CN101161204A - Method and apparatus for recognizing intelligently ultrasound womb contraceptive ring image - Google Patents
Method and apparatus for recognizing intelligently ultrasound womb contraceptive ring image Download PDFInfo
- Publication number
- CN101161204A CN101161204A CNA2007101246297A CN200710124629A CN101161204A CN 101161204 A CN101161204 A CN 101161204A CN A2007101246297 A CNA2007101246297 A CN A2007101246297A CN 200710124629 A CN200710124629 A CN 200710124629A CN 101161204 A CN101161204 A CN 101161204A
- Authority
- CN
- China
- Prior art keywords
- image
- womb
- candidate
- recognition
- ultrasound
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Landscapes
- Image Processing (AREA)
- Ultra Sonic Daignosis Equipment (AREA)
Abstract
A ultrasonic womb contraceptive ring image intelligent identifying method is provided in the present invention, comprising the following steps: a, collecting ultrasonic womb image by B-ultrasonic equipment; b, processing to eliminate noise and gray scale linear intercept to the original womb image; c, processing to fixed threshold value separate and morphology treatment to processed womb image, through region extraction to acquire a plurality of candidate object like contraceptive ring; d, extracting the characteristic value of the candidate object, using contraceptive ring image data parameter base and multi-classifying fusion structure identifying system to process intelligent identification to the acquired candidate object, judging whether there is contraceptive ring object in the ultrasonic womb image. The present invention also provides an equipment used by the method. The present invention can effectively be used for assisting ultrasonic ring-checking member to operation, promote the accuracy of ring-checking examination.
Description
Technical field
The present invention relates to a kind of method and apparatus that intrauterine device object in the ultrasound womb image is carried out Intelligent Recognition.
Background technology
China's population surpasses 1,300,000,000 at present, and huge population pressure has brought white elephant for many aspects such as China's economic development, social stability and environmental conservation.Family planning has been made huge contribution to slowing down China's rate of population increase with the pressure that the alleviation overpopulation brings as the fundamental state policy of China.
In family planning work, ensure that the smooth execution effectively in vast basic unit of every birth-control measures is the most important thing in every work.Ensure effective execution of birth-control measures, just need implement of the right age women is carried out the ultrasonic ring inspection of looking on schedule, look into the practical situation of the number of times of ring inspection, be generally 1 year 3 to 4 times,, also have 1 year 2 times if this area's economic condition is not high by each department.However, still exist to rob and give birth to the excusing from death phenomenon, this is to utilize to look into the leak of ring work, becomes pregnant after once looking into the ring inspection, as false retrieval omission situation occurs in looking into ring work next time, then can escape the birth control remedial measure and cause excusing from death.Look into that to occur false retrieval omission phenomenon in the ring work be because vast basic unit to a great extent, especially remote countryside is geographic ultrasonicly looks into that ring worker medical level is limited to be caused, simultaneously ultrasonicly look into the ring flow of personnel and run off very serious, often just just trained and run off, ultrasonic ultrasonoscopy training of looking into the ring personnel is very heavy, and above present situation has brought very large hidden danger for whole family planning work.
In view of this, be necessary to research and develop and a kind ofly can effectively and accurately carry out the method and apparatus that Intelligent Recognition is judged, by Real time identification and point out current B ultrasonic whether to exist contraceptive ring image assist raising basic unit to look into the accuracy of ring work in the image of uterus as shown on screen intrauterine device object in the ultrasound womb image; Simultaneously also can be used as training tool, to alleviate the burden that basic unit looks into the training of ring personnel ultrasonoscopy.
Summary of the invention
The purpose of this invention is to provide a kind of intelligently ultrasound womb contraceptive ring image recognition methods and equipment, whether this method and apparatus can be judged existing the intrauterine device object to carry out Intelligent Recognition in the ultrasound womb image effectively.
For achieving the above object, the intelligently ultrasound womb contraceptive ring image recognition methods of the present invention's proposition comprises the following steps:
A. gather the ultrasound womb image by B ultrasonic equipment;
B. original uterus image is carried out Flame Image Process, comprise the intercepting of denoising and gray scale;
C. the uterus image after handling is carried out that fixed threshold is cut apart and morphology is handled, obtain candidate's object of a plurality of doubtful intrauterine devices by extracted region;
D. extract the eigenvalue of candidate's object, use the recognition system of contraceptive ring image data parameters storehouse and multiple Classifiers Combination structure that the candidate's object that obtains is carried out Intelligent Recognition, judge in this ultrasound womb image whether contain the intrauterine device object;
E. by display device display image recognition result.
Image processing process described in the step b comprises the following steps: at first, analyze the whole signal noise ratio level and the speckle noise order of severity of ultrasonoscopy and carry out the adaptive-filtering processing, then, intercepting accounts for the bright area of the gross area 20%, its tonal range linearity is drawn high in 0 to 255 gray space, and the dark areas gray scale that will account for all the other 80% areas to transfer to be 0.
Image segmentation described in the step c adopts fixed threshold to cut apart and the morphology processing is carried out image segmentation to the ultrasound womb image, and extracts candidate's object, specifically comprises the following steps: at first, and use the fixed threshold method to cut apart, segmentation threshold is 80; Then, adopt closing operation of mathematical morphology to handle to the uterus image after cutting apart to optimize segmentation effect; At last, adopt extracted region to obtain candidate's object of a plurality of doubtful intrauterine devices to the uterus image after cutting apart.
The recognition feature value of candidate's object of the doubtful intrauterine device described in the steps d can be selected following part or all of feature: the gray feature in local background zone around candidate's object reaches comprises object gray average, object gray variance and background gray average; With the material standed for body weight heart is the picture element gray matrix of topography zone after the pyramid sample process at center; The morphological characteristic of candidate's object comprises object area, width and class circularity etc.
In the steps d different characteristic value of candidate's object imported respectively and carry out Classification and Identification at least three graders, obtain recognition credibility, comprehensively obtain recognition result in the credibility input fusion device that then each grader is drawn, to judge in this ultrasound womb image whether contain the intrauterine device object.
The present invention also provides a kind of intelligently ultrasound womb contraceptive ring image identification equipment, comprise ultrasonoscopy sniffer, ultrasonoscopy generating apparatus, ultrasonoscopy display device, it is characterized in that: also comprise the Flame Image Process and the recognition device that are arranged between video generation device and the ultrasonoscopy display device.
Described Flame Image Process and recognition device comprise a uterus view data parameter library, a Flame Image Process and cut apart module and an intrauterine device object identification module.
Deposit all kinds of eigenvalues that extracted from the contraceptive ring image storehouse in the view data parameter library of described uterus and train the optimal parameter that is obtained after the test job through the specific classification device.
Described intrauterine device object identification module comprises double-layer structure, and ground floor is made up of at least three tagsort devices; The second layer is made up of the fusion device of classification more than.
Intrauterine device object identification module can link to each other with a recognition result display device, and the recognition result display device is used for output and shows recognition result.
Whether intelligently ultrasound womb contraceptive ring image recognition methods and equipment that the present invention proposes can accurately to existing the intrauterine device object to carry out Intelligent Recognition in the ultrasound womb image, and export judged result and point out looking into the ring worker, improve the purpose of looking into ring work accuracy to reach.
Description of drawings
The present invention is described in detail below in conjunction with preferred embodiment and accompanying drawing, wherein:
Fig. 1 is the theory diagram of intelligently ultrasound womb contraceptive ring image recognition methods of the present invention;
Fig. 2 is the particular flow sheet of intelligently ultrasound womb contraceptive ring image recognition methods of the present invention;
Fig. 3 is that the present invention uses greyscale transformation and fixed threshold method ultrasonoscopy to be carried out the flow chart of image segmentation.
The specific embodiment
Key problem in technology of the present invention is the ultrasound womb image is carried out Intelligent Recognition, automatically identifies whether contain the intrauterine device object in the current ultrasonoscopy by machine.Know-why of the present invention comprises Flame Image Process, pattern recognition and artificial intelligence technology.
Fig. 1 is the theory diagram that the present invention is applied in a better embodiment in the ultrasonoscopy.General ultrasonic device comprises a ultrasonoscopy sniffer 11, a ultrasonoscopy generating apparatus 12 and a ultrasonovision 14.Ultrasonoscopy sniffer 11 adopts medical Type B ultrasonic device that position, women checked people uterus is surveyed; The ultrasonic signal that ultrasonoscopy generating apparatus 12 is used for detecting according to ultrasonoscopy sniffer 11 generates the ultrasound womb image; 14 of ultrasonoscopy display devices are used to show the uterus image.The present invention also is provided with a Flame Image Process and recognition device 13 between ultrasonoscopy generating apparatus 12 and ultrasonoscopy display device 14, be used for that the uterus image that ultrasonoscopy generating apparatus 12 generates is carried out Intelligent Recognition and handle, judge whether contain the intrauterine device object among the figure.In implementation process, can be equipped with simultaneously recognition result display device 15 and be used for the intrauterine device recognition result that display image is handled and recognition device 13 is exported.
Flame Image Process and recognition device 13 comprise a womb contraceptive ring image data parameters storehouse 16, a Flame Image Process and image segmentation module 17 and an intrauterine device object identification module 18.
Womb contraceptive ring image data parameters storehouse 16 among the present invention is formed from 2300 width of cloth typical case ultrasound womb image according to the concrete supervised learning grader that adopts and is obtained the image library, there is the intrauterine device object in 1037 width of cloth uterus images in this image library, do not have intrauterine device in 1263 width of cloth uterus images.All images is all checked the working site collection from the ultrasonic ring of looking in the image library, adopts ultrasonic probe crosscut and two kinds of standard maneuvers of rip cutting.The particular location that whether has intrauterine device object and intrauterine device object in the image of uterus is veteranly looked into the ring worker and is confirmed jointly by several.This image library has fully comprised in difference checks the dissimilar contraceptive ring images of different women checked people's intrauterine under the maneuver, is a standard picture storehouse describing ultrasound womb contraceptive ring image.
Flame Image Process and image segmentation module 17 are used for the uterus image that ultrasonoscopy generating apparatus 12 generates is carried out Flame Image Process and image segmentation, to obtain candidate's object of a plurality of doubtful intrauterine devices.Intrauterine device object identification module 18 is used for candidate's object of doubtful intrauterine device is carried out Intelligent Recognition, judges in the current ultrasound womb image whether contain the intrauterine device object.
Recognition result display device 15 is used for the intrauterine device recognition result that display image is handled and recognition device 13 is exported, and it can select different implementations for use according to different application scenarios.Recognition result display device 15 both can be designed to directly show the hardware device of intrauterine device information, also can be designed to the intrauterine device identifying information is outputed to the interface that other softwares or hardware system show.
Fig. 2 carries out the particular flow sheet of intelligent identification Method to ultrasound womb contraceptive ring image for the present invention.The key step of the inventive method is as follows:
1, the original ultrasound womb image to input carries out Flame Image Process and image segmentation, and this step is finished by Flame Image Process and image segmentation module 17.
Because the restriction of ultrasonic device image-forming principle, the ultrasonoscopy signal to noise ratio is low and the speckle noise phenomenon is serious, therefore need to adopt auto-adaptive filtering technique that original ultrasonoscopy is carried out denoising, this technology is at first analyzed the whole signal noise ratio level and the speckle noise order of severity of ultrasonoscopy, then each picture element in the image is all judged that by local neighborhood intensity profile situation around it this pixel is positioned at homogeneous region or different filtering modes is selected in the non-homogeneous zone, homogeneous region is adopted mean filter, medium filtering is adopted in the non-homogeneous zone.Under the occasion that processing speed is had relatively high expectations, also can directly use traditional filtering modes such as mean filter that image is carried out denoising.
The ultrasound womb image is under all multifactor common influences such as the super equipment of different B, different probe positions, different tested person's position, different intrauterine device types, make that the imaging difference of intrauterine device object is very big, if the conventional art means such as ultrasonoscopy applying template similarity coupling after the denoising are unpractical and do not have feasibility.For further extracting the information in the image, need carry out image segmentation to the image after the denoising and handle.For realizing after filtering and noise reduction and greyscale transformation, to use fixed threshold method and morphology to handle ultrasonoscopy is carried out the image segmentation processing to automatically the cutting apart fast of ultrasound womb image.
Fig. 3 carries out the intercepting of filtering and noise reduction and gray scale earlier to handle among the present invention, the back uses fixed threshold method and morphology operations to carry out the flow chart of image partition method.The intrauterine device object image-forming has two big characteristics in the ultrasound womb image, and the one, intrauterine device overall intensity level is higher, and the 2nd, the uterine region overall intensity level around the intrauterine device is lower.After denoising, in order to regulate the overall intensity level of the ultrasound womb image of gathering under the different condition, utilize the intrauterine device imaging characteristics, image is carried out the gray scale intercepting to be handled, be specially image pixel by gray value from secretly to bright ordering, gray value from dark end distance this pixel from for whole pixel number 80% ratio the time is a reference value, be lower than reference value as gray value and then be adjusted into 0, and the tonal range linearity of reference value to 255 drawn high in 0 to 255 gray space, this process can be adjusted to close intensity profile with the uterus image that generates under the different condition, to improve the order of accuarcy of next step image segmentation.Use the fixed threshold method to carry out image segmentation to adjusted image subsequently, segmentation threshold uses 80 usually.Adopt morphology to handle to optimize segmentation effect to the image after cutting apart, adopting radius usually is that the structural element of 3 pixels carries out closed operation and handles.Adopt extracted region to obtain candidate's object of a plurality of doubtful intrauterine devices to split image at last.
2, whether the doubtful candidate's object of identification decision is the intrauterine device imaging, and this step is finished by intrauterine device object identification module 18.
Doubtful candidate's object quantity of every width of cloth ultrasound womb image gained through Flame Image Process and after cutting apart has nothing in common with each other according to image is different, usually about 10 to 20.If include the intrauterine device object image-forming in the image, these candidate's objects have one according to the situation of cutting apart and belong to the intrauterine device object to several objects, and candidate's object of all the other non-intrauterine devices need be excluded; If do not comprise the intrauterine device object image-forming in the image, then all candidate's objects all need to be excluded.
Because ultrasonograph quality is relatively poor, in order to improve the accuracy rate of identification, this method uses the multiple Classifiers Combination framework whether each candidate's object is discerned judgement as the intrauterine device imaging.Classifier design all belongs to aforementioned supervised learning technology, and concrete grader can be the BP neural network classifier in this method, and also can be other main flows has supervised classification device (as SVM).
In the present embodiment, the multiple Classifiers Combination framework is a double-layer structure, and ground floor is made up of three graders, can be provided with two or more than three graders as required, each grader is all accepted the different characteristic vector of doubtful candidate's object and is imported as grader, and grader is output as recognition credibility.The eigenvalue of doubtful candidate's object is selected to comprise following three kinds of modes: 1, and the morphological characteristic of candidate's object comprises object area, width and class circularity etc.; 2, the gray feature in local background zone comprised object gray average, object gray variance and background gray average etc. around candidate's object reached; 3, be the picture element gray matrix of topography zone after the pyramid sample process at center with the material standed for body weight heart.
Grader training and test process based on the ultrasound womb image library are as follows: use the interior parts of images of image library as training image, through extracting above-mentioned three kinds of characteristic parameters of candidate's object among the figure after the image segmentation respectively, ultrasonicly look into the ring worker and judge that wherein which is the imaging of intrauterine device object by experienced again, which is the imaging of other objects, to all equal label of material standed for body tag (Label) information, form training sample.Use BP neutral net or support vector machine main flow graders such as (SVM) that training sample is trained, with another part image in the image library as test pattern, the characteristic parameter of candidate's object is tested among the extraction figure respectively, selects the classifier parameters of recognition effect the best.The optimal parameter that obtains is deposited into view data parameter library 16.When using, use the relevant parameter of depositing in the view data parameter library 16 that grader is carried out parameter configuration, need doubtful candidate's object of identification will obtain recognition credibility by the grader that configures parameter, be used as down the input of one deck grader.
The second layer is made up of a fusion device, that can adopt main flow has a supervised classification device, the weighting ballot method graders that adopt more, this grader is collected the credibility output of all graders of ground floor and is formed a credibility characteristic vector, input feature vector as this grader, the credibility that grader is output as weighting after handling, output valve are between 0 to 1, and numerical value is big more, and to show that processed doubtful candidate's object belongs to the probability of intrauterine device big more.The optimal weights parameter that weighting ballot method grader obtains in training and test process also will be deposited into view data parameter library 16.When using, use the respective weights parameter of depositing in the view data parameter library 16 that grader is configured, need the credibility characteristic vector of identification will obtain final weighting credibility, as the final result of identification by the grader that configures parameter.In the present embodiment, the selection output valve belongs to the intrauterine device imaging greater than doubtful candidate's object of 0.8, and the doubtful candidate's object smaller or equal to 0.8 does not belong to the intrauterine device imaging.
Whole doubtful candidate's object in the image is discerned judgement, as whole doubtful candidate's object output valves in the piece image all smaller or equal to 0.8, then this image is identified to be judged as and does not contain the intrauterine device object image-forming, all greater than 0.8, then this image is identified to be judged as and contains the intrauterine device object image-forming if any doubtful candidate's object output valve.The output of recognition result is finished by recognition result display device 15, and it can select different implementations for use according to different application scenarios.Recognition result display device 15 both can be designed to directly show the hardware device of intrauterine device information, also can be designed to the intrauterine device identifying information is outputed to the interface that other softwares or hardware system show.
Whether said method and equipment can realize existing the intrauterine device object to carry out Intelligent Recognition accurately in the ultrasound womb image, and the output judged result points out looking into the ring worker, improve the purpose of looking into ring work accuracy to reach.
Claims (13)
1. intelligently ultrasound womb contraceptive ring image recognition methods is characterized in that this method comprises the following steps:
A. gather the ultrasound womb image by B ultrasonic equipment;
B. original uterus image is carried out Flame Image Process, comprise the intercepting of denoising and gray scale;
C. the uterus image after handling is carried out that fixed threshold is cut apart and morphology is handled, obtain candidate's object of a plurality of doubtful intrauterine devices by extracted region;
D. extract the eigenvalue of candidate's object, use the recognition system of contraceptive ring image data parameters storehouse and multiple Classifiers Combination structure that the candidate's object that obtains is carried out Intelligent Recognition, judge in this ultrasound womb image whether contain the intrauterine device object;
E. by display device display image recognition result.
2. the method for claim 1, it is characterized in that: the image processing process described in the step b comprises the following steps: at first, analyze the whole signal noise ratio level and the speckle noise order of severity of ultrasonoscopy and carry out the adaptive-filtering processing, then, intercepting accounts for the bright area of the gross area 20%, its tonal range linearity is drawn high in 0 to 255 gray space, and the dark areas gray scale that will account for all the other 80% areas to transfer to be 0.
3. the method for claim 1, it is characterized in that: the image segmentation described in the step c adopts fixed threshold to cut apart and the morphology processing is carried out image segmentation to the ultrasound womb image, and extraction candidate object, specifically comprise the following steps: at first, use the fixed threshold method to cut apart, segmentation threshold is 80; Then, adopt closing operation of mathematical morphology to handle to the uterus image after cutting apart to optimize segmentation effect; At last, adopt extracted region to obtain candidate's object of a plurality of doubtful intrauterine devices to the uterus image after cutting apart.
4. the method for claim 1, it is characterized in that: the recognition feature value selection mode of candidate's object of the doubtful intrauterine device described in the steps d comprises object gray average, object gray variance and background gray average for candidate's object reaches the gray feature in local background zone on every side.
5. the method for claim 1 is characterized in that: the recognition feature value selection mode of candidate's object of the doubtful intrauterine device described in the steps d is for being the picture element gray matrix of topography zone after the pyramid sample process at center with the material standed for body weight heart.
6. the method for claim 1 is characterized in that: the recognition feature value selection mode of candidate's object of the doubtful intrauterine device described in the steps d comprises object area, width and class circularity for the morphological characteristic of candidate's object.
7. the method for claim 1, it is characterized in that: in the steps d different characteristic value of candidate's object imported respectively and carry out Classification and Identification at least three graders, obtain recognition credibility, comprehensively obtain recognition result in the credibility input fusion device that then each grader is drawn, to judge in this ultrasound womb image whether contain the intrauterine device object.
8. the equipment that uses of the described method of a claim 1, comprise ultrasonoscopy sniffer, ultrasonoscopy generating apparatus, ultrasonoscopy display device, it is characterized in that: also comprise the Flame Image Process and the recognition device that are arranged between video generation device and the ultrasonoscopy display device.
9. equipment as claimed in claim 8 is characterized in that: described Flame Image Process and recognition device comprise a uterus view data parameter library, a Flame Image Process and cut apart module and an intrauterine device object identification module.
10. equipment as claimed in claim 9 is characterized in that: deposit all kinds of eigenvalues that extracted from the contraceptive ring image storehouse in the view data parameter library of described uterus and train the optimal parameter that is obtained after the test job through the specific classification device.
11. equipment as claimed in claim 9 is characterized in that: described intrauterine device object identification module comprises double-layer structure, and ground floor is made up of at least three tagsort devices; The second layer is made up of the fusion device of classification more than.
12. equipment as claimed in claim 11 is characterized in that: described tagsort device is BP neural network classifier or SVM (support vector machine) grader, and described many classification fusion devices are weighting ballot method or BP neural network classifier.
13. equipment as claimed in claim 8 is characterized in that: described intrauterine device object identification module links to each other with a recognition result display device.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2007101246297A CN100500100C (en) | 2007-11-20 | 2007-11-20 | Method and apparatus for recognizing intelligently ultrasound womb contraceptive ring image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2007101246297A CN100500100C (en) | 2007-11-20 | 2007-11-20 | Method and apparatus for recognizing intelligently ultrasound womb contraceptive ring image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101161204A true CN101161204A (en) | 2008-04-16 |
| CN100500100C CN100500100C (en) | 2009-06-17 |
Family
ID=39296243
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2007101246297A Expired - Fee Related CN100500100C (en) | 2007-11-20 | 2007-11-20 | Method and apparatus for recognizing intelligently ultrasound womb contraceptive ring image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100500100C (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102113897A (en) * | 2009-12-31 | 2011-07-06 | 深圳迈瑞生物医疗电子股份有限公司 | Method and device for extracting target-of-interest from image and method and device for measuring target-of-interest in image |
| CN105705098A (en) * | 2013-09-20 | 2016-06-22 | 透壁生物技术公司 | Image analysis techniques for diagnosing diseases |
| CN108985366A (en) * | 2018-07-06 | 2018-12-11 | 武汉兰丁医学高科技有限公司 | B ultrasound image recognition algorithm and system based on convolution depth network |
| CN109602445A (en) * | 2018-12-06 | 2019-04-12 | 余姚市华耀工具科技有限公司 | Spleen defect detection platform |
| CN109620294A (en) * | 2018-12-08 | 2019-04-16 | 余姚市华耀工具科技有限公司 | Malignancy appraisal organization |
| CN110448273A (en) * | 2019-08-29 | 2019-11-15 | 江南大学 | A kind of low-power consumption epileptic prediction circuit based on support vector machines |
-
2007
- 2007-11-20 CN CNB2007101246297A patent/CN100500100C/en not_active Expired - Fee Related
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102113897A (en) * | 2009-12-31 | 2011-07-06 | 深圳迈瑞生物医疗电子股份有限公司 | Method and device for extracting target-of-interest from image and method and device for measuring target-of-interest in image |
| US8699766B2 (en) | 2009-12-31 | 2014-04-15 | Shenzhen Mindray Bio-Medical Electronics Co., Ltd. | Method and apparatus for extracting and measuring object of interest from an image |
| CN102113897B (en) * | 2009-12-31 | 2014-10-15 | 深圳迈瑞生物医疗电子股份有限公司 | Method and device for extracting target-of-interest from image and method and device for measuring target-of-interest in image |
| CN105705098A (en) * | 2013-09-20 | 2016-06-22 | 透壁生物技术公司 | Image analysis techniques for diagnosing diseases |
| CN108985366A (en) * | 2018-07-06 | 2018-12-11 | 武汉兰丁医学高科技有限公司 | B ultrasound image recognition algorithm and system based on convolution depth network |
| CN109602445A (en) * | 2018-12-06 | 2019-04-12 | 余姚市华耀工具科技有限公司 | Spleen defect detection platform |
| CN109620294A (en) * | 2018-12-08 | 2019-04-16 | 余姚市华耀工具科技有限公司 | Malignancy appraisal organization |
| CN110448273A (en) * | 2019-08-29 | 2019-11-15 | 江南大学 | A kind of low-power consumption epileptic prediction circuit based on support vector machines |
| CN110448273B (en) * | 2019-08-29 | 2021-03-30 | 江南大学 | Low-power-consumption epilepsy prediction circuit based on support vector machine |
Also Published As
| Publication number | Publication date |
|---|---|
| CN100500100C (en) | 2009-06-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100462054C (en) | Method for shielding sex part on foetus image for preventing recognizing foetus sex | |
| Nhat-Duc et al. | Automatic recognition of asphalt pavement cracks using metaheuristic optimized edge detection algorithms and convolution neural network | |
| Hoang | Automatic detection of asphalt pavement raveling using image texture based feature extraction and stochastic gradient descent logistic regression | |
| CN100500100C (en) | Method and apparatus for recognizing intelligently ultrasound womb contraceptive ring image | |
| WO2019104767A1 (en) | Fabric defect detection method based on deep convolutional neural network and visual saliency | |
| EP1593094B1 (en) | Image analysis for the purpose of assessing cancer | |
| CN112232138B (en) | Channel slope damage intelligent identification method based on superpixel characteristics | |
| CN104794502A (en) | Image processing and mode recognition technology-based rice blast spore microscopic image recognition method | |
| CN109241871A (en) | A kind of public domain stream of people's tracking based on video data | |
| CN108460744A (en) | A kind of enhancing of cement cutting pavement image noise reduction and FRACTURE CHARACTERISTICS extracting method | |
| CN104008551B (en) | A kind of Citrus Huanglongbing pathogen detection method based on visible images | |
| CN108764186A (en) | Personage based on rotation deep learning blocks profile testing method | |
| CN107133638A (en) | Multi-parameter MRI prostate cancer CAD method and system based on two graders | |
| CN109003275A (en) | The dividing method of weld defect image | |
| CN109948501A (en) | The detection method of personnel and safety cap in a kind of monitor video | |
| CN107273852A (en) | Escalator floor plates object and passenger behavior detection algorithm based on machine vision | |
| CN110264444A (en) | Damage detecting method and device based on weak segmentation | |
| CN105607288A (en) | Intelligent glasses omnibearing vehicle part completeness detection method based on acoustic detection assistance | |
| CN106326834A (en) | Human body gender automatic identification method and apparatus | |
| CN105023272A (en) | Crop leaf insect pest detection method and system | |
| CN103325123A (en) | Image edge detection method based on self-adaptive neural fuzzy inference systems | |
| Wang et al. | An edge detection method by combining fuzzy logic and neural network | |
| CN111053579B (en) | Classification method and system for pelvic floor state, ultrasonic equipment and computer storage medium | |
| CN104331864A (en) | Breast imaging processing based on non-subsampled contourlet and visual salient model | |
| Yunhong et al. | Edge Detection Algorithm of MRI Medical Image Based on Artificial Neural Network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20090617 Termination date: 20181120 |