CN111160204B - Geological radar image recognition method and system based on principal component analysis BP neural network - Google Patents
Geological radar image recognition method and system based on principal component analysis BP neural network Download PDFInfo
- Publication number
- CN111160204B CN111160204B CN201911342790.0A CN201911342790A CN111160204B CN 111160204 B CN111160204 B CN 111160204B CN 201911342790 A CN201911342790 A CN 201911342790A CN 111160204 B CN111160204 B CN 111160204B
- Authority
- CN
- China
- Prior art keywords
- geological radar
- radar image
- sample data
- data set
- neural network
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 56
- 238000000034 method Methods 0.000 title claims abstract description 45
- 238000000513 principal component analysis Methods 0.000 title claims abstract description 37
- 238000003708 edge detection Methods 0.000 claims abstract description 36
- 230000009467 reduction Effects 0.000 claims abstract description 31
- 238000012549 training Methods 0.000 claims abstract description 28
- 230000008030 elimination Effects 0.000 claims abstract description 20
- 238000003379 elimination reaction Methods 0.000 claims abstract description 20
- 230000000877 morphologic effect Effects 0.000 claims abstract description 20
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims abstract description 18
- 125000004122 cyclic group Chemical group 0.000 claims abstract description 14
- 239000011435 rock Substances 0.000 claims abstract description 14
- 238000002372 labelling Methods 0.000 claims abstract description 10
- 230000008569 process Effects 0.000 claims description 16
- 238000004590 computer program Methods 0.000 claims description 14
- 239000011159 matrix material Substances 0.000 claims description 12
- 238000001514 detection method Methods 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 9
- 238000001914 filtration Methods 0.000 claims description 8
- 238000010276 construction Methods 0.000 claims description 6
- 239000000945 filler Substances 0.000 claims description 6
- 230000005540 biological transmission Effects 0.000 claims description 4
- 230000009977 dual effect Effects 0.000 claims description 4
- 230000002708 enhancing effect Effects 0.000 claims description 4
- 238000000605 extraction Methods 0.000 claims description 4
- 238000009499 grossing Methods 0.000 claims description 4
- 230000004044 response Effects 0.000 claims description 4
- 230000001629 suppression Effects 0.000 claims description 4
- 230000000007 visual effect Effects 0.000 claims description 4
- 230000001066 destructive effect Effects 0.000 claims description 3
- 230000002401 inhibitory effect Effects 0.000 claims 1
- 238000010586 diagram Methods 0.000 description 7
- 230000005856 abnormality Effects 0.000 description 6
- 239000010410 layer Substances 0.000 description 5
- 230000008859 change Effects 0.000 description 4
- 230000007547 defect Effects 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 230000002159 abnormal effect Effects 0.000 description 2
- 238000010521 absorption reaction Methods 0.000 description 2
- 230000002411 adverse Effects 0.000 description 2
- 239000003673 groundwater Substances 0.000 description 2
- 239000011229 interlayer Substances 0.000 description 2
- 238000011835 investigation Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000009933 burial Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000013277 forecasting method Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000001902 propagating effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/13—Satellite images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/28—Quantising the image, e.g. histogram thresholding for discrimination between background and foreground patterns
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/30—Noise filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Multimedia (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computing Systems (AREA)
- Evolutionary Biology (AREA)
- Mathematical Physics (AREA)
- Molecular Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Remote Sensing (AREA)
- Astronomy & Astrophysics (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention provides a geological radar image recognition method and system based on principal component analysis (BP) neural network. The geological radar image identification method based on principal component analysis BP neural network comprises labeling a label of a geological radar image, wherein the label comprises complete rock, a fault fracture zone, a water-rich zone and a karst cave; sequentially carrying out noise elimination, binarization and morphological edge detection on the geological radar image marked with the label to obtain a digital image, and forming a sample data set; reducing the dimension of the sample data set by using a principal component analysis algorithm, and simultaneously keeping the characteristic of the sample data set with the greatest contribution to the difference; performing cyclic training on the BP neural network by using the sample data set after dimension reduction; and receiving the geological radar image in real time, sequentially carrying out noise elimination, binarization, morphological edge detection and dimension reduction, inputting the processed geological radar image into a BP neural network after training is completed, and outputting a geological radar image recognition result.
Description
Technical Field
The invention belongs to the field of geological radar image processing, and particularly relates to a geological radar image recognition method and system based on principal component analysis (BP) neural network.
Background
The statements in this section merely provide background information related to the present disclosure and may not necessarily constitute prior art.
In the prior advanced geological forecasting methods of a plurality of tunnels, the geological radar has the characteristics of high efficiency, no damage to a target, high resolution of detected data, stronger anti-interference capability and the like, and is widely applied. The principle of geological radar detection is that when an abnormal medium and surrounding medium have electrical difference, pulse electromagnetic waves emitted by the geological radar are reflected when propagating to an abnormal interface, and reflected signals are received and recorded by a receiving antenna. By analyzing the reflected signal, information such as the spatial position of the abnormality and the depth of burial can be deduced.
The inventor finds that in the actual detection process of the geological radar, the operation environment is generally disordered, the radar imaging quality is low due to the influence of noise, and the abnormality is difficult to accurately identify, so that the abnormality detection accuracy is too dependent on the experience and level of technicians; and when the tunnel is longer or the data volume is larger, the anomaly identification often needs to consume a great deal of time and labor, which limits the application and popularization of the geological radar technology to a certain extent.
Disclosure of Invention
In order to solve the problems, the invention provides a geological radar image recognition method and a geological radar image recognition system based on principal component analysis BP neural network, which realize intelligent recognition of geological radar image anomalies, greatly improve recognition accuracy, avoid the defect of recognition by experience, save time and improve efficiency.
In order to achieve the above purpose, the present invention adopts the following technical scheme:
the first aspect of the invention provides a geological radar image recognition method based on principal component analysis (BP) neural network, which comprises the following steps:
labeling a geological radar image, wherein the label comprises a complete rock, a fault fracture zone, a water-rich zone and a karst cave;
sequentially carrying out noise elimination, binarization and morphological edge detection on the geological radar image marked with the label to obtain a digital image, and forming a sample data set;
reducing the dimension of the sample data set by using a principal component analysis algorithm, and simultaneously keeping the characteristic of the sample data set with the greatest contribution to the difference;
performing cyclic training on the BP neural network by using the sample data set after dimension reduction;
and receiving the geological radar image in real time, sequentially carrying out noise elimination, binarization, morphological edge detection and dimension reduction, inputting the processed geological radar image into a BP neural network after training is completed, and outputting a geological radar image recognition result.
A second aspect of the present invention provides a geological radar image recognition system based on principal component analysis, BP, neural network, comprising:
the image tag labeling module is used for labeling tags of geological radar images, and the tags comprise complete rocks, fault fracture zones, water-rich zones and karst caverns;
the sample data set construction module is used for sequentially carrying out noise elimination, binarization and morphological edge detection on the geological radar image marked with the tag to obtain a digital image and form a sample data set;
the dimension reduction module is used for reducing the dimension of the sample data set by utilizing a principal component analysis algorithm and simultaneously keeping the characteristic of the sample data set with the greatest contribution to the difference;
the BP neural network training module is used for carrying out cyclic training on the BP neural network by utilizing the sample data set after dimension reduction;
the image real-time identification module is used for receiving the geological radar image in real time, sequentially carrying out noise elimination, binarization, morphological edge detection and dimension reduction, inputting the processed image into the BP neural network after training, and outputting a geological radar image identification result.
A third aspect of the present invention provides a computer readable storage medium having stored thereon a computer program which when executed by a processor performs the steps in a method for identifying a geological radar image based on principal component analysis, BP, neural network as described above.
A fourth aspect of the invention provides a computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor executing the program to perform the steps in a method for identifying a geological radar image based on principal component analysis, BP, neural network as described above.
The beneficial effects of the invention are as follows:
the method comprises the steps of sequentially carrying out noise elimination, binarization and morphological edge detection on a geological radar image marked with a label to obtain a digital image, and forming a sample data set; reducing the dimension of the sample data set by using a principal component analysis algorithm, and simultaneously keeping the characteristic of the sample data set with the greatest contribution to the difference; performing cyclic training on the BP neural network by using the sample data set after dimension reduction; finally, receiving the geological radar image in real time, sequentially carrying out noise elimination, binarization, morphological edge detection and dimension reduction treatment, inputting the obtained geological radar image into a BP neural network after training is completed, and outputting a geological radar image recognition result, so that intelligent recognition of geological radar image abnormality is realized, recognition accuracy is greatly improved, the defect of recognition by experience is avoided, time is saved, and efficiency is improved.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the invention.
FIG. 1 is a flowchart of a geological radar image recognition method based on principal component analysis BP neural network provided by an embodiment of the invention;
fig. 2 is a schematic diagram of a BP neural network with a 5-layer structure according to an embodiment of the present invention.
Detailed Description
The invention will be further described with reference to the drawings and examples.
It should be noted that the following detailed description is illustrative and is intended to provide further explanation of the invention. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of exemplary embodiments according to the present invention. As used herein, the singular is also intended to include the plural unless the context clearly indicates otherwise, and furthermore, it is to be understood that the terms "comprises" and/or "comprising" when used in this specification are taken to specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof.
Example 1
As shown in fig. 1, the present embodiment provides a geological radar image recognition method based on principal component analysis BP neural network, which includes:
step S101: labeling geological radar images, wherein the labels comprise complete rocks, fault fracture zones, water-rich zones and karst caverns.
In specific implementation, the waveform characteristics of the geological radar image and the change rules of the detailed characteristics such as the frequency, the amplitude, the phase and the electromagnetic wave energy absorption condition can represent different geological phenomena. Among them, in engineering investigation, common adverse geological phenomena are: fault fracture zone, water rich zone, karst cave and lithology change zone, etc.
The complete rock mass is generally relatively uniform in medium, very small in electrical difference, free of obvious reflection interfaces, and radar images and waveform characteristics are generally expressed as follows: the energy clusters are uniformly distributed or only locally have strong reflection fine bright stripes; the low-amplitude reflection wave group is generally formed, the wave shape is uniform, no clutter reflection exists, and the automatic gain gradient is relatively small.
Faults are destructive geological structures in which broken rock bodies, mud, groundwater and the like usually develop, the medium is extremely uneven and has large electrical differences, and the rock bodies on two sides of the faults often develop joints and folds, so that the medium uniformity is poor. Cracks generally exist in fault affected zones, dikes and weak interlayers, and various non-uniform fillers exist in the cracks, so that the dielectric difference is large. The method is characterized in that fault and crack interface reflection is strong in geological radar images, amplitude near a reflecting surface is remarkably enhanced and changed greatly, energy clusters are unevenly distributed, diffraction and scattering are often generated in a broken belt and a crack belt, waveforms are disordered, a phase axis is staggered, and even the deep part is blurred.
The appearance of the rich water band in the geological radar image is as follows: the geological radar wave generates strong amplitude reflection on the surface of the aquifer, and when the electromagnetic wave penetrates through the aquifer, the electromagnetic wave generates multiple strong reflections with a certain rule, and diffraction and scattering are generated in the water-rich belt.
The karst cave appears in the geological radar image as: the cavity is composed of a plurality of hyperbolic strong reflection waves, and the reflection waves are generally high-amplitude, low-frequency and equidistant multiple reflection wave groups on the side wall of the cavity, and particularly are stronger when no filler exists or the cavity is filled with water.
Step S102: and sequentially carrying out noise elimination, binarization and morphological edge detection on the geological radar image marked with the label to obtain a digital image, and forming a sample data set.
Specifically, before denoising the geological radar image, the method further comprises the following steps: and carrying out image enhancement processing on the original geological radar image, enhancing useful information in the image, and improving the visual effect of the image.
In the implementation, the noise in the original geological radar image is eliminated by adopting low-pass filtering, so that the influence of the noise generated by interference in the acquisition and transmission processes on the radar image is reduced; and the high-pass filtering is adopted to enhance the high-frequency signal information of the target body outline in the geological radar image, so that the useful image characteristics are highlighted.
Specifically, edge extraction is carried out on the geological radar image by using a Canny edge detection algorithm; the process comprises the following steps:
noise reduction using gaussian smoothing;
calculating the gradient intensity and direction of each pixel point in the image, and reserving the maximum gradient value and direction at each point;
eliminating spurious response brought by edge detection by utilizing non-maximum suppression;
determining true and potential edges using dual threshold detection; the specific process of determining the real and potential edges by using the dual-threshold detection is to use a large threshold first to detect the edge points which are believed to be relatively confident. The tracking is then performed along the previously derived gradient direction, using a smaller threshold value during tracking until returning to the original starting point. Thereby obtaining a binary image, each point representing whether it is an edge point;
edge detection is ultimately accomplished by suppressing isolated weak edges.
Wherein the Canny edge detection algorithm is optimal for step-type edges affected by white noise. The Canny edge detection algorithm aims at returning a binary image, wherein a non-zero value represents the existence of an edge in the image, and returns scale and direction information related to the edge.
Step S103: the dimension of the sample data set is reduced using a principal component analysis algorithm while maintaining the features in the sample data set that contribute most to the difference.
In this embodiment, the dimension of the sample data set is reduced by using the principal component analysis algorithm, and the process of maintaining the feature with the greatest contribution to the difference in the sample data set is as follows:
normalizing the samples in the sample data set;
solving a covariance matrix of the sample characteristics;
selecting k maximum eigenvalues to form an eigenvector matrix; wherein k is a positive integer greater than or equal to 2;
the sample data is projected onto a feature vector matrix to determine principal components.
This reduces the original multi-dimensional problem in dimension and thus greatly simplifies it.
Step S104: and performing cyclic training on the BP neural network by using the sample data set after dimension reduction.
For example: the BP neural network of the present embodiment is a 5-layer BP neural network structure, as shown in fig. 2.
Specifically, the process of performing cyclic training on the BP neural network by using the sample data set after dimension reduction is as follows:
initializing a weight and a threshold, wherein the weight and the threshold are random values in a (-1, 1) interval;
the input signal is transmitted forward, the square of the checking error is calculated, so as to correct weights and threshold values, error information is transmitted backward from the output layer, and each weight is corrected to reduce the error;
and when the square error is smaller than the preset target error value, ending the iteration, outputting a weight vector, otherwise, continuing forward propagation of the input signal until the square error is smaller than the preset target error value or the preset iteration times are reached.
Step S105: and receiving the geological radar image in real time, sequentially carrying out noise elimination, binarization, morphological edge detection and dimension reduction, inputting the processed geological radar image into a BP neural network after training is completed, and outputting a geological radar image recognition result.
In the embodiment, the geological radar image marked with the label is subjected to noise elimination, binarization and morphological edge detection in sequence to obtain a digital image, and a sample data set is formed; reducing the dimension of the sample data set by using a principal component analysis algorithm, and simultaneously keeping the characteristic of the sample data set with the greatest contribution to the difference; performing cyclic training on the BP neural network by using the sample data set after dimension reduction; finally, receiving the geological radar image in real time, sequentially carrying out noise elimination, binarization, morphological edge detection and dimension reduction treatment, inputting the obtained geological radar image into a BP neural network after training is completed, and outputting a geological radar image recognition result, so that intelligent recognition of geological radar image abnormality is realized, recognition accuracy is greatly improved, the defect of recognition by experience is avoided, time is saved, and efficiency is improved.
Example two
The embodiment provides a geological radar image recognition system based on principal component analysis BP neural network, which comprises:
(1) The image tag labeling module is used for labeling tags of geological radar images, and the tags comprise complete rocks, fault fracture zones, water-rich zones and karst caverns;
in specific implementation, the waveform characteristics of the geological radar image and the change rules of the detailed characteristics such as the frequency, the amplitude, the phase and the electromagnetic wave energy absorption condition can represent different geological phenomena. Among them, in engineering investigation, common adverse geological phenomena are: fault fracture zone, water rich zone, karst cave and lithology change zone, etc.
The complete rock mass is generally relatively uniform in medium, very small in electrical difference, free of obvious reflection interfaces, and radar images and waveform characteristics are generally expressed as follows: the energy clusters are uniformly distributed or only locally have strong reflection fine bright stripes; the low-amplitude reflection wave group is generally formed, the wave shape is uniform, no clutter reflection exists, and the automatic gain gradient is relatively small.
Faults are destructive geological structures in which broken rock bodies, mud, groundwater and the like usually develop, the medium is extremely uneven and has large electrical differences, and the rock bodies on two sides of the faults often develop joints and folds, so that the medium uniformity is poor. Cracks generally exist in fault affected zones, dikes and weak interlayers, and various non-uniform fillers exist in the cracks, so that the dielectric difference is large. The method is characterized in that fault and crack interface reflection is strong in geological radar images, amplitude near a reflecting surface is remarkably enhanced and changed greatly, energy clusters are unevenly distributed, diffraction and scattering are often generated in a broken belt and a crack belt, waveforms are disordered, a phase axis is staggered, and even the deep part is blurred.
The appearance of the rich water band in the geological radar image is as follows: the geological radar wave generates strong amplitude reflection on the surface of the aquifer, and when the electromagnetic wave penetrates through the aquifer, the electromagnetic wave generates multiple strong reflections with a certain rule, and diffraction and scattering are generated in the water-rich belt.
The karst cave appears in the geological radar image as: the cavity is composed of a plurality of hyperbolic strong reflection waves, and the reflection waves are generally high-amplitude, low-frequency and equidistant multiple reflection wave groups on the side wall of the cavity, and particularly are stronger when no filler exists or the cavity is filled with water.
(2) The sample data set construction module is used for sequentially carrying out noise elimination, binarization and morphological edge detection on the geological radar image marked with the tag to obtain a digital image and form a sample data set;
specifically, before denoising the geological radar image, the method further comprises the following steps: and carrying out image enhancement processing on the original geological radar image, enhancing useful information in the image, and improving the visual effect of the image.
In the implementation, the noise in the original geological radar image is eliminated by adopting low-pass filtering, so that the influence of the noise generated by interference in the acquisition and transmission processes on the radar image is reduced; and the high-pass filtering is adopted to enhance the high-frequency signal information of the target body outline in the geological radar image, so that the useful image characteristics are highlighted.
Specifically, edge extraction is carried out on the geological radar image by using a Canny edge detection algorithm; the process comprises the following steps:
noise reduction using gaussian smoothing;
calculating the gradient intensity and direction of each pixel point in the image, and reserving the maximum gradient value and direction at each point;
eliminating spurious response brought by edge detection by utilizing non-maximum suppression;
determining true and potential edges using dual threshold detection; the specific process of determining the real and potential edges by using the dual-threshold detection is to use a large threshold first to detect the edge points which are believed to be relatively confident. The tracking is then performed along the previously derived gradient direction, using a smaller threshold value during tracking until returning to the original starting point. Thereby obtaining a binary image, each point representing whether it is an edge point;
edge detection is ultimately accomplished by suppressing isolated weak edges.
Wherein the Canny edge detection algorithm is optimal for step-type edges affected by white noise. The Canny edge detection algorithm aims at returning a binary image, wherein a non-zero value represents the existence of an edge in the image, and returns scale and direction information related to the edge.
(3) The dimension reduction module is used for reducing the dimension of the sample data set by utilizing a principal component analysis algorithm and simultaneously keeping the characteristic of the sample data set with the greatest contribution to the difference;
in this embodiment, the dimension of the sample data set is reduced by using the principal component analysis algorithm, and the process of maintaining the feature with the greatest contribution to the difference in the sample data set is as follows:
normalizing the samples in the sample data set;
solving a covariance matrix of the sample characteristics;
selecting k maximum eigenvalues to form an eigenvector matrix; wherein k is a positive integer greater than or equal to 2;
the sample data is projected onto a feature vector matrix to determine principal components.
This reduces the original multi-dimensional problem in dimension and thus greatly simplifies it.
(4) The BP neural network training module is used for carrying out cyclic training on the BP neural network by utilizing the sample data set after dimension reduction;
for example: the BP neural network of the present embodiment is a 5-layer BP neural network structure, as shown in fig. 2.
Specifically, the process of performing cyclic training on the BP neural network by using the sample data set after dimension reduction is as follows:
initializing a weight and a threshold, wherein the weight and the threshold are random values in a (-1, 1) interval;
the input signal is transmitted forward, the square of the checking error is calculated, so as to correct weights and threshold values, error information is transmitted backward from the output layer, and each weight is corrected to reduce the error;
and when the square error is smaller than the preset target error value, ending the iteration, outputting a weight vector, otherwise, continuing forward propagation of the input signal until the square error is smaller than the preset target error value or the preset iteration times are reached.
(5) The image real-time identification module is used for receiving the geological radar image in real time, sequentially carrying out noise elimination, binarization, morphological edge detection and dimension reduction, inputting the processed image into the BP neural network after training, and outputting a geological radar image identification result.
In the embodiment, the geological radar image marked with the label is subjected to noise elimination, binarization and morphological edge detection in sequence to obtain a digital image, and a sample data set is formed; reducing the dimension of the sample data set by using a principal component analysis algorithm, and simultaneously keeping the characteristic of the sample data set with the greatest contribution to the difference; performing cyclic training on the BP neural network by using the sample data set after dimension reduction; finally, receiving the geological radar image in real time, sequentially carrying out noise elimination, binarization, morphological edge detection and dimension reduction treatment, inputting the obtained geological radar image into a BP neural network after training is completed, and outputting a geological radar image recognition result, so that intelligent recognition of geological radar image abnormality is realized, recognition accuracy is greatly improved, the defect of recognition by experience is avoided, time is saved, and efficiency is improved.
Example III
The present embodiment provides a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps in a geological radar image recognition method based on principal component analysis BP neural network as described above.
Example IV
The embodiment provides a computer device, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the steps in the geological radar image identification method based on principal component analysis BP neural network are realized when the processor executes the program.
It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of a hardware embodiment, a software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, magnetic disk storage, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
Those skilled in the art will appreciate that implementing all or part of the above-described methods in accordance with the embodiments may be accomplished by way of a computer program stored on a computer readable storage medium, which when executed may comprise the steps of the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a Random access Memory (Random AccessMemory, RAM), or the like.
The above description is only of the preferred embodiments of the present invention and is not intended to limit the present invention, but various modifications and variations can be made to the present invention by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (7)
1. A geological radar image recognition method based on principal component analysis BP neural network is characterized by comprising the following steps:
labeling a geological radar image, wherein the label comprises a complete rock, a fault fracture zone, a water-rich zone and a karst cave;
the complete rock mass medium is relatively uniform, has small electrical difference and no obvious reflection interface,
faults are destructive geological structures, various heterogeneous fillers are also arranged in cracks, and dielectric differences are large; the method is characterized in that fault and crack interface reflection is strong in geological radar images, amplitude near a reflecting surface is remarkably enhanced and changed greatly, energy clusters are unevenly distributed, diffraction and scattering are often generated in a broken belt and a crack belt, waveforms are disordered, a phase axis is staggered, and even the deep part is blurred;
the appearance of the rich water band in the geological radar image is as follows: the geological radar wave generates strong amplitude reflection on the surface of the aquifer, and when the electromagnetic wave penetrates through the aquifer, the electromagnetic wave generates multiple strong reflections with a certain rule, and diffraction and scattering are generated in the water-rich belt;
the karst cave appears in the geological radar image as: the cavity is composed of a plurality of hyperbolic strong reflection waves, wherein the side wall of the cavity is provided with high-amplitude, low-frequency and equidistant multiple reflection wave groups, and the reflection waves are stronger when no filler exists or the cavity is filled with water;
sequentially carrying out noise elimination, binarization and morphological edge detection on the geological radar image marked with the label to obtain a digital image, and forming a sample data set;
reducing the dimension of the sample data set by using a principal component analysis algorithm, and simultaneously keeping the characteristic of the sample data set with the greatest contribution to the difference;
performing cyclic training on the BP neural network by using the sample data set after dimension reduction;
receiving a geological radar image in real time, sequentially carrying out noise elimination, binarization, morphological edge detection and dimension reduction treatment, inputting the obtained geological radar image into a BP neural network after training is completed, and outputting a geological radar image recognition result;
noise in an original geological radar image is eliminated by adopting low-pass filtering, and the influence of noise generated by interference in the acquisition and transmission processes on the radar image is reduced; the high-pass filtering is adopted to enhance the high-frequency signal information of the target body outline in the geological radar image, so that the useful image characteristics are highlighted;
the process of reducing the dimension of the sample data set by using the principal component analysis algorithm and simultaneously maintaining the features of the sample data set which have the greatest contribution to the difference comprises the following steps:
normalizing the samples in the sample data set;
solving a covariance matrix of the sample characteristics;
selecting k maximum eigenvalues to form an eigenvector matrix; wherein k is a positive integer greater than or equal to 2;
projecting the sample data onto a feature vector matrix to determine a principal component;
performing edge extraction on the geological radar image by using a Canny edge detection algorithm; the process comprises the following steps:
noise reduction using gaussian smoothing;
calculating the gradient intensity and direction of each pixel point in the image, and reserving the maximum gradient value and direction at each point;
eliminating spurious response brought by edge detection by utilizing non-maximum suppression;
determining true and potential edges using dual threshold detection;
edge detection is ultimately accomplished by suppressing isolated weak edges.
2. The method for identifying a geological radar image based on principal component analysis BP neural network according to claim 1, further comprising, before denoising the geological radar image: and carrying out image enhancement processing on the original geological radar image, enhancing useful information in the image, and improving the visual effect of the image.
3. The geological radar image recognition method based on principal component analysis (BP) neural network as claimed in claim 1, wherein the process of performing cyclic training on the BP neural network by using the dimension-reduced sample data set is as follows:
initializing a weight and a threshold, wherein the weight and the threshold are random values in a (-1, 1) interval;
the input signal is transmitted forward, the square of the checking error is calculated, so as to correct weights and threshold values, error information is transmitted backward from the output layer, and each weight is corrected to reduce the error;
and when the square error is smaller than the preset target error value, ending the iteration, outputting a weight vector, otherwise, continuing forward propagation of the input signal until the square error is smaller than the preset target error value or the preset iteration times are reached.
4. A geological radar image recognition system based on a principal component analysis BP neural network, realized by the geological radar image recognition method based on a principal component analysis BP neural network according to claim 1, comprising:
the image tag labeling module is used for labeling tags of geological radar images, and the tags comprise complete rocks, fault fracture zones, water-rich zones and karst caverns;
the sample data set construction module is used for sequentially carrying out noise elimination, binarization and morphological edge detection on the geological radar image marked with the tag to obtain a digital image and form a sample data set;
the dimension reduction module is used for reducing the dimension of the sample data set by utilizing a principal component analysis algorithm and simultaneously keeping the characteristic of the sample data set with the greatest contribution to the difference;
the BP neural network training module is used for carrying out cyclic training on the BP neural network by utilizing the sample data set after dimension reduction;
the image real-time identification module is used for receiving the geological radar image in real time, sequentially carrying out noise elimination, binarization, morphological edge detection and dimension reduction, inputting the processed image into the BP neural network after training, and outputting a geological radar image identification result.
5. The system for identifying a geological radar image based on principal component analysis BP neural network of claim 4, wherein before denoising the geological radar image in the sample dataset construction module, further comprises: performing image enhancement processing on an original geological radar image, enhancing useful information in the image, and improving the visual effect of the image;
or (b)
In the sample data set construction module, noise in an original geological radar image is eliminated by adopting low-pass filtering, and the influence of noise generated by interference in the acquisition and transmission processes on the radar image is reduced; the high-pass filtering is adopted to enhance the high-frequency signal information of the target body outline in the geological radar image, so that the useful image characteristics are highlighted;
or (b)
In the sample data set construction module, edge extraction is carried out on the geological radar image by using a Canny edge detection algorithm; the process comprises the following steps:
noise reduction using gaussian smoothing;
calculating the gradient intensity and direction of each pixel point in the image, and reserving the maximum gradient value and direction at each point;
eliminating spurious response brought by edge detection by utilizing non-maximum suppression;
determining true and potential edges using dual threshold detection;
finally completing edge detection by inhibiting isolated weak edges;
or (b)
In the dimension reduction module, the dimension of the sample data set is reduced by utilizing a principal component analysis algorithm, and the process of keeping the characteristic of the sample data set with the largest contribution to the difference is as follows:
normalizing the samples in the sample data set;
solving a covariance matrix of the sample characteristics;
selecting k maximum eigenvalues to form an eigenvector matrix; wherein k is a positive integer greater than or equal to 2;
projecting the sample data onto a feature vector matrix to determine a principal component;
or (b)
In the BP neural network training module, the process of carrying out cyclic training on the BP neural network by using the sample data set after dimension reduction comprises the following steps:
initializing a weight and a threshold, wherein the weight and the threshold are random values in a (-1, 1) interval;
the input signal is transmitted forward, the square of the checking error is calculated, so as to correct weights and threshold values, error information is transmitted backward from the output layer, and each weight is corrected to reduce the error;
and when the square error is smaller than the preset target error value, ending the iteration, outputting a weight vector, otherwise, continuing forward propagation of the input signal until the square error is smaller than the preset target error value or the preset iteration times are reached.
6. A computer-readable storage medium, on which a computer program is stored, characterized in that the program, when being executed by a processor, implements the steps in the method for identifying geological radar images based on principal component analysis BP neural network according to any one of claims 1-3.
7. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor, when executing the program, implements the steps of the method for identifying a geological radar image based on principal component analysis, BP, neural network according to any one of claims 1-3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911342790.0A CN111160204B (en) | 2019-12-23 | 2019-12-23 | Geological radar image recognition method and system based on principal component analysis BP neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911342790.0A CN111160204B (en) | 2019-12-23 | 2019-12-23 | Geological radar image recognition method and system based on principal component analysis BP neural network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111160204A CN111160204A (en) | 2020-05-15 |
| CN111160204B true CN111160204B (en) | 2024-01-30 |
Family
ID=70558127
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911342790.0A Active CN111160204B (en) | 2019-12-23 | 2019-12-23 | Geological radar image recognition method and system based on principal component analysis BP neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111160204B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113424055B (en) | 2019-10-09 | 2022-09-02 | 山东大学 | Multi-scale tunnel structure disease detection and intelligent diagnosis system and method |
| CN112001885A (en) * | 2020-07-14 | 2020-11-27 | 武汉理工大学 | PCB fault detection method, storage medium and system based on machine vision |
| CN112801035B (en) * | 2021-02-24 | 2023-04-07 | 山东大学 | Knowledge and data dual-drive-based carrying type lithology intelligent identification method and system |
| CN113486971B (en) * | 2021-07-19 | 2023-10-27 | 国网山东省电力公司日照供电公司 | User state identification method and system based on principal component analysis and neural network |
| CN113570139A (en) * | 2021-07-28 | 2021-10-29 | 新疆建筑科学研究院（有限责任公司） | Neural network surrounding rock grading prediction method and system based on geophysical prospecting data |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102722759A (en) * | 2012-05-17 | 2012-10-10 | 河海大学 | Method for predicting power supply reliability of power grid based on BP neural network |
| CN104200284A (en) * | 2014-09-01 | 2014-12-10 | 山东科技大学 | Mine fault structure prediction method |
| CN107092990A (en) * | 2017-05-03 | 2017-08-25 | 西安电子科技大学 | The shield construction ground settlement forecast system and method analyzed based on big data |
| CN108182427A (en) * | 2018-01-30 | 2018-06-19 | 电子科技大学 | A kind of face identification method based on deep learning model and transfer learning |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10146835B2 (en) * | 2016-08-23 | 2018-12-04 | Interana, Inc. | Methods for stratified sampling-based query execution |
-
2019
- 2019-12-23 CN CN201911342790.0A patent/CN111160204B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102722759A (en) * | 2012-05-17 | 2012-10-10 | 河海大学 | Method for predicting power supply reliability of power grid based on BP neural network |
| CN104200284A (en) * | 2014-09-01 | 2014-12-10 | 山东科技大学 | Mine fault structure prediction method |
| CN107092990A (en) * | 2017-05-03 | 2017-08-25 | 西安电子科技大学 | The shield construction ground settlement forecast system and method analyzed based on big data |
| CN108182427A (en) * | 2018-01-30 | 2018-06-19 | 电子科技大学 | A kind of face identification method based on deep learning model and transfer learning |
Non-Patent Citations (1)
| Title |
|---|
| 蒋林权."基于机器视觉的砀山梨表面缺陷识别".《中国优秀硕士学位论文全文数据库 (农业科技辑)》.2019,全文. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111160204A (en) | 2020-05-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111160204B (en) | Geological radar image recognition method and system based on principal component analysis BP neural network | |
| CN104991269B (en) | A kind of margin guide and the full waveform inversion fast method of structural constraint | |
| AU2013289171B2 (en) | System and method for estimating and attenuating noise in seismic data | |
| CN101982835B (en) | Level set method for edge detection of SAR images of airport roads | |
| CN103226832B (en) | Based on the multi-spectrum remote sensing image change detecting method of spectral reflectivity mutation analysis | |
| CN110133643B (en) | Plant root system detection method and device | |
| Gong et al. | Automatic subway tunnel crack detection system based on line scan camera | |
| Zheng et al. | A novel edge detection for buried target extraction after SVD-2D wavelet processing | |
| He et al. | GPR image denoising with NSST-UNET and an improved BM3D | |
| CN114677525B (en) | Edge detection method based on binary image processing | |
| Ma et al. | Rebar Clutter Suppression and Road Defects Localization in GPR B-Scan Images Based on SuppRebar-GAN and EC-Yolov7 Networks | |
| CN107121705A (en) | A kind of ground penetrating radar echo signals Denoising Algorithm compared based on automatic anti-phase correction and kurtosis value | |
| CN105652256A (en) | High-frequency surface wave radar TBD method based on polarization information | |
| CN117351321A (en) | Single-stage lightweight subway lining cavity recognition method and related equipment | |
| Liu et al. | The target detection for GPR images based on curve fitting | |
| Li et al. | Fracture extraction from FMI based on multiscale mathematical morphology | |
| JPH116879A (en) | Method and device for three dimensional survey | |
| CN112558156B (en) | Processing method and processing system for seismic strong amplitude anomalies | |
| CN115345216A (en) | FMCW radar interference elimination method fusing prior information | |
| CN111399038B (en) | Slope parameter extraction method and device and computer readable storage medium | |
| Ouadfeul et al. | Noise attenuation from GPR data using wavelet transform and artificial neural network | |
| Wu et al. | Non-destructive location technology of concealed power facilities based on a two-step detection method | |
| CN117310696B (en) | Self-focusing backward projection imaging method and device for ground penetrating radar | |
| CN117152622B (en) | Boundary optimization model training, boundary optimization method, device, equipment and medium | |
| CN112379427B (en) | Method and device for suppressing surface wave noise |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |