CN110517258A - A kind of cervical carcinoma pattern recognition device and system based on high light spectrum image-forming technology - Google Patents

A kind of cervical carcinoma pattern recognition device and system based on high light spectrum image-forming technology Download PDF

Info

Publication number
CN110517258A
CN110517258A CN201910815815.8A CN201910815815A CN110517258A CN 110517258 A CN110517258 A CN 110517258A CN 201910815815 A CN201910815815 A CN 201910815815A CN 110517258 A CN110517258 A CN 110517258A
Authority
CN
China
Prior art keywords
image
hyperion
uterine neck
person
identified
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
Application number
CN201910815815.8A
Other languages
Chinese (zh)
Inventor
李玮
谢龙跃
韩毅
袁双虎
孟祥伟
马志祥
李希元
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shandong University
Original Assignee
Shandong University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Shandong University filed Critical Shandong University
Priority to CN201910815815.8A priority Critical patent/CN110517258A/en
Publication of CN110517258A publication Critical patent/CN110517258A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/084—Backpropagation, e.g. using gradient descent
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0012—Biomedical image inspection
    • H—ELECTRICITY
    • H03—ELECTRONIC CIRCUITRY
    • H03H—IMPEDANCE NETWORKS, e.g. RESONANT CIRCUITS; RESONATORS
    • H03H17/00—Networks using digital techniques
    • 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
    • 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
    • G06T2207/20081—Training; Learning
    • 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
    • G06T2207/20084—Artificial neural networks [ANN]
    • 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
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30004—Biomedical image processing
    • G06T2207/30096—Tumor; Lesion
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V2201/00—Indexing scheme relating to image or video recognition or understanding
    • G06V2201/03—Recognition of patterns in medical or anatomical images
    • H—ELECTRICITY
    • H03—ELECTRONIC CIRCUITRY
    • H03H—IMPEDANCE NETWORKS, e.g. RESONANT CIRCUITS; RESONATORS
    • H03H17/00—Networks using digital techniques
    • H03H2017/0072—Theoretical filter design
    • H03H2017/0081—Theoretical filter design of FIR filters

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • General Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • General Health & Medical Sciences (AREA)
  • General Engineering & Computer Science (AREA)
  • Evolutionary Computation (AREA)
  • Health & Medical Sciences (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Biophysics (AREA)
  • Computing Systems (AREA)
  • Quality & Reliability (AREA)
  • Computational Linguistics (AREA)
  • Radiology & Medical Imaging (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Molecular Biology (AREA)
  • Biomedical Technology (AREA)
  • Medical Informatics (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • Image Analysis (AREA)
  • Investigating Or Analysing Materials By Optical Means (AREA)

Abstract

The invention discloses a kind of cervical carcinoma pattern recognition device and system based on high light spectrum image-forming technology improves high spectrum image accuracy of identification and speed；The device includes: data acquisition module, is used for EO-1 hyperion uterine neck image；Database module establishes the large database concept of EO-1 hyperion uterine neck image for the EO-1 hyperion uterine neck image of EO-1 hyperion uterine neck image and each period cervical carcinogenesis patient using obtained normal person；Neural network module carries out feature extraction and classification for the EO-1 hyperion uterine neck image to different times cervical carcinogenesis patient, and establishes the BP neuroid of each classification respectively；3-D image establishes module, for handling using EO-1 hyperion uterine neck image of the BP neuroid established to person to be identified, establishes the three-dimensional high spectrum image of person to be identified；Data fitting module obtains fitting image for being fitted the three-dimensional high spectrum image of person to be identified and uterine neck image in large database concept using CEM algorithm.

Description

A kind of cervical carcinoma pattern recognition device and system based on high light spectrum image-forming technology
Technical field
This disclosure relates to image identification technical field, and in particular to a kind of cervical carcinoma image based on high light spectrum image-forming technology Identification device and cervical carcinoma image identification system.
Background technique
With the development of science and technology, the identification of all kinds of images starts to develop to automation direction, the application of artificial intelligence is more next It is wider, a large amount of manpower and material resources are not only saved, precision is also very high.
High light spectrum image-forming technology is the image data technology based on very more narrow-bands, it is by imaging technique and spectral technique Combine, detect target two-dimensional geometry space and one-dimensional spectral information, obtain continuous, narrow-band the figure of high spectral resolution As data.High light spectrum image-forming technology is quickly grown at present, similarly can be used using the image that high light spectrum image-forming technology generates The method automatic identification of artificial intelligence.
Inventor has found that the imaging method of traditional cervical carcinoma image is by vaginoscopy, hand in R&D process Method is complicated, low efficiency, and traditional images recognition methods is to judge that the method precision is low according to the experience of oneself by doctor, Time-consuming length, and low efficiency.
Summary of the invention
In order to overcome the above-mentioned deficiencies of the prior art, present disclose provides a kind of cervical carcinomas based on high light spectrum image-forming technology Pattern recognition device and cervical carcinoma image identification system improve high spectrum image accuracy of identification and speed.
A kind of technical side of on the one hand cervical carcinoma pattern recognition device based on high light spectrum image-forming technology that the disclosure provides Case is:
A kind of cervical carcinoma pattern recognition device based on high light spectrum image-forming technology, the device include:
Data acquisition module, for obtaining the EO-1 hyperion uterine neck image of normal person and the height of each period cervical carcinogenesis patient Spectrum uterine neck image and the EO-1 hyperion uterine neck image of person to be identified；
Database module, for the EO-1 hyperion uterine neck image and each period cervical carcinogenesis using obtained normal person The EO-1 hyperion uterine neck image of patient, establishes the large database concept of EO-1 hyperion uterine neck image；
Neural network module carries out feature for the EO-1 hyperion uterine neck image to different times cervical carcinogenesis patient and mentions It takes and classifies, and establish the BP neuroid of each classification respectively；
3-D image establishes module, for utilizing established BP neuroid to the EO-1 hyperion uterine neck figure of person to be identified As being handled, the three-dimensional high spectrum image of person to be identified is established；
Data fitting module, for using CEM algorithm by the three-dimensional high spectrum image of person to be identified and large database concept Middle Palace Neck image is fitted, and obtains fitting image.
A kind of technical side of on the one hand cervical carcinoma image identification system based on high light spectrum image-forming technology that the disclosure provides Case is:
A kind of cervical carcinoma image identification system based on high light spectrum image-forming technology, the system include Hyperspectral imaging devices and Processor；
The Hyperspectral imaging devices, for generate respectively normal person EO-1 hyperion uterine neck image and each period cervical carcinoma Become the uterine neck image of patient, and generate the EO-1 hyperion uterine neck image of person to be identified, extremely by the EO-1 hyperion uterine neck image transmitting of generation Processor；
The processor, for obtaining the EO-1 hyperion uterine neck image of normal person and the uterine neck of each period cervical carcinogenesis patient Image establishes large database concept；Feature extraction and classification are carried out to the EO-1 hyperion uterine neck image of different times cervical carcinogenesis patient, and The BP neuroid for establishing each classification respectively is carried out using EO-1 hyperion uterine neck image of the BP neuroid to person to be identified Processing, establishes the three-dimensional high spectrum image of person to be identified, by uterine neck in the three-dimensional high spectrum image of person to be identified and large database concept Image is fitted, and obtains fitting image.
Through the above technical solutions, the beneficial effect of the disclosure is:
(1) disclosure can be combined with hyperspectral technique, in conjunction with the characteristics of high light spectrum image-forming, improve image recognition Accuracy.
(2) disclosure carries out image procossing by CEM algorithm, improves the accuracy of image recognition.
(3) method that the disclosure uses big data classification analysis, reduces error, reduces the influence of external factor.
Detailed description of the invention
The Figure of description for constituting a part of this disclosure is used to provide further understanding of the disclosure, and the disclosure is shown Meaning property embodiment and its explanation do not constitute the improper restriction to the disclosure for explaining the application.
Fig. 1 is the structure chart of cervical carcinoma pattern recognition device of the embodiment one based on high light spectrum image-forming technology；
Flow chart when Fig. 2 is data fitting module progress data fitting in embodiment one；
Fig. 3 is the structure chart of cervical carcinoma image identification system of the embodiment two based on high light spectrum image-forming technology
Fig. 4 is the structure chart of Hyperspectral imaging devices in embodiment two.
Specific embodiment
The disclosure is described further with embodiment with reference to the accompanying drawing.
It is noted that following detailed description is all illustrative, it is intended to provide further instruction to the disclosure.Unless another It indicates, all technical and scientific terms that the disclosure uses have logical with disclosure person of an ordinary skill in the technical field The identical meanings understood.
It should be noted that term used herein above is merely to describe specific embodiment, and be not intended to restricted root According to the illustrative embodiments of the application.As used herein, unless the context clearly indicates otherwise, otherwise singular Also it is intended to include plural form, additionally, it should be understood that, when in the present specification using term "comprising" and/or " packet Include " when, indicate existing characteristics, step, operation, device, component and/or their combination.
Embodiment one
The present embodiment provides a kind of cervical carcinoma pattern recognition device based on high light spectrum image-forming technology, please refers to attached drawing 1, should Device includes data acquisition module, Database module, neural network module, 3-D image establishes module and data are quasi- Mold block, in which:
The data acquisition module, for obtain normal person EO-1 hyperion uterine neck image and each period cervical carcinogenesis patient EO-1 hyperion uterine neck image and person to be identified EO-1 hyperion uterine neck image.
The Database module, for the EO-1 hyperion uterine neck image and each period uterine neck using obtained normal person The EO-1 hyperion uterine neck image of canceration patient, establishes the large database concept of normal person and the high spectrum image in patient's each period.
The neural network module carries out special for the EO-1 hyperion uterine neck image to different times cervical carcinogenesis patient Sign is extracted and classification, separately designs a BP neuroid to each classification.
Specifically, the neural network module is specifically used for:
Using the similitude of image information, feature is carried out to the EO-1 hyperion uterine neck image of different times cervical carcinogenesis patient and is mentioned It takes；
The reflection spectrum curve of each pixel is generated according to the feature of extraction using map sorting algorithm；
Based on the reflection spectrum curve of each pixel, the BP neuroid of every class image is established.
The BP neuroid is the learning process of error-duration model error backpropagation algorithm, by the forward-propagating of information With two process compositions of backpropagation of error.Each neuron of input layer is responsible for receiving from extraneous input information, and transmits Give middle layer each neuron；Middle layer is internal information process layer, is responsible for information transformation, according to the demand of information change ability, Middle layer can be designed as single hidden layer or more hidden layer configurations；The last one hidden layer is transmitted to the information of each neuron of output layer, After after further treatment, the forward-propagating treatment process that once learns is completed, by output layer outwardly output information processing result. When reality output and desired output are not inconsistent, into the back-propagation phase of error.Error is by output layer, by under error gradient The mode of drop corrects each layer weight, to hidden layer, the layer-by-layer anti-pass of input layer.Information forward-propagating and error in cycles reversely passes Process is broadcast, is the process of process and neural network learning training that each layer weight constantly adjusts, this process is performed until net Until the error of network output is reduced to acceptable degree or preset study number.
The 3-D image establishes module, for utilizing established BP neuroid to the EO-1 hyperion palace of person to be identified Neck image is handled, and the three-dimensional high spectrum image of person to be identified is established.
The data fitting module, for using CEM algorithm by the three-dimensional high spectrum image and large database concept of person to be identified Middle uterine neck image is fitted, and obtains fitting image.
Attached drawing 2 is please referred to, the data fitting module is specifically used for:
Obtain the three-dimensional high spectrum image of person to be identified；
The three-dimensional high spectrum image of obtained person to be identified is pre-processed, including two dimensionization and normalization, is obtained pre- Treated hyperspectral image data r (L*N)；
According to hyperspectral image data r, the autocorrelation matrix R of image is acquired, and is inverted to it, auto-correlation square is obtained The inverse matrix of battle array；
Determine target optical spectrum vector d (d size is L*1)；
FIR filter is designed, the filter of FIR filter is acquired in conjunction with the prior information of the object vector got from library of spectra Wave vector；
Pretreated hyperspectral image data r is passed through into FIR filter, the image data y after being fitted, expression Formula are as follows:
In the present embodiment, the expression formula of the autocorrelation matrix R are as follows:
Wherein, riFor the hyperspectral image data after two dimensionization；N is pixel number；
The design formula of the FIR linear filter are as follows:
Wherein, R is the autocorrelation matrix of matrix r；D is target optical spectrum information to be detected.
The expression formula of the FIR linear filter are as follows:
W=[w1, w2..., wL]T
Wherein, wiFor the filter factor under different pixels.
The cervical carcinoma pattern recognition device based on high light spectrum image-forming technology that the present embodiment proposes, can be with hyperspectral technique It combines, in conjunction with the characteristics of high light spectrum image-forming, improves the accuracy of image recognition；Image procossing is carried out by CEM algorithm, is mentioned The high accuracy of image recognition；Using the method for big data classification analysis, reduce error, reduces the shadow of external factor It rings.
Embodiment two
The present embodiment provides a kind of cervical carcinoma image identification system based on high light spectrum image-forming technology, using high light spectrum image-forming Technology generates the uterine neck image of EO-1 hyperion, and the curve of spectrum of uterine neck image is generated using map combination algorithm, establishes large database concept, The uterine neck image of person to be identified is fitted with big data.
Attached drawing 3 is please referred to, the cervical carcinoma image identification system includes Hyperspectral imaging devices, processor and display dress It sets, in which:
The Hyperspectral imaging devices, for generate respectively normal person EO-1 hyperion uterine neck image and each period cervical carcinoma Become the EO-1 hyperion uterine neck image of patient, and generate the EO-1 hyperion uterine neck image of person to be identified, extremely by the uterine neck image transmitting of generation Processor.
The processor, for obtaining the EO-1 hyperion uterine neck image of normal person and the bloom of each period cervical carcinogenesis patient Uterine neck image is composed, large database concept is established；For each uterine neck image, a BP neuroid is constructed respectively；It obtains to be identified The EO-1 hyperion uterine neck image of person is handled using EO-1 hyperion uterine neck image of the BP neuroid to person to be identified, establish to The three-dimensional high spectrum image of identification person intends the three-dimensional high spectrum image of person to be identified and uterine neck image in large database concept It closes, obtains fitting image.
The display device for the fitting image in reading processor, and is shown.
Attached drawing 4 is please referred to, the Hyperspectral imaging devices include light source, camera lens I, slit, camera lens II, grating and fluorescence Screen emits on light source to human body uterus neck, and camera lens I, slit and camera lens II of the cell tissue reflected light by imaging reflex to On grating, after grating, reflected light is imaged on fluorescent screen by wavelength color.
The Hyperspectral imaging devices that the present embodiment proposes, using hyperspectral technique, when light source is radiated at human body uterus neck When, cell tissue reflected light pass through imaging lens, slit, then via optical devices such as gratings after, imaged in by wavelength color glimmering On optical screen.
In the present embodiment, the processor includes data acquisition module, Database module, neural network mould Block, 3-D image establish module and data fitting module, in which:
The data acquisition module, uterine neck image and each period cervical carcinogenesis for obtaining the normal person of EO-1 hyperion are suffered from The uterine neck image of person and the EO-1 hyperion uterine neck image of person to be identified.
The Database module, uterine neck image and each period palace for the normal person using obtained EO-1 hyperion Neck cancer becomes the uterine neck image of patient, establishes the large database concept of normal person and the high spectrum image in patient's each period.
The neural network module carries out feature extraction and classification for the high spectrum image to different times, right Each classification separately designs a BP neuroid.
Specifically, the neural network module is specifically used for:
Using the similitude of image information, feature extraction is carried out to the high spectrum image to different times, and utilize map Sorting algorithm generates the reflection spectrum curve of each pixel；
Using the reflection spectrum curve of each pixel, the BP neuroid of every class image is established.
The 3-D image establishes module, for utilizing established BP neuroid to the EO-1 hyperion palace of person to be identified Neck image is handled, and the three-dimensional high spectrum image of person to be identified is established.
The data fitting module, for using CEM algorithm by the three-dimensional high spectrum image and large database concept of person to be identified Middle uterine neck image is fitted, and obtains fitting image.
Attached drawing 2 is please referred to, the data fitting module is specifically used for:
Obtain the three-dimensional high spectrum image of person to be identified；
The three-dimensional high spectrum image of obtained person to be identified is pre-processed, including two dimensionization and normalization, is obtained pre- Treated hyperspectral image data r (L*N)；
According to hyperspectral image data r, the autocorrelation matrix R of image is acquired, and is inverted to it, auto-correlation square is obtained The inverse matrix of battle array；
Determine target optical spectrum vector d (d size is L*1)；
FIR filter is designed, the filter of FIR filter is acquired in conjunction with the prior information of the object vector got from library of spectra Wave vector；
Pretreated hyperspectral image data r is passed through into FIR filter, the image data y after being fitted, expression Formula are as follows:
In the present embodiment, the expression formula of the autocorrelation matrix R are as follows:
Wherein, riFor the hyperspectral image data after two dimensionization；N is pixel number；
The design formula of the FIR linear filter are as follows:
Wherein, R is the autocorrelation matrix of matrix r；D is target optical spectrum information to be detected.
The expression formula of the FIR linear filter are as follows:
W=[w1, w2..., wL]T
Wherein, wiFor the filter factor under different pixels.
The cervical carcinoma image identification system based on high light spectrum image-forming technology that the present embodiment proposes, can be with hyperspectral technique It combines, in conjunction with the characteristics of high light spectrum image-forming, improves the accuracy of image recognition；Image procossing is carried out by CEM algorithm, is mentioned The high accuracy of image recognition；Using the method for big data classification analysis, reduce error, reduces the shadow of external factor It rings.
Although above-mentioned be described in conjunction with specific embodiment of the attached drawing to the disclosure, model not is protected to the disclosure The limitation enclosed, those skilled in the art should understand that, on the basis of the technical solution of the disclosure, those skilled in the art are not Need to make the creative labor the various modifications or changes that can be made still within the protection scope of the disclosure.

Claims (10)

1. a kind of cervical carcinoma pattern recognition device based on high light spectrum image-forming technology, characterized in that include:
Data acquisition module, for obtaining the EO-1 hyperion uterine neck image of normal person and the EO-1 hyperion of each period cervical carcinogenesis patient Uterine neck image and the EO-1 hyperion uterine neck image of person to be identified；
Database module, for utilizing the obtained EO-1 hyperion uterine neck image of normal person and each period cervical carcinogenesis patient EO-1 hyperion uterine neck image, establish the large database concept of EO-1 hyperion uterine neck image；
Neural network module, for different times cervical carcinogenesis patient EO-1 hyperion uterine neck image carry out feature extraction and Classification, and the BP neuroid of each classification is established respectively；
3-D image establishes module, for utilize established BP neuroid to the EO-1 hyperion uterine neck image of person to be identified into Row processing, establishes the three-dimensional high spectrum image of person to be identified；
Data fitting module, for using CEM algorithm by uterine neck figure in the three-dimensional high spectrum image of person to be identified and large database concept As being fitted, fitting image is obtained.
2. the cervical carcinoma pattern recognition device according to claim 1 based on high light spectrum image-forming technology, characterized in that described Neural network module is specifically used for:
Using the similitude of image information, feature extraction is carried out to the EO-1 hyperion uterine neck image of different times cervical carcinogenesis patient；
The reflection spectrum curve of each pixel is generated according to the feature of extraction using map sorting algorithm；
Based on the reflection spectrum curve of each pixel, the BP neuroid of every class image is established.
3. the cervical carcinoma pattern recognition device according to claim 1 based on high light spectrum image-forming technology, characterized in that described Data fitting module is specifically used for:
Obtain the three-dimensional high spectrum image of person to be identified；
The three-dimensional high spectrum image of obtained person to be identified is pre-processed, pretreated hyperspectral image data is obtained；
According to hyperspectral image data, the autocorrelation matrix of image is acquired, and is inverted to it, the inverse of autocorrelation matrix is obtained Matrix；
Determine target optical spectrum vector；
Design FIR filter, in conjunction with the object vector got from library of spectra prior information acquire the filtering of FIR filter to Amount；
Pretreated hyperspectral image data is passed through into FIR filter, the image data after being fitted.
4. the cervical carcinoma pattern recognition device according to claim 3 based on high light spectrum image-forming technology, characterized in that described The design formula of FIR linear filter are as follows:
Wherein, R is the autocorrelation matrix of matrix r；D is target optical spectrum information to be detected；
The expression formula of the FIR linear filter are as follows:
W=[w1, w2..., wL]T
Wherein, wiFor the filter factor under different pixels.
5. the cervical carcinoma pattern recognition device according to claim 4 based on high light spectrum image-forming technology, characterized in that described The expression formula of autocorrelation matrix R are as follows:
Wherein, riFor the hyperspectral image data after two dimensionization；N is pixel number.
6. a kind of cervical carcinoma image identification system based on high light spectrum image-forming technology, characterized in that including Hyperspectral imaging devices And processor；
The Hyperspectral imaging devices, EO-1 hyperion uterine neck image and each period cervical carcinogenesis for generating normal person respectively are suffered from The uterine neck image of person, and the EO-1 hyperion uterine neck image of person to be identified is generated, by the EO-1 hyperion uterine neck image transmitting of generation to processing Device；
The processor, for obtaining the EO-1 hyperion uterine neck image of normal person and the uterine neck figure of each period cervical carcinogenesis patient Picture establishes large database concept；Feature extraction and classification are carried out to the EO-1 hyperion uterine neck image of different times cervical carcinogenesis patient, and divided The BP neuroid for not establishing each classification, using BP neuroid to the EO-1 hyperion uterine neck image of person to be identified at Reason, establishes the three-dimensional high spectrum image of person to be identified, by uterine neck figure in the three-dimensional high spectrum image of person to be identified and large database concept As being fitted, fitting image is obtained.
7. the cervical carcinoma image identification system according to claim 6 based on high light spectrum image-forming technology, characterized in that described Hyperspectral imaging devices include light source, camera lens I, slit, camera lens II, grating and fluorescent screen, transmitting light source to human body uterus neck On, human body reflected light reflexes on grating by camera lens I, slit and camera lens II, and after grating, reflected light is imaged by wavelength color In on fluorescent screen.
8. the cervical carcinoma image identification system according to claim 6 based on high light spectrum image-forming technology, characterized in that described Processor includes:
Data acquisition module, for obtaining the EO-1 hyperion uterine neck image of normal person and the EO-1 hyperion of each period cervical carcinogenesis patient Uterine neck image and the EO-1 hyperion uterine neck image of person to be identified；
Database module, for utilizing the obtained EO-1 hyperion uterine neck image of normal person and each period cervical carcinogenesis patient EO-1 hyperion uterine neck image, establish the large database concept of EO-1 hyperion uterine neck image；
Neural network module, for different times cervical carcinogenesis patient EO-1 hyperion uterine neck image carry out feature extraction and Classification, and the BP neuroid of each classification is established respectively；
3-D image establishes module, for utilize established BP neuroid to the EO-1 hyperion uterine neck image of person to be identified into Row processing, establishes the three-dimensional high spectrum image of person to be identified；
Data fitting module, for using CEM algorithm by uterine neck figure in the three-dimensional high spectrum image of person to be identified and large database concept As being fitted, fitting image is obtained.
9. the cervical carcinoma image identification system according to claim 8 based on high light spectrum image-forming technology, characterized in that described Neural network module is specifically used for:
Using the similitude of image information, feature extraction is carried out to the EO-1 hyperion uterine neck image of different times cervical carcinogenesis patient；
The reflection spectrum curve of each pixel is generated according to the feature of extraction using map sorting algorithm；
Based on the reflection spectrum curve of each pixel, the BP neuroid of every class image is established.
10. the cervical carcinoma image identification system according to claim 8 based on high light spectrum image-forming technology, characterized in that institute Data fitting module is stated to be specifically used for:
Obtain the three-dimensional high spectrum image of person to be identified；
The three-dimensional high spectrum image of obtained person to be identified is pre-processed, pretreated hyperspectral image data is obtained；
According to hyperspectral image data, the autocorrelation matrix of image is acquired, and is inverted to it, the inverse of autocorrelation matrix is obtained Matrix；
Determine target optical spectrum vector；
Design FIR filter, in conjunction with the object vector got from library of spectra prior information acquire the filtering of FIR filter to Amount；
Pretreated hyperspectral image data is passed through into FIR filter, the image data after being fitted.
CN201910815815.8A 2019-08-30 2019-08-30 A kind of cervical carcinoma pattern recognition device and system based on high light spectrum image-forming technology Pending CN110517258A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910815815.8A CN110517258A (en) 2019-08-30 2019-08-30 A kind of cervical carcinoma pattern recognition device and system based on high light spectrum image-forming technology

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910815815.8A CN110517258A (en) 2019-08-30 2019-08-30 A kind of cervical carcinoma pattern recognition device and system based on high light spectrum image-forming technology

Publications (1)

Publication Number Publication Date
CN110517258A true CN110517258A (en) 2019-11-29

Family

ID=68629774

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910815815.8A Pending CN110517258A (en) 2019-08-30 2019-08-30 A kind of cervical carcinoma pattern recognition device and system based on high light spectrum image-forming technology

Country Status (1)

Country Link
CN (1) CN110517258A (en)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20020146160A1 (en) * 2001-01-19 2002-10-10 Parker Mary F. Method and apparatus for generating two-dimensional images of cervical tissue from three-dimensional hyperspectral cubes
CN106097355A (en) * 2016-06-14 2016-11-09 山东大学 The micro-Hyperspectral imagery processing method of gastroenteric tumor based on convolutional neural networks
CN107064019A (en) * 2017-05-18 2017-08-18 西安交通大学 The device and method for gathering and splitting for dye-free pathological section high spectrum image
CN108596213A (en) * 2018-04-03 2018-09-28 中国地质大学（武汉） A kind of Classification of hyperspectral remote sensing image method and system based on convolutional neural networks
CN109766858A (en) * 2019-01-16 2019-05-17 中国人民解放军国防科技大学 Three-dimensional convolution neural network hyperspectral image classification method combined with bilateral filtering

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20020146160A1 (en) * 2001-01-19 2002-10-10 Parker Mary F. Method and apparatus for generating two-dimensional images of cervical tissue from three-dimensional hyperspectral cubes
CN106097355A (en) * 2016-06-14 2016-11-09 山东大学 The micro-Hyperspectral imagery processing method of gastroenteric tumor based on convolutional neural networks
CN107064019A (en) * 2017-05-18 2017-08-18 西安交通大学 The device and method for gathering and splitting for dye-free pathological section high spectrum image
CN108596213A (en) * 2018-04-03 2018-09-28 中国地质大学（武汉） A kind of Classification of hyperspectral remote sensing image method and system based on convolutional neural networks
CN109766858A (en) * 2019-01-16 2019-05-17 中国人民解放军国防科技大学 Three-dimensional convolution neural network hyperspectral image classification method combined with bilateral filtering

Non-Patent Citations (6)

* Cited by examiner, † Cited by third party
Title
孙俊 等: "高光谱图像技术在掺假大米检测中的应用", 《农业工程学报》 *
寻丽娜 等: "基于CEM的高光谱图像小目标检测算法", 《光电工程》 *
徐德忠 主编: "《循证医学入门 临床科研方法与实例评价》", 31 May 2006, 第四军医大学出版社 *
邓小玲 等: "基于主成分分析和BP神经网络的柑橘黄龙病诊断技术", 《光子学报》 *
郑欣 等: "基于神经网络的显微高光谱乳腺癌组织图像研究", 《第二军医大学学报》 *
陈华 等: "基于神经网络的三维宽场显微图像复原研究", 《光子学报》 *

Similar Documents

Publication Publication Date Title
US11991344B2 (en) Systems, methods and apparatuses for stereo vision and tracking
CN105675549B (en) A kind of Portable rural crop parameter measurement and growing way intellectual analysis device and method
CN105651384B (en) A kind of all-optical information acquisition system
Neal et al. Measuring shape
US10706262B2 (en) Intelligent body measurement
CN110443882A (en) Light field microscopic three-dimensional method for reconstructing and device based on deep learning algorithm
CN106934394B (en) Dual wavelength image acquisition system and method
CN107525588A (en) A kind of method for fast reconstruction of the double camera spectrum imaging system based on GPU
JP7462890B2 (en) Method and system for calibrating a plenoptic camera system - Patents.com
CN105631859B (en) Three-degree-of-freedom bionic stereo visual system
CN114241031A (en) Fish body ruler measurement and weight prediction method and device based on double-view fusion
CN106952341A (en) The underwater scene three-dimensional point cloud method for reconstructing and its system of a kind of view-based access control model
CN108446596A (en) Iris 3D 4 D datas acquisition system based on Visible Light Camera matrix and method
CN108363964A (en) A kind of pretreated wrinkle of skin appraisal procedure and system
CN108520230A (en) A kind of 3D four-dimension hand images data identification method and equipment
CN115115776A (en) Single image three-dimensional human body reconstruction method and device based on shadow
Yu et al. An intelligent measurement scheme for basic characters of fish in smart aquaculture
Xu et al. Self-supervised monocular depth estimation with 3-d displacement module for laparoscopic images
Leménager et al. Studying flowers in 3D using photogrammetry
Zheng et al. Tomato recognition and localization method based on improved YOLOv5n-seg model and binocular stereo vision
CN110517258A (en) A kind of cervical carcinoma pattern recognition device and system based on high light spectrum image-forming technology
Hirai et al. High-dynamic-range spectral imaging system for omnidirectional scene capture
Sadafi et al. Red blood cells segmentation: a fully convolutional network approach
CN111239044A (en) Cell detection method, device and system
CN110147773A (en) A kind of remote sensing images recognition methods

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