CN110211143A - A kind of medical image analysis method based on computer vision - Google Patents
A kind of medical image analysis method based on computer vision Download PDFInfo
- Publication number
- CN110211143A CN110211143A CN201910496499.2A CN201910496499A CN110211143A CN 110211143 A CN110211143 A CN 110211143A CN 201910496499 A CN201910496499 A CN 201910496499A CN 110211143 A CN110211143 A CN 110211143A
- Authority
- CN
- China
- Prior art keywords
- image
- hospital
- treated
- pixel
- threshold
- 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
- 238000003703 image analysis method Methods 0.000 title claims abstract description 14
- 238000012545 processing Methods 0.000 claims abstract description 20
- 238000000034 method Methods 0.000 claims abstract description 16
- 238000003709 image segmentation Methods 0.000 claims abstract description 8
- 230000009466 transformation Effects 0.000 claims abstract description 6
- 230000005540 biological transmission Effects 0.000 claims abstract description 4
- 230000011218 segmentation Effects 0.000 claims description 17
- 238000001228 spectrum Methods 0.000 claims description 14
- 238000006243 chemical reaction Methods 0.000 claims description 8
- 238000001914 filtration Methods 0.000 claims description 8
- 238000007781 pre-processing Methods 0.000 claims description 8
- 238000005516 engineering process Methods 0.000 claims description 7
- 230000002452 interceptive effect Effects 0.000 claims description 7
- 238000003745 diagnosis Methods 0.000 claims description 5
- 238000004445 quantitative analysis Methods 0.000 claims description 5
- 238000005192 partition Methods 0.000 claims description 4
- 230000008569 process Effects 0.000 claims description 4
- 238000012360 testing method Methods 0.000 claims description 4
- 230000009471 action Effects 0.000 claims description 3
- 238000001514 detection method Methods 0.000 claims description 3
- 230000002708 enhancing effect Effects 0.000 claims description 3
- 238000013507 mapping Methods 0.000 claims description 3
- 238000000926 separation method Methods 0.000 claims description 3
- 206010003694 Atrophy Diseases 0.000 claims description 2
- 238000013528 artificial neural network Methods 0.000 claims description 2
- 230000037444 atrophy Effects 0.000 claims description 2
- 230000009977 dual effect Effects 0.000 claims description 2
- 238000003708 edge detection Methods 0.000 claims description 2
- 238000000605 extraction Methods 0.000 claims description 2
- 230000003993 interaction Effects 0.000 claims description 2
- 230000001629 suppression Effects 0.000 claims 1
- 238000002059 diagnostic imaging Methods 0.000 abstract description 4
- 230000008859 change Effects 0.000 abstract description 2
- 238000004321 preservation Methods 0.000 abstract description 2
- 230000008030 elimination Effects 0.000 abstract 1
- 238000003379 elimination reaction Methods 0.000 abstract 1
- 238000007726 management method Methods 0.000 description 6
- 230000007812 deficiency Effects 0.000 description 2
- 235000013399 edible fruits Nutrition 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000004171 remote diagnosis Methods 0.000 description 2
- 238000007794 visualization technique Methods 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000002591 computed tomography Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000000155 isotopic effect Effects 0.000 description 1
- 238000002595 magnetic resonance imaging Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000000877 morphologic effect Effects 0.000 description 1
- 210000000056 organ Anatomy 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 230000005855 radiation Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 238000002604 ultrasonography Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
- 238000012800 visualization Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/40—Image enhancement or restoration using histogram techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20024—Filtering details
- G06T2207/20032—Median filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20048—Transform domain processing
- G06T2207/20064—Wavelet transform [DWT]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
Abstract
The present invention is claimed a kind of medical image analysis method based on computer vision and completes pretreatment work to the medical image sequences of input, mainly includes greyscale transformation, the elimination of smooth and noise realizes the image segmentation to pretreated medical image.It realizes the digital management to medical image, reduces the difficulty and cost of management, and management more standardizes, and is more suitable the inquiry and retrieval of image；Image is stored in the form of digitized, and the information of image is completely kept down, and window position can change with window width, and can use some image processing techniques and be pocessed；The preservation of image is carried out using disk, it is no longer necessary to which film facilitates the transmission of data, and data are less likely to be damaged, go bad, lose；Using dicom standard, it can be compatible with the data format of most of medical imaging device at present, greatly improve the utilization rate and circulation of data.
Description
Background technique
The present invention relates to technical field of image processing, in particular to a kind of medical image analysis side based on computer vision
Method is suitable for target following, target identification, video monitoring, video conference, biomedicine, information security, remote sensing telemetering, space flight
Images match in aviation or video multimedia.
Background technique
Since the seventies, with computed tomography, the appearance of the medical imaging devices such as Magnetic resonance imaging with
Using the two-dimensional digital tomographic sequence of the available human body of people and its internal.Radiation technician can pass through observation
Image, to find and diagnose the state of an illness, but this diagnostic mode has following main deficiency: (1) comparison of radioscopic image (containing CT)
Degree is very low, is easy mistaken diagnosis (2) core isotopic image and ultrasound image is relatively rough, resolution ratio is very low；(3) from one group of two dimensional image
It is difficult and inaccurate to conceive three-dimensional structure.For the accuracy and science for improving medical diagnosis and treatment planning, need two
Dimension tomographic sequence is transformed into the image with intuitive stereoscopic effect, shows the three-dimensional structure and form of human organ, from
And several anatomical informations that can not be obtained with traditional means are provided, and provide visual interactive hand for further simulated operation
Section.Reconstruction of medical images and visualization technique are exactly to propose in this context, which just obtains after proposition
It has arrived a large amount of research and has been widely applied.
During the display of 3-D image, doctor must complete the use to medical data by display instrument, so
And traditional three-dimensional visualization method can not show that this is just by initial data is distortionless on common display instrument
Doctor brings difficulty using these medical images, while also counteracting medical image three-dimensional visualization in wide range
Interior application.
Although current computer technology has obtained very big development, arithmetic speed is many relative to improving in the past,
Need processed data volume sometimes very huge in image geometry matching, therefore, the calculating speed of computer is again without reaching
To real-time target.And due to different illumination and noise etc., the image of identical content can be made also to have difference, to improve
The matched difficulty of image geometry.Therefore, the image geometry matching algorithm that matching precision is high, real-time is good, at medical image
The fields such as reason, remote sensing image processing, pattern-recognition have very important application value and theory significance.
The digital management of Medical image, the difficulty and higher cost of management, management is lack of standardization, is not suitable for image
Inquiry and retrieval；Storage needs film, it has not been convenient to the transmission of data, and also data are easily damaged, go bad, lose；It cannot be compatible with
The data format of major part medical imaging device at present, is managed various medical images, cannot adequately utilize internet,
Realize data sharing, remote diagnosis mostly is held a consultation, and resource, the utilization rate and circulation of data can not be provided for medical educational
Difference.
Summary of the invention
In view of the deficiencies of the prior art, the present invention provides medical image analysis method based on computer vision,
And there is certain versatility.
A kind of medical image analysis method based on computer vision is claimed in the present invention characterized by comprising
Patient treats treatment region image using the mobile phone for being equipped with the software and is acquired: if acquired image is askew
Tiltedly, obscure etc., mobile phone prompt patient re-shoots area image to be treated, obtains clear image；
Image is read in, and complete reading and conversion and the display of image and the Threshold segmentation of image: server terminal uses
Template matching algorithm based on shape feature and the detection algorithm based on SIFT feature are detected, and OpenCV vision class libraries is called
In packaged image processing operators, corresponding C++ code is completed in programming, and utilizes each control button in software interface
Realize the function of algorithm；
Mobile phone starts the system software and carries out image preprocessing, and carries out handset identity: image preprocessing completes image filter
Wave processing and the acquisition of target subgraph and its histogram equalization processing, are used to fix sufferer by include in medical image
Bracket or other supporter garbages delete, method is filled on erasing rubber institute route via with the gray value of background
Original gray value, to provide for medical diagnosis from qualitative to quantitative, more objective information passes through original DICOM data
The mapping ground mode that reconciles is converted into 8 BMP grayscale images, the contrast of enhancing output image；
Regional image information to be treated is obtained, then obtained information is issued hospital by mobile phone: image segmentation and automation
Quantitative analysis obtains the outer contoured features of target, on the basis of image segmentation, obtains target area longest and most wide respectively
Distance in include number of pixels, and in proportion ruler coefficient conversion at entity long width values, it is reconstructed obtain threedimensional model after, it is auxiliary
Functional module is helped mainly to complete the three-dimensional view angle map function to model；
According to interactive regional choice, show the tissue of different piece: hospital sends area image letter to be treated in real time
It ceases to patient, realizes quantitative background separation function using mouse action or keyboard input.
The present invention completes the pretreatment work to the medical image sequences of input, mainly includes greyscale transformation！It is smooth and
Noise is eliminated, and the image segmentation to pretreated medical image is realized.It proposes and is combined by calculating automatically with interactive mode
Threshold segmentation, morphological method carry out region finishing, realize the digital management to medical image, reduce the difficulty of management
Degree and cost, management more standardize, and are more suitable the inquiry and retrieval of image；Image is stored in the form of digitized, figure
The information of picture is completely kept down, and window position can change with window width, and be can use some image processing techniques and be subject to
Processing；Use disk to carry out the preservation of image, it is no longer necessary to film facilitates the transmission of data, and data be less likely to be damaged,
Rotten, loss；Using dicom standard, it can be compatible with the data format of most of medical imaging device at present, to various medicine shadows
As being managed；Internet can be adequately utilized, realizes data sharing, remote diagnosis mostly is held a consultation, while being also medical treatment religion
Provide many resources, greatly improves the utilization rate and circulation of data.
Detailed description of the invention
Technical solution in ord to more clearly illustrate embodiments of the present application, below will be to needed in the embodiment attached
Figure is briefly described, it should be understood that the following drawings illustrates only some embodiments of the application, therefore is not construed as pair
The restriction of range for those of ordinary skill in the art without creative efforts, can also be according to this
A little attached drawings obtain other relevant attached drawings.
Fig. 1 is a kind of work flow diagram of medical image analysis method based on computer vision according to the present invention.
Specific embodiment
Exemplary embodiments of the present disclosure are described in more detail below with reference to accompanying drawings.Although showing the disclosure in attached drawing
Exemplary embodiment, it being understood, however, that may be realized in various forms the disclosure without should be by embodiments set forth here
It is limited.On the contrary, these embodiments are provided to facilitate a more thoroughly understanding of the present invention, and can be by the scope of the present disclosure
It is fully disclosed to those skilled in the art.
It is a kind of workflow of medical image analysis method based on computer vision according to the present invention referring to attached drawing 1
Cheng Tu.
A kind of medical image analysis method based on computer vision is claimed in the present invention characterized by comprising
Patient treats treatment region image using the mobile phone for being equipped with the software and is acquired: if acquired image is askew
Tiltedly, obscure etc., mobile phone prompt patient re-shoots area image to be treated, obtains clear image；
Image is read in, and complete reading and conversion and the display of image and the Threshold segmentation of image: server terminal uses
Template matching algorithm based on shape feature and the detection algorithm based on SIFT feature are detected, and OpenCV vision class libraries is called
In packaged image processing operators, corresponding C++ code is completed in programming, and utilizes each control button in software interface
Realize the function of algorithm；
Mobile phone starts the system software and carries out image preprocessing, and carries out handset identity: image preprocessing completes image filter
Wave processing and the acquisition of target subgraph and its histogram equalization processing, are used to fix sufferer by include in medical image
Bracket or other supporter garbages delete, method is filled on erasing rubber institute route via with the gray value of background
Original gray value, to provide for medical diagnosis from qualitative to quantitative, more objective information passes through original DICOM data
The mapping ground mode that reconciles is converted into 8 BMP grayscale images, the contrast of enhancing output image；
Regional image information to be treated is obtained, then obtained information is issued hospital by mobile phone: image segmentation and automation
Quantitative analysis obtains the outer contoured features of target, on the basis of image segmentation, obtains target area longest and most wide respectively
Distance in include number of pixels, and in proportion ruler coefficient conversion at entity long width values, it is reconstructed obtain threedimensional model after, it is auxiliary
Functional module is helped mainly to complete the three-dimensional view angle map function to model；
According to interactive regional choice, show the tissue of different piece: hospital sends area image letter to be treated in real time
It ceases to patient, realizes quantitative background separation function using mouse action or keyboard input.
Preferably, the patient treats treatment region image image using the mobile phone for being equipped with the software and is acquired, and has
Body further include:
The collecting part of system is made of camera lens, camera and industrial personal computer etc.；
Collection process after being converted to picture signal, is transmitted by system camera photographic subjects background by USB2.0 interface
It is handled to processing server；
In image acquisition process, need to digitize image, generally using digitized image pick-up card at
Reason.
Preferably, described image is read in, and completes reading and conversion and the display of image and the Threshold segmentation of image, specifically
Further include:
It determines the partition threshold needed, partition threshold is divided into pixel compared with pixel value, determines a gray value, so
It will be greater than afterwards, be judged to object equal to the pixel of this gray value and indicated with a gray value, and the pixel of this gray value will be less than
Sentence, background and to be indicated with another gray value, or conversely, the gray scale size relation of object, background is exchanged, the knot of binaryzation
Fruit will be so that gray scale image becomes bianry image；
Determine threshold values be the key that segmentation, different threshold values may cause it is entirely different as a result, the number of threshold values and
Value is determined according to priori knowledge or test statistics data；
Pretreated subgraph is split using iterative threshold segmentation technology, and according to the segmentation knot in subgraph
Fruit is in original image；
The profile of target is extracted, automation quantitative analysis is carried out for computer and creates conditions；
According in segmentation result image, the shape feature of the consistency of feature and all subregion is commented in the same area
The quality of valence segmentation result.
Further, the mobile phone starts system software progress image preprocessing, and carries out handset identity, specifically also wraps
It includes:
Multistage median filtering and edge detecting technology is selected to complete the pretreatment to image, wherein multistage median filtering is used
In the pretreatment to full figure, for any pixel, the average gray value of pixel in its certain neighborhood is taken to substitute original pixel；
Using the adaptive weighted averaging filter method based on the ratio between filter window internal variance and mean value, filter window is covered
Before cover area executes median filtering operation, first the pixel in window is assigned according to the ratio size of variance and mean value different
Weight.Weighted median filtering weight coefficient expression formula are as follows:
ω (i, j)=[ω (k+1, k+1)-cd σ2/m]
The local window size of its median filtering used is (2k+1) * (2k+1), during ω (k+1, k+1, k+1) is in formula
Heart point weight；C is constant, by manual adjustment；D is the distance that point (i, j) arrives local window center；σ2It is local window with m
Variance and mean value, [x], which indicates to work as in bracket, takes immediate integer in the presence of result, if result is not present, takes zero.
In above formula, because of c, d, σ2It is all larger than with m or equal to 0, it is clear that the maximum weight of central point.In general, exist
Those relatively uniform regions, if being mutated, mainly caused by noise.In these regions, local variance very little, cd
σ2The value very little of/m, the weight of each pixel is roughly equal in region, is equal to general median filter, catastrophe point is gone
It removes.In the region that those include detailed information or boundary, local variance is very big, cd σ2The value of/m is very big, therefore pixel in region
Point weight is reduced rapidly with increasing with central point distance, so that the gray value near window center is retained, is reached reservation
The purpose of details.
In filter, the size of window determines noise removal capability, and constant c and central point weights omega (k+1, k+1, k+1) are certainly
Surely the ability on boundary and details is protected；Take central point weight 150, customized parameter c=7, filter window is 5 × 5, achieve compared with
For satisfied effect.
Edge detection further increases the differentiation of target and background for the processing to the subgraph comprising target area
Degree provides the image of high quality for segmentation, carries out convolution algorithm with input picture and Gaussian function and obtain filtered image, so
Amplitude and the direction for calculating image gradient using differential operator afterwards, then traverse whole image, carry out to gradient magnitude non-very big
Value inhibit, finally using dual thresholds method carry out edge extraction, more than higher thresholds pixel be edge, it is on the contrary then not
It is that, if testing result is greater than Low threshold but be less than high threshold, that just compares in the neighborhood territory pixel of this pixel either with or without super
The edge pixel for crossing high threshold, it is on the contrary then be not if there is that then should be edge.
Preferably, described to obtain regional image information to be treated, then obtained information is issued hospital by mobile phone, specific to go back
Include:
The rotation angle between area image to be treated and hospital's image is found out using Fourier transformation；
Area image to be treated and hospital's image are inputted first, and hospital's image and area image to be treated are to Same Scene
Institute has between certain rotation differential seat angle namely area image to be treated and hospital's image that there are one at image between the two
Fixed rotation differential seat angle；
Area image to be treated and hospital's image are transformed into Fourier frequency domain, obtain area image to be treated and doctor
Then the frequency spectrum of institute's image is treated the frequency spectrum for the treatment of region image and the Fourier frequency spectrum difference modulus value of hospital's image, is established
One, about the equation for rotating angle between two images, is found out the pass of area image frequency spectrum to be treated and hospital's image spectrum modulus value
System；
Rotation angle equation is transformed under polar coordinates again, it is corresponding under polar coordinates to obtain two modulus value of Fourier frequency spectrum
Relational equation finally carries out Fourier transformation to the equation under polar coordinates, acquires region to be treated according to cross power spectrum formula
Rotation angle between image and hospital's image.
Preferably, described according to interactive regional choice, show the tissue of different piece, specifically further include:
The image denoising that noisy image is carried out to the atrophy of Stationary Wavelet Transform neighbour coefficient, obtains sub-band coefficients: low respectively
Frequency coefficient, level detail coefficient, vertical detail coefficient and diagonal detail coefficient human-computer interaction choose sub-block region and save correspondence
Point coordinate, obtains scaling factor；
Filtered image is shown in new dialog box, its size can be zoomed in and out according to actual needs, and storage allocation is empty
Between save corresponding sub-image data and scaling factor；
Low frequency coefficient is remained unchanged, to the level detail coefficient, vertical detail coefficient and diagonal detail coefficient point of each layer
It carry out not adjacent region threshold processing；
Noise image is handled using Pulse Coupled Neural Network, sub-block bitmap after display processing；
The subgraph comprising target area manually determined by doctor is enhanced using histogram equalization technology, is changed
The partially dark phenomenon of image.
The above description is only an example of the present application, is not intended to limit this application.For those skilled in the art
For, various changes and changes are possible in this application.All any modifications made within the spirit and principles of the present application are equal
Replacement, improvement etc., should be included within the scope of the claims of this application.
Claims (6)
1. a kind of medical image analysis method based on computer vision characterized by comprising
Patient treats treatment region image using the mobile phone for being equipped with the software and is acquired: if acquired image is crooked, mould
Paste etc., mobile phone prompt patient re-shoot area image to be treated, obtain clear image；
Image is read in, and completes reading and conversion and the display of image and the Threshold segmentation of image: server terminal is using being based on
The template matching algorithm of shape feature and detection algorithm based on SIFT feature are detected, and are called and are sealed in OpenCV vision class libraries
The image processing operators installed, corresponding C++ code is completed in programming, and is realized using control button each in software interface
The function of algorithm；
Mobile phone starts the system software and carries out image preprocessing, and carries out handset identity: image preprocessing is completed at image filtering
The acquisition of reason and target subgraph and its histogram equalization processing, the branch for being used to fix sufferer that will include in medical image
Frame or other supporter garbages are deleted, and method is the original filled on erasing rubber institute route via with the gray value of background
Beginning gray value, to provide for medical diagnosis from qualitative to quantitative, original DICOM data are passed through reconciliation by more objective information
Mapping ground mode is converted into 8 BMP grayscale images, the contrast of enhancing output image；
Obtain regional image information to be treated, then obtained information is issued hospital by mobile phone: image segmentation and automation are quantitative
Analysis, obtains the outer contoured features of target, on the basis of image segmentation, obtain respectively target area longest and most width from
The number of pixels for inside including, and in proportion ruler coefficient conversion at entity long width values, it is reconstructed obtain threedimensional model after, assist function
Energy module mainly completes the three-dimensional view angle map function to model；
According to interactive regional choice, show the tissue of different piece: hospital send in real time regional image information to be treated to
Patient realizes quantitative background separation function using mouse action or keyboard input.
2. a kind of medical image analysis method based on computer vision as described in claim 1 characterized by comprising
The patient treats treatment region image image using the mobile phone for being equipped with the software and is acquired, specifically further include:
The collecting part of system is made of camera lens, camera and industrial personal computer etc.；
Collection process is by system camera photographic subjects background, after being converted to picture signal, everywhere by the transmission of USB2.0 interface
Reason server is handled；
It in image acquisition process, needs to digitize image, generally be handled using digitized image pick-up card.
3. a kind of medical image analysis method based on computer vision as described in claim 1 characterized by comprising
Described image is read in, and completes reading and conversion and the display of image and the Threshold segmentation of image, specifically further include:
It determines the partition threshold needed, partition threshold is divided into pixel compared with pixel value, determines a gray value, then will
Pixel greater than, equal to this gray value is judged to object and is indicated with a gray value, and the pixel for being less than this gray value is judged to
Background is simultaneously indicated with another gray value, or conversely, the gray scale size relation of object, background is exchanged, the result of binaryzation will
So that gray scale image becomes bianry image；
Determine threshold values be the key that segmentation, different threshold values may cause it is entirely different as a result, threshold values number and value
It is to be determined according to priori knowledge or test statistics data；
Pretreated subgraph is split using iterative threshold segmentation technology, and is existed according to the segmentation result in subgraph
In original image；
The profile of target is extracted, automation quantitative analysis is carried out for computer and creates conditions；
According in segmentation result image, the shape feature of the consistency of feature and all subregion is evaluated point in the same area
Cut the quality of result.
4. a kind of medical image analysis method based on computer vision as described in claim 1 characterized by comprising
The mobile phone starts the system software and carries out image preprocessing, and carries out handset identity, specifically further include:
Select multistage median filtering and edge detecting technology to complete the pretreatment to image, wherein multistage median filtering for pair
The pretreatment of full figure takes the average gray value of pixel in its certain neighborhood to substitute original pixel any pixel；
Edge detection further increases the discrimination of target and background, is for the processing to the subgraph comprising target area
Segmentation provides the image of high quality, carries out convolution algorithm with input picture and Gaussian function and obtains filtered image, then sharp
Amplitude and the direction that image gradient is calculated with differential operator, then traverse whole image, carry out non-maximum suppression to gradient magnitude
System finally carries out the extraction at edge using the method for dual thresholds, and the pixel more than higher thresholds is edge, on the contrary then be not,
If testing result is greater than Low threshold but be less than high threshold, that just compares in the neighborhood territory pixel of this pixel either with or without more than height
The edge pixel of threshold value, it is on the contrary then be not if there is that then should be edge.
5. a kind of medical image analysis method based on computer vision as described in claim 1 characterized by comprising
Described to obtain regional image information to be treated, then obtained information is issued hospital by mobile phone, specifically further include:
The rotation angle between area image to be treated and hospital's image is found out using Fourier transformation；
Input area image to be treated and hospital's image first, hospital's image and area image to be treated be to Same Scene institute at
Image exists between certain rotation differential seat angle namely area image to be treated and hospital's image between the two and exists centainly
Rotate differential seat angle；
Area image to be treated and hospital's image are transformed into Fourier frequency domain, obtain area image to be treated and hospital's figure
Then the frequency spectrum of picture treats the frequency spectrum for the treatment of region image and the Fourier frequency spectrum difference modulus value of hospital's image, establishes one
About the equation for rotating angle between two images, the relationship of area image frequency spectrum to be treated and hospital's image spectrum modulus value is found out；
Rotation angle equation is transformed under polar coordinates again, obtains two modulus value of Fourier frequency spectrum corresponding relationship under polar coordinates
Equation finally carries out Fourier transformation to the equation under polar coordinates, acquires area image to be treated according to cross power spectrum formula
Rotation angle between hospital's image.
6. a kind of medical image analysis method based on computer vision as described in claim 1 characterized by comprising
The regional choice according to interactive mode, shows the tissue of different piece, specifically further include:
The image denoising that noisy image is carried out to the atrophy of Stationary Wavelet Transform neighbour coefficient, obtains sub-band coefficients: low frequency system respectively
Number, level detail coefficient, vertical detail coefficient and diagonal detail coefficient human-computer interaction choose sub-block region and save corresponding points seat
Mark obtains scaling factor；
Filtered image is shown in new dialog box, its size can be zoomed in and out according to actual needs, and storage allocation space is protected
Deposit corresponding sub-image data and scaling factor；
Low frequency coefficient is remained unchanged, to the level detail coefficient, vertical detail coefficient and diagonal detail coefficient of each layer respectively into
The processing of row adjacent region threshold；
Noise image is handled using Pulse Coupled Neural Network, sub-block bitmap after display processing；
The subgraph comprising target area manually determined by doctor is enhanced using histogram equalization technology, changes image
Partially dark phenomenon.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910496499.2A CN110211143A (en) | 2019-06-10 | 2019-06-10 | A kind of medical image analysis method based on computer vision |
| CN202010519537.4A CN112070785A (en) | 2019-06-10 | 2020-06-09 | Medical image analysis method based on computer vision |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910496499.2A CN110211143A (en) | 2019-06-10 | 2019-06-10 | A kind of medical image analysis method based on computer vision |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110211143A true CN110211143A (en) | 2019-09-06 |
Family
ID=67791592
Family Applications (2)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910496499.2A Pending CN110211143A (en) | 2019-06-10 | 2019-06-10 | A kind of medical image analysis method based on computer vision |
| CN202010519537.4A Withdrawn CN112070785A (en) | 2019-06-10 | 2020-06-09 | Medical image analysis method based on computer vision |
Family Applications After (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010519537.4A Withdrawn CN112070785A (en) | 2019-06-10 | 2020-06-09 | Medical image analysis method based on computer vision |
Country Status (1)
| Country | Link |
|---|---|
| CN (2) | CN110211143A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111078090A (en) * | 2019-11-29 | 2020-04-28 | 上海联影医疗科技有限公司 | Display method, device, equipment and storage medium |
| CN112401829A (en) * | 2020-11-26 | 2021-02-26 | 温州眼视光国际创新中心 | Remote intelligent control system of slit lamp |
| CN113160153A (en) * | 2021-04-06 | 2021-07-23 | 宁波大学医学院附属医院 | Lung nodule screening method and system based on deep learning technology |
| CN113764072B (en) * | 2021-05-13 | 2023-04-18 | 腾讯科技（深圳）有限公司 | Medical image reconstruction method, device, equipment and storage medium |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112652382B (en) * | 2020-12-31 | 2022-05-06 | 山东大学齐鲁医院 | Gallbladder-pancreas disease multidisciplinary combined consultation and consultation system based on mobile terminal |
| CN116630425B (en) * | 2023-07-21 | 2023-09-22 | 长春市天之城科技有限公司 | Intelligent food detection system based on X rays |
| CN117541983A (en) * | 2023-11-08 | 2024-02-09 | 广东理致技术有限公司 | Model data quality analysis method and system based on machine vision |
-
2019
- 2019-06-10 CN CN201910496499.2A patent/CN110211143A/en active Pending
-
2020
- 2020-06-09 CN CN202010519537.4A patent/CN112070785A/en not_active Withdrawn
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111078090A (en) * | 2019-11-29 | 2020-04-28 | 上海联影医疗科技有限公司 | Display method, device, equipment and storage medium |
| CN112401829A (en) * | 2020-11-26 | 2021-02-26 | 温州眼视光国际创新中心 | Remote intelligent control system of slit lamp |
| CN113160153A (en) * | 2021-04-06 | 2021-07-23 | 宁波大学医学院附属医院 | Lung nodule screening method and system based on deep learning technology |
| CN113764072B (en) * | 2021-05-13 | 2023-04-18 | 腾讯科技（深圳）有限公司 | Medical image reconstruction method, device, equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112070785A (en) | 2020-12-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110211143A (en) | A kind of medical image analysis method based on computer vision | |
| AlZubi et al. | Multiresolution analysis using wavelet, ridgelet, and curvelet transforms for medical image segmentation | |
| EP0766205B1 (en) | Image processing process and device for detecting regions with a predetermined cancer type in an intensity image | |
| CN111539930A (en) | Dynamic ultrasonic breast nodule real-time segmentation and identification method based on deep learning | |
| CN107590809A (en) | Lung dividing method and medical image system | |
| Sinha et al. | Medical image processing | |
| Yang | Multimodal medical image fusion through a new DWT based technique | |
| CN105488781B (en) | A kind of dividing method based on CT images liver neoplasm lesion | |
| Dehghani et al. | A method for improve preprocessing images mammography | |
| CN114241077B (en) | CT image resolution optimization method and device | |
| Nie et al. | Feature extraction for medical CT images of sports tear injury | |
| CN113850328A (en) | Non-small cell lung cancer subtype classification system based on multi-view deep learning | |
| Edwin et al. | Liver and tumour segmentation from abdominal CT images using adaptive threshold method | |
| Mandour | An Exhaustive Review of Neutrosophic Logic in Addressing Image Processing Issues | |
| Muthiah et al. | Fusion of MRI and PET images using deep learning neural networks | |
| Guo et al. | Thyroid nodule ultrasonic imaging segmentation based on a deep learning model and data augmentation | |
| Guo | Real-time medical image denoising and information hiding model based on deep wavelet multiscale autonomous unmanned analysis | |
| Shi et al. | Texture features based microscopic image classification of liver cellular granuloma using artificial neural networks | |
| Noviana et al. | Axial segmentation of lungs CT scan images using canny method and morphological operation | |
| CN110084772B (en) | MRI/CT fusion method based on bending wave | |
| Ivković et al. | Segmentation and classification of Bi-Rads medical images with the imaging biomarkers according to level of detail | |
| CN111462065A (en) | Breast cancer detection method and system based on ultrasonic and infrared sequence image fusion | |
| Shaziya et al. | Comprehensive review of automatic lung segmentation techniques on pulmonary CT images | |
| Hou et al. | Research on tumor image segmentation in medical imaging based on extremum adaptive median filtering | |
| Karthikeyan et al. | Wavelet-based Image Enhancement Techniques for Improving Visual Quality of Ultrasonic Images |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20190906 |
|
| WD01 | Invention patent application deemed withdrawn after publication |