The level of detail image of back is done similar calculating.Like this, size of images reduces by half on two-dimensional direction one by one, up to the size of down-sampled images less than template size.
Details is fewer and feweri in decomposable process, and a series of lamination factor difference has been expressed the details on the yardstick separately, in frequency domain, each yardstick just corresponding each frequency range of original signal spectrum, the frequency range of each layering is that lap is arranged.
Step 2: to the adjustment of the every layer coefficients of details
Multiple dimensioned image intensifying will adjust to the level of detail image coefficient that decomposition draws, and these adjustment comprise contrast compensation and fringing coefficient adjustment, dynamic range compression operation.In the edge image of each layer, the little trickle details of coefficient representative, they need be enhanced the visibility that improves corresponding details, those big coefficients are represented those stronger edges simultaneously, they influence whole dynamic range to a great extent, suitable is pressed, and also can not influence their visibility, has but improved the overall contrast of image.
A. the contrast variation strengthens the contrast of integral image
Contrast compensation is done mapping by a nonlinear curve and is handled, and curve is as follows:
Wherein, x represents the coefficient on the level of detail image, and promptly exploded view looks like to obtain the coefficient of each layer edge image, and y is the coefficient of layered image after the corresponding conversion.Earlier x is standardized on [1,1] scope, the dynamic range that factor a is used for adjusting figure output as a result is consistent with original image.The degree of crook of coefficient p control curve, in order to meet the demands: (1) all edges all are enhanced; (2) the enhancing amplitude at the edge a little less than the signal then requires p<1 greater than the strong edge of signal.Multiple dimensioned contrast compensation strengthens all detailed information of full figure, rather than just on former figure yardstick.The edge of sharpening has strengthened, and the zone of low contrast also is enhanced simultaneously, and like this, soft tissue also can be faintly visible when bone strengthens.As can be seen, strengthen with respect to single yardstick edge from example, the multi-scale image contrast compensation can obtain good visual effect does not have " grain effect " simultaneously.
B. fringing coefficient adjustment, dynamic range compression are improved the contrast of regional area.
Contrast compensation is the basic model of enhancement process, just can so operate in most of inspection areas, but the image border, zone (as brothers) that has is very faint, just needs to strengthen emphatically, multiply by an amplification factor ae on the fringing coefficient of corresponding scale
kOn the contrary, the zone that has (as shoulder, belly) gray scale is crossed over the tonal range of full figure, needs to select suitable compressibility factor al
k, these regional tonal ranges are compressed, otherwise are influenced the demonstration in other zones.
Multiply by ae
kAdjust the coefficient of small scale and be fringing coefficient adjustment, ae
kShown in the following formula:
Wherein, f
e(f
e1) control adjusts the amplitude of parameter, n
eThe level of detail image number of plies that appointment needs fringing coefficient to adjust.When the number of plies increases, adjust coefficient ae
kWith
Speed successively decrease, and only at the high-end n of frequency spectrum
eCarry out ae in the follow-up layer in the layer
kKeep 1.The adjustment of this gradual change coefficient has been arranged, just had " the grain effect " in the violent zone of grey scale change minimized.
Equally, multiply by al
kOn large scale, be dynamic range compression al
kShown in the following formula:
Wherein, L is the number of plies sum of picture breakdown, n
lAppointment needs the number of plies of gray scale dynamic range compression, al
kWhen the number of plies increases according to by 1 the beginning with
Speed successively decrease.The amplitude of scope compression is by parameter f
l(f
l〉=1) controls.
In the experiment, choose empirical value and obtain effect preferably: n
eGet 3, n
lGet 5, f
eGet 1.7, f
lGet 1.4, draw ae
kAnd al
kIn the number of plies is distribution on 10 tomographic images, as shown in table 1.
Each layer of table 1 adjusted coefficient
Because multiple dimensioned coefficient adjustment contrast variation's algorithm for image enhancement is detailed information to be strengthened on a plurality of yardsticks at entire image, and consider that dynamic range is excessive on the faint and large scale of local edge on the small scale, thereby not only on the contrast of general image, be improved, and on subrange, also obtain desirable effect.
Step 3: laplacian pyramid decomposes inverse transformation
In fact inverse transformation process is exactly an inverse process that decomposes, as Fig. 2.Obtain new detail pictures g with carrying out sum operation with the residual image of last layer behind the detail pictures gL filtering interpolation of the bottom
L-1', with the method, obtain g successively
L-2', g
L-3' ... g
0'.Work as g
0' carry out sum operation with b0 ' again behind the filtering interpolation and obtain final needed image.Promptly on out to out, i.e. approximate image from drawing at last, interpolation is adjusted image and is gone on the size of last layer, does addition then and when the detail signal of anterior layer.Interpolation and addition repeat up to extensive palinspastic map size.If the interpolation in sampling in the decomposable process and the inverse transformation is corresponding, the result should be consistent with former figure.
Claims (1)
1. multi-scale self-adaptive contrast variation's medical image enhancement method is characterized in that this method mainly may further comprise the steps:
(1) utilizes the laplacian pyramid decomposition transform, medical image is resolved into the image collection that resolution that Pyramid arranges progressively reduces, be the details that different level of detail images is represented different scale respectively, a series of level of detail image coefficient value has been expressed the details on the yardstick separately;
(2) lamination factor that decomposition is drawn adjusts, and comprises
A. adopt contrast variation's method to strengthen the contrast of every tomographic image integral body, the contrast variation does mapping by nonlinear curve and handles, and the function representation of nonlinear curve is
Wherein x represents the coefficient on the level of detail image, be the coefficient that exploded view looks like to obtain each layer edge image, y is the coefficient of level of detail image after the corresponding conversion, and the dynamic range that factor a is used for adjusting figure output as a result is consistent with original image, the degree of crook of coefficient p control curve, p<1;
B. adopting the method for fringing coefficient adjustment and dynamic range compression to strengthen the contrast in level of detail image local zone, specifically is in the level of detail image border when faint, and the level of detail image coefficient multiply by amplification factor ae
kAdjust the coefficient of small scale, when the level of detail gradation of image was crossed over the tonal range of full images, the layered image coefficient multiply by compressibility factor al
k
Amplification factor ae
k, computing method:
Compressibility factor al
kComputing method:
(3) utilize laplacian pyramid to decompose inverse transformation, with each level of detail image image after the composite artwork image intensifying again of adjusting behind the coefficient.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CNB2007100676936A CN100461218C (en) | 2007-03-29 | 2007-03-29 | Method for enhancing medical image with multi-scale self-adaptive contrast change |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CNB2007100676936A CN100461218C (en) | 2007-03-29 | 2007-03-29 | Method for enhancing medical image with multi-scale self-adaptive contrast change |
Publications (2)
Publication Number | Publication Date |
---|---|
CN101030298A CN101030298A (en) | 2007-09-05 |
CN100461218C true CN100461218C (en) | 2009-02-11 |
Family
ID=38715623
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CNB2007100676936A Expired - Fee Related CN100461218C (en) | 2007-03-29 | 2007-03-29 | Method for enhancing medical image with multi-scale self-adaptive contrast change |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN100461218C (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101668196B (en) * | 2009-09-25 | 2012-02-08 | 西安电子科技大学 | Low code rate image compression method based on down sampling and interpolation |
Families Citing this family (28)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP2071513A1 (en) * | 2007-12-10 | 2009-06-17 | Agfa HealthCare NV | Method of generating a multiscale contrast enhanced image |
US8086065B2 (en) * | 2008-04-02 | 2011-12-27 | Himax Imaging, Inc. | Apparatus and method for contrast enhancement |
CN101727658B (en) * | 2008-10-14 | 2012-12-26 | 深圳迈瑞生物医疗电子股份有限公司 | Image processing method and device |
EP2449526A1 (en) | 2009-06-29 | 2012-05-09 | Thomson Licensing | Zone-based tone mapping |
WO2011008239A1 (en) * | 2009-06-29 | 2011-01-20 | Thomson Licensing | Contrast enhancement |
CN102306376B (en) * | 2009-11-03 | 2014-12-17 | 蒋慧琴 | Method for adaptive medical image enhancement |
CN102646269B (en) * | 2012-02-29 | 2015-09-23 | 中山大学 | A kind of image processing method of laplacian pyramid and device thereof |
JP5633550B2 (en) * | 2012-09-05 | 2014-12-03 | カシオ計算機株式会社 | Image processing apparatus, image processing method, and program |
US8995783B2 (en) * | 2012-09-19 | 2015-03-31 | Qualcomm Incorporation | System for photograph enhancement by user controlled local image enhancement |
TWI571762B (en) * | 2012-11-08 | 2017-02-21 | 國立台灣科技大學 | Real time image cloud system and management method |
CN103295191B (en) * | 2013-04-19 | 2016-03-23 | 北京航科威视光电信息技术有限公司 | Multiple scale vision method for adaptive image enhancement and evaluation method |
CN103325098A (en) * | 2013-07-02 | 2013-09-25 | 南京理工大学 | High dynamic infrared image enhancement method based on multi-scale processing |
CN104166974B (en) * | 2013-08-01 | 2015-05-13 | 上海联影医疗科技有限公司 | CT locating film image enhancing method and CT locating film image enhancing device |
CN103500442B (en) * | 2013-09-29 | 2017-06-06 | 华南理工大学 | X-ray image multi-scale detail enhancing method in integrated antenna package |
CN104574284A (en) * | 2013-10-24 | 2015-04-29 | 南京普爱射线影像设备有限公司 | Digital X-ray image contrast enhancement processing method |
CN103996168B (en) * | 2014-01-21 | 2017-02-01 | 公安部第一研究所 | X-ray safety inspection image enhancing method based on region self-adaptive processing |
CN105488762A (en) * | 2014-10-10 | 2016-04-13 | 杨大刚 | Image restoration method for digital dental panoramic machine |
CN106485665B (en) * | 2015-08-31 | 2019-01-08 | 辽宁开普医疗系统有限公司 | A kind of low dosage DR image processing method and its device |
BR112018013602A2 (en) | 2015-12-31 | 2019-04-09 | Shanghai United Imaging Healthcare Co., Ltd. | image processing methods and systems |
CN106127712B (en) * | 2016-07-01 | 2020-03-31 | 上海联影医疗科技有限公司 | Image enhancement method and device |
CN107256559B (en) * | 2017-06-01 | 2020-07-14 | 北京环境特性研究所 | Method for complex background suppression |
CN108648155B (en) * | 2018-05-07 | 2020-11-06 | 河北省科学院应用数学研究所 | Image enhancement method based on compressed domain and terminal equipment |
CN108629788B (en) * | 2018-05-08 | 2022-03-15 | 苏州大学 | Image edge detection method, device and equipment and readable storage medium |
CN108830915B (en) * | 2018-05-28 | 2023-03-31 | 牙博士医疗控股集团有限公司 | Method and device for realizing 3D (three-dimensional) simulation animation of oral cavity image |
CN109949233B (en) * | 2019-02-18 | 2022-12-13 | 深圳蓝影医学科技股份有限公司 | Method, system, device and storage medium for filtering scattered rays in X-ray image |
CN110740253B (en) * | 2019-04-01 | 2020-11-10 | 徐成晓 | Shooting mode self-adaptive switching method |
CN111292267B (en) * | 2020-02-04 | 2020-10-23 | 北京锐影医疗技术有限公司 | Image subjective visual effect enhancement method based on Laplacian pyramid |
CN112308793A (en) * | 2020-10-21 | 2021-02-02 | 淮阴工学院 | Novel method for enhancing contrast and detail of non-uniform illumination image |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5742892A (en) * | 1995-04-18 | 1998-04-21 | Sun Microsystems, Inc. | Decoder for a software-implemented end-to-end scalable video delivery system |
US5963676A (en) * | 1997-02-07 | 1999-10-05 | Siemens Corporate Research, Inc. | Multiscale adaptive system for enhancement of an image in X-ray angiography |
CN1283291A (en) * | 1997-12-31 | 2001-02-07 | 萨尔诺夫公司 | Apparatus and method for performing scalable hierarchical motion estimation |
CN1545062A (en) * | 2003-11-27 | 2004-11-10 | 上海交通大学 | Pyramid image merging method being integrated with edge and texture information |
CN1889125A (en) * | 2006-07-26 | 2007-01-03 | 深圳市嘉易通医疗科技有限公司 | Medical radiation image detail enhancing method |
-
2007
- 2007-03-29 CN CNB2007100676936A patent/CN100461218C/en not_active Expired - Fee Related
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5742892A (en) * | 1995-04-18 | 1998-04-21 | Sun Microsystems, Inc. | Decoder for a software-implemented end-to-end scalable video delivery system |
US5963676A (en) * | 1997-02-07 | 1999-10-05 | Siemens Corporate Research, Inc. | Multiscale adaptive system for enhancement of an image in X-ray angiography |
CN1283291A (en) * | 1997-12-31 | 2001-02-07 | 萨尔诺夫公司 | Apparatus and method for performing scalable hierarchical motion estimation |
CN1545062A (en) * | 2003-11-27 | 2004-11-10 | 上海交通大学 | Pyramid image merging method being integrated with edge and texture information |
CN1889125A (en) * | 2006-07-26 | 2007-01-03 | 深圳市嘉易通医疗科技有限公司 | Medical radiation image detail enhancing method |
Non-Patent Citations (2)
Title |
---|
基于视觉特性的多尺度对比度塔图像融合及性能评价. 张新曼,韩九强.西安交通大学学报,第38卷第4期. 2004 |
基于视觉特性的多尺度对比度塔图像融合及性能评价. 张新曼,韩九强.西安交通大学学报,第38卷第4期. 2004 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101668196B (en) * | 2009-09-25 | 2012-02-08 | 西安电子科技大学 | Low code rate image compression method based on down sampling and interpolation |
Also Published As
Publication number | Publication date |
---|---|
CN101030298A (en) | 2007-09-05 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN100461218C (en) | Method for enhancing medical image with multi-scale self-adaptive contrast change | |
Heinrich et al. | Residual U-net convolutional neural network architecture for low-dose CT denoising | |
CN108780571B (en) | Image processing method and system | |
CN100410969C (en) | Medical radiation image detail enhancing method | |
Veldkamp et al. | The value of scatter removal by a grid in full field digital mammography | |
McLoughlin et al. | Noise equalization for detection of microcalcification clusters in direct digital mammogram images | |
JP4895204B2 (en) | Image component separation device, method, and program, and normal image generation device, method, and program | |
CN101129266B (en) | X-ray image processing system | |
EP1743299B1 (en) | Method, computer program product and apparatus for enhancing a computerized tomography image | |
JP4919408B2 (en) | Radiation image processing method, apparatus, and program | |
CN101779962B (en) | Method for enhancing medical X-ray image display effect | |
CN105447866A (en) | X-ray chest radiograph bone marrow suppression processing method based on convolution neural network | |
DE69922983T2 (en) | Imaging system and method | |
Irrera et al. | A flexible patch based approach for combined denoising and contrast enhancement of digital X-ray images | |
CN104574284A (en) | Digital X-ray image contrast enhancement processing method | |
CN104091309B (en) | Balanced display method and system for flat-plate X-ray image | |
Kelm et al. | Optimizing non-local means for denoising low dose CT | |
Bhairannawar | Efficient medical image enhancement technique using transform HSV space and adaptive histogram equalization | |
Yang et al. | Automatic tissue classification for high-resolution breast CT images based on bilateral filtering | |
CN113205461B (en) | Low-dose CT image denoising model training method, denoising method and device | |
Xia et al. | Dedicated breast computed tomography: Volume image denoising via a partial‐diffusion equation based technique | |
JP2013119021A (en) | X-ray ct device and image processing method | |
Ren et al. | Deep-learning-based denoising of X-ray differential phase and dark-field images | |
Abir | Contrast enhancement of digital mammography based on multi-scale analysis | |
Huddin et al. | Enhancement techniques for MRI human spine images |
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 | ||
C17 | Cessation of patent right | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20090211 Termination date: 20120329 |