CN106503649B - A kind of short-wave signal detection recognition method based on computer vision - Google Patents
A kind of short-wave signal detection recognition method based on computer vision Download PDFInfo
- Publication number
- CN106503649B CN106503649B CN201610917977.9A CN201610917977A CN106503649B CN 106503649 B CN106503649 B CN 106503649B CN 201610917977 A CN201610917977 A CN 201610917977A CN 106503649 B CN106503649 B CN 106503649B
- Authority
- CN
- China
- Prior art keywords
- short
- frequency
- wave signal
- time
- signal
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
- G06F2218/10—Feature extraction by analysing the shape of a waveform, e.g. extracting parameters relating to peaks
-
- 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
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Signal Processing (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Image Analysis (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention discloses a kind of short-wave signal detection recognition methods based on computer vision, according to the classifier building process of the time-frequency matrix character of short-wave signal and machine learning, invent a kind of short-wave signal based on computer vision detection and method of discrimination, compared to traditional signal detection discriminant approach based on manual type and other existing signal detection method of discrimination, this method can realize automatic detection and identification of the computer to short-wave signal, more other signal detection method of discrimination simultaneously, noise circumstance locating for signal is required lower, suitable for noise down to the channel circumstance of 0dB.By having carried out a large amount of real data tests under the conditions of different channels, the detection of short-wave signal differentiates accuracy 90% or more.
Description
Technical field
The present invention relates to technical field of computer vision and machine learning field, more particularly, to one kind to based on computer
The short-wave signal feature that vision is extracted classifies via SVM (Support Vector Machine) and then realizes short-wave signal positioning
With the detection method of common type identification.
Background technique
Short-wave signal is one of signal set most commonly seen in radio communication.According to Consultative Committee on International Radio (CCIR)
(CCIR) division, shortwave are defined as wavelength in 100m~10m, and frequency is the electromagnetic wave of 3MHz~30MHz, is carried out using shortwave
Radio communication become short wave communication, also known as high frequency (HF) communicate.In usually used, in order to make full use of shortwave closely
The advantages of communication, the frequency range of short wave communication actual use are 1.5MHz~30MHz.
Technology of the present invention is the detection identification of short-wave signal based on computer vision.Wherein, computer vision
Introducing it is particularly important.
Computer vision refers to be replaced human eye to identify target, tracks, measuring with computer, is similar to realize
The computer technology of human visual function.It mainly contains the content of two aspects of image procossing and pattern-recognition: figure
It is main to realize that the visual information to input computer carries out target position analysis as treatment process, that is, target position fixing process, often
Method includes image denoising, dimensionality reduction etc.；Mode identification procedure refers to the various forms of (numerical value to characterization things or phenomenon
, text and logical relation) information handled and analyzed, to be described, recognize to things or phenomenon, reconciliation of classifying
The process released is the important component of information science and artificial intelligence.Common mode identification procedure is based on machine learning
Classifier realize.It mainly includes data training set building, the feature extraction of training data, the feature training of classifier
And etc., eventually by testing classification device for known and unknown data classifier performance, the performance superiority and inferiority of classification of assessment device.
Currently, short-wave signal detection also mainly in a manner of manually scouting based on, i.e., by business personnel manually to short-wave reception
The signal that machine receives is listened to.When listening to a certain suspicious or neutral signal, it is also necessary to by instrument to signal wave
Shape carries out visual analyzing, and receives control accordingly according to acquired signal parameter development and keep equipment.The process is to business personnel
Short-wave signal listen to that skill requirement is very high, and artificial training cost is very big.
The deficiencies of present invention is directed to the high labor cost of traditional artificial short-wave signal detection method, time-consuming is high, proposes
A kind of short-wave signal detection recognition method based on computer vision, by computer to common class signal in short-wave signal
The positioning and analysis automated realizes the autonomous type identification of computer of short-wave signal, effectively increases signal detection
Timeliness.
Summary of the invention
The technical problem to be solved by the present invention is to for the short-wave signal artificial detection identification in noisy communication channel
Problem time-consuming high, complexity is high provides a kind of computer vision feature modeling by signal and is determined by SVM classifier
Short-wave signal automatic detection and knowledge method for distinguishing are effectively improved in the case where guaranteeing short-wave signal detection identification accuracy
The timeliness of the process.
The technical scheme of the invention to solve the technical problem is: a kind of short-wave signal based on computer vision
Detection and localization and time-frequency characteristics extract, and the SVM classifier based on machine learning, match to the time-frequency characteristics of signal, most
The automatic detection and identification of short-wave signal are realized eventually.Technical solution of the present invention process is as shown in Figure 1.
Method of the invention comprises the following steps:
1) frequency when time-frequency is obtained according to frequency direction every 300kHz piecemeal to signal time-frequency data matrix to be detected
According to submatrix.
2) to each time-frequency data submatrix, following signal framing detection is carried out, concrete operations are as follows:
2.1) clock synchronization frequency projects on the frequency axis according to submatrix obtains one-dimensional row and vector V, and V is represented to be believed in frequency direction
Number energy it is cumulative, as shown in Fig. 2, wherein Fig. 2 (a) is time-frequency data submatrix figure, Fig. 2 (b) is based on one-dimensional row and vector V
The drop shadow curve of drafting.
2.2) distribution histogram of each element in V is analyzed.Locating threshold T=(V_med-V_min)+V_med,
Middle V_med indicates that the mode of V element, V_min indicate the minimum value of element in V.
2.3) it combines the locating threshold T in previous step as thresholding and 2.1) drop shadow curve in step, reference projection is bent
It is higher than the frequency separation of thresholding in line, corresponding frequency separation in time-frequency matrix is intercepted, currently processed time-frequency is obtained
The set of the time frequency signal submatrix of short-wave signal present in data submatrix.
3) the corresponding time frequency signal submatrix set of each time-frequency data submatrix obtained first two steps merges, and is made
For the time frequency signal set of matrices of the whole short-wave signals for including in the signal time-frequency data matrix to be detected of input.
4) set constituted to the time frequency signal matrix of each short-wave signal, carries out signal characteristic abstraction one by one.
4.1) kurtosis feature
The time frequency signal matrix of short-wave signal is projected on the frequency axis and obtains one-dimension array Y.Y is carried out at windowed traversal
Reason, as shown in Figure 3.The difference of the corresponding numerical value of point of proximity is calculated separately, and is substituted into following formula.
The value of point each in Y is replaced with into corresponding Dt, finally obtains kurtosis curve Y '.
It is then grouped sequence: using the zero in kurtosis curve as partitioning standards, kurtosis curve being divided into multiple
Spike group, the corresponding kurtosis curve regions maximum value of statistics each group obtain the row of spike group maximum value sequence progress from big to small
Sequence.
The spike group maximum value sequence sorted from large to small obtained in packet sequencing is divided into two classes by subsequent given threshold,
Threshold value is the maximum value of the corresponding kurtosis curve of maximum sharpness group divided by constant C (empirical value C=10^3), takes a group sequence smaller portions
As local spike group maximum value sequence, and to the part spike group maximum value sequence.
Ratio successively asked to two adjacent values of spike group maximum value sequence, statistics obtain maximum ratio it is corresponding two most
Spike group maximum value sequence is divided into two parts, takes sequence order smaller by big value position using two maximum values as division limits
Part as effective spike group maximum value sequence, the maximum value number that the partial sequence includes is counted, as effective spike
Number, and the result as kurtosis feature is exported；
4.2) spatial distribution characteristic
The time frequency signal matrix of short-wave signal is projected on the frequency axis and obtains one-dimension array Y.Evaluation domain section | V (i) |
∈ [a, b] divides k and (by studying the relationship of k value and resolution ratio, obtains corresponding relationship if section size is k (k=b-a)
Array), the scatterplot quantity in each section is counted, the ratio for the total scatterplot number of scatterplot number Zhan for including in each section is calculated.
It is exported the scatterplot proportionality percentage array in each section as spatial distribution characteristic.
4.3) symmetric characteristics
The time frequency signal matrix of short-wave signal is projected on the frequency axis and obtains one-dimension array Y.Evaluation domain section [a, b],
Thresholding C is set, the part that C is higher than in curve is effective coverage.Wherein, C=0.2* (b-a)+a.It is starting with curve highest point
Starting point and two sides end point are asked as end point in point, position when looking for curve to decline to C or less for the first time to two sides respectively
Distance d_right and d_left.The ratio between d_right and d_left are asked, when symmetrical, the result is close to 1.
4.4) Variance feature
The time frequency signal matrix of short-wave signal is projected on the frequency axis and obtains one-dimension array Y, extracts maximum value in array Y
The corresponding frequency row Y (N) in position calculates the variance of Y (N) row, exports as Variance feature.
4.5) rectangle morphological feature
For each short-wave signal time-frequency data matrix, even rectangular degree Pr is calculated.Calculation method is Pr=As/Ar (As
It is the area of connected domain S in image, Ar is the minimum circumscribed rectangle area for surrounding the connected domain).
It is corresponding to single short-wave signal time-frequency data matrix 4.1) to be integrated to 4.5) feature, obtain the short-wave signal
The eigenmatrix of time-frequency data matrix.
5) 4) each short-wave signal Input matrix obtained is obtained into classifier into preparatory trained SVM classifier
Output is exported according to classifier as a result, short-wave signal type is marked.
It is described 5) in, the SVM classifier trained has been used, based on the extracted kurtosis feature of machine vision, space point
Cloth feature, symmetric characteristics, Variance feature, rectangle morphological feature, identify short-wave signal type
Compared with the prior art, the advantages of the present invention are as follows: traditional short-wave signal detection recognition method relies primarily on hand
Work mode detects, high labor cost, and uses duration.Meanwhile the more already present short-wave signal detection for being related to computer technology
Recognition methods application scenarios are more harsh, very high for the susceptibility of noise, and identification types are more single.The present invention is sufficiently sharp
Short-wave signal time-frequency matrix is carried out manifold with the information of short-wave signal time-frequency matrix in conjunction with computer vision technique
It extracts, and efficient machine learning SVM classifier is selected to be trained and identify.Whole process is compared to traditional short-wave signal
Detection method effectively reduces cost of labor and the used time of detection while guaranteeing accuracy rate, improves short-wave signal detection
The timeliness of identification.
The experimental results showed that the average detected of short-wave signal and identification are just under the noisy channel that noise level lower limit is 0dB
True rate is up to 90% or more.
Detailed description of the invention
Fig. 1 is the flow chart of the method for the present invention；
Fig. 2 is time-frequency data submatrix perspective view on the frequency axis, and wherein Fig. 2 (a) is time-frequency data submatrix figure,
Fig. 2 (b) is the drop shadow curve drawn based on one-dimensional row and vector V.
Fig. 3 is the windowing process schematic diagram of kurtosis feature.
Specific embodiment
The present invention is further elaborated below in conjunction with attached drawing.
The present invention devises a kind of shortwave based on machine vision with identification for the short-wave signal detection in noisy channel
Automatic signal detection recognition methods.It is actual in use, computer by by this method come the noisy short-wave signal to input
Carry out detection identification.Method and step of the invention is as follows:
Step 1: to signal time-frequency data matrix to be detected, according to the every 300kHz piecemeal of frequency direction, when obtaining time-frequency
Frequency is according to submatrix.
Step 2: carrying out following signal framing detection to each time-frequency data submatrix, concrete operations are as follows:
1) clock synchronization frequency projects on the frequency axis according to submatrix obtains one-dimensional row and vector V, V represent signal in frequency direction
Energy adds up, as shown in Fig. 2, wherein Fig. 2 (a) is time-frequency data submatrix figure, Fig. 2 (b) is to be drawn based on one-dimensional row and vector V
The drop shadow curve of system.
2) distribution histogram of each element in V is analyzed.Locating threshold T=(V_med-V_min)+V_med, wherein
V_med indicates that the mode of V element, V_min indicate the minimum value of element in V.
As thresholding and 2.1) 3) locating threshold T in the previous step drop shadow curve in step, reference drop shadow curve are combined
In be higher than thresholding frequency separation, corresponding frequency separation in time-frequency matrix is intercepted, currently processed when frequency is obtained
According to the set of the time frequency signal submatrix of short-wave signal present in submatrix.
Step 3: the corresponding time frequency signal submatrix set of each time-frequency data submatrix that first two steps are obtained merges,
Obtain the time frequency signal set of matrices of the whole short-wave signals for including in signal time-frequency data matrix to be detected as input.
Step 4: carrying out signal characteristic abstraction one by one to the set that the time frequency signal matrix of each short-wave signal is constituted.
1) kurtosis feature
The time frequency signal matrix of short-wave signal is projected on the frequency axis and obtains one-dimension array Y.Y is carried out at windowed traversal
Reason, as shown in Figure 3.The difference of the corresponding numerical value of point of proximity is calculated separately, and is substituted into following formula.
The value of point each in Y is replaced with into corresponding Du, finally obtains kurtosis curve Y '.The value of point each in Y is replaced
For corresponding Du, kurtosis curve Y ' is finally obtained.
It is then grouped sequence: using the zero in kurtosis curve as partitioning standards, kurtosis curve being divided into multiple
Spike group, the corresponding kurtosis curve regions maximum value of statistics each group obtain the row of spike group maximum value sequence progress from big to small
Sequence.
The spike group maximum value sequence sorted from large to small obtained in packet sequencing is divided into two classes by subsequent given threshold,
Threshold value is the maximum value of the corresponding kurtosis curve of maximum sharpness group divided by constant C (empirical value C=10^3), takes a group sequence smaller portions
As local spike group maximum value sequence, and to the part spike group maximum value sequence.
Ratio successively asked to two adjacent values of spike group maximum value sequence, statistics obtain maximum ratio it is corresponding two most
Spike group maximum value sequence is divided into two parts, takes sequence order smaller by big value position using two maximum values as division limits
Part as effective spike group maximum value sequence, the maximum value number that the partial sequence includes is counted, as effective spike
Number, and the result as kurtosis feature is exported；
2) spatial distribution characteristic
The time frequency signal matrix of short-wave signal is projected on the frequency axis and obtains one-dimension array Y.Evaluation domain section | V (i) |
∈ [a, b] divides k and (by studying the relationship of k value and resolution ratio, obtains corresponding relationship if section size is k (k=b-a)
Array), the scatterplot quantity in each section is counted, the ratio for the total scatterplot number of scatterplot number Zhan for including in each section is calculated.
It is exported the scatterplot proportionality percentage array in each section as spatial distribution characteristic.
3) symmetric characteristics
The time frequency signal matrix of short-wave signal is projected on the frequency axis and obtains one-dimension array Y.Evaluation domain section [a, b],
Thresholding C is set, the part that C is higher than in curve is effective coverage.Wherein, C=0.2* (b-a)+a.It is starting with curve highest point
Starting point and two sides end point are asked as end point in point, position when looking for curve to decline to C or less for the first time to two sides respectively
Distance d_right and d_left.The ratio between d_right and d_left are asked, when symmetrical, the result is close to 1.
4) Variance feature
The time frequency signal matrix of short-wave signal is projected on the frequency axis and obtains one-dimension array Y, extracts maximum value in array Y
The corresponding frequency row Y (N) in position calculates the variance of Y (N) row, exports as Variance feature.
5) rectangle morphological feature
For each short-wave signal time-frequency data matrix, even rectangular degree Pr is calculated.Calculation method is Pr=As/Ar (As
It is the area of connected domain S in image, Ar is the minimum circumscribed rectangle area for surrounding the connected domain).
It is corresponding to single short-wave signal time-frequency data matrix 4.1) to be integrated to 4.5) feature, obtain the short-wave signal
The eigenmatrix of time-frequency data matrix.
Step 5: each short-wave signal Input matrix that the 4th step is obtained is obtained into preparatory trained SVM classifier
It exports to classifier, exports according to classifier as a result, short-wave signal type is marked.
In order to examine the performance of method proposed by the invention, by method of the invention to the time-frequency data largely acquired into
Test is gone.Test platform is visual studio 2010, and system configuration is windows 10 (16G memory, CPU i5).
Test process are as follows: the short-wave signal data .dat file that actual samples obtain is input to the C language journey based on the method for the present invention
In sequence, according to the short-wave signal type of detection identification output, the short-wave signal collection of the short-wave signal type of the input data is constructed
It closes, by manual review, counts the accuracy of short-wave signal classification.It is surveyed by carrying out a large amount of time-frequency data to the method for the present invention
Examination, verifies the actual performance of inventive algorithm.Test result shows signal detection anti-noise of the inventive algorithm in noisy channel
Sound ability is strong, can carry out effective detection and localization to short-wave signal in the case where noise level lower limit is the noisy channel of 0dB.Together
When, the actual implementation process of the method for the present invention is based on computer and carries out automatically, avoid because testing staff it is artificial it is unstable because
Detection recognition accuracy fluctuation caused by plain, the average detected recognition correct rate of the short-wave signal of this method is up to 90% or more.
With stable detection discrimination, timeliness is good.
Table 1 is carried out based on the method for the present invention once based on the test of actual channel, and what is obtained divides according to short-wave signal type
The short-wave signal set of class.Wherein, the corresponding number of signal number is the signal number that this method is classified, and is artificial in bracket
The correct signal number for meeting the short-wave signal type determined after check.By statistics, collects described in table 1 and amount to comprising signal
1106, wherein correct category signal 1031, accuracy 93.21%.
Short-wave signal set of the table 1 based on short-wave signal type
Table 2 compared the method for the present invention and traditional artificial shortwave detection recognition method, existing time domain short-wave signal detect
Recognition methods is in Average Accuracy, accuracy rate stability, the difference for handling the time.Since the method for the present invention is based on computer vision
Effective extraction has been carried out to short-wave signal time-frequency characteristics, has effectively removed the characteristic information of redundancy, thus in terms of accuracy rate compared with
Traditional artificial detection method has a distinct increment with existing tim e- domain detection recognition methods.Due to being automatically processed using computer,
Therefore accuracy rate stability is not influenced by time and personnel's fatigue, and more traditional artificial detection recognition methods is more stable.?
In terms of handling the time, due to being to carry out detection identification to the feature of the short-wave signal refined, the data volume of processing is compared with people
Work detection method and Time Domain Analysis are less, while the SVM classifier for introducing machine learning carries out classifier identification, processing
Time is shorter.
Table 2 compares the method for the present invention and other shortwave detection recognition methods
| Method type | Traditional artificial shortwave detects identification | The detection identification of time domain shortwave | Shortwave detection identification based on computer vision |
| Average Accuracy | 60% | 70% | 93.21% |
| Accuracy rate stability | ± 20% | ± 5% | ± 5% |
| Handle the time | > 1s | < 500ms | < 200ms |
Claims (2)
1. a kind of short-wave signal detection recognition method based on computer vision, it is characterised in that: this method is one kind based on
The short-wave signal detection and localization of calculation machine vision and time-frequency characteristics extract, and the SVM classifier based on machine learning, to signal when
Frequency feature is matched, the final automatic detection and identification for realizing short-wave signal；
This method comprises the following steps:
1) time-frequency data submatrix is obtained according to frequency direction every 300kHz piecemeal to signal time-frequency data matrix to be detected；
2) to each time-frequency data submatrix, following signal framing detection is carried out, concrete operations are as follows:
2.1) clock synchronization frequency projects on the frequency axis according to submatrix obtains one-dimensional row and vector V, V represent signal energy in frequency direction
Amount adds up；
2.2) distribution histogram of each element in V is analyzed；It is described 2.2) in use the threshold value thresholding based on histogram
Calculation method, calculation formula are T=(V_med-V_min)+V_min, and wherein V_med indicates that the mode of V element, V_min indicate V
The minimum value of middle element；
As thresholding and 2.1) 2.3) locating threshold T in the previous step drop shadow curve in step is combined, in reference drop shadow curve
Higher than the frequency separation of thresholding, corresponding frequency separation in time-frequency matrix is intercepted, obtains currently processed time-frequency data
The set of the time frequency signal submatrix of short-wave signal present in submatrix；
3) the corresponding time frequency signal submatrix set of each time-frequency data submatrix obtained first two steps merges, and obtains as defeated
The time frequency signal set of matrices of the whole short-wave signals for including in signal time-frequency data matrix to be detected entered；
4) set constituted to the time frequency signal matrix of each short-wave signal, carries out signal characteristic abstraction one by one；
4.1) kurtosis feature
The time frequency signal matrix of short-wave signal is projected on the frequency axis and obtains one-dimension array Y；Windowed traversal processing is carried out to Y；
The difference of the corresponding numerical value of point of proximity is calculated separately, and is substituted into following formula；
The value of point each in Y is replaced with into corresponding Dt, finally obtain kurtosis curve Y '；
It is then grouped sequence: using the zero in kurtosis curve as partitioning standards, kurtosis curve being divided into multiple spikes
Group, the corresponding kurtosis curve regions maximum value of statistics each group obtain the sequence of spike group maximum value sequence progress from big to small；
The spike group maximum value sequence sorted from large to small obtained in packet sequencing is divided into two classes, threshold value by subsequent given threshold
For the corresponding kurtosis curve of maximum sharpness group maximum value divided by constant CHill, CHill=10^3 takes group sequence smaller portions as office
Portion's spike group maximum value sequence, and to the part spike group maximum value sequence；
Ratio is successively asked to two adjacent values of spike group maximum value sequence, statistics obtains corresponding two maximum values of maximum ratio
Spike group maximum value sequence is divided into two parts, takes the lesser portion of sequence order by position using two maximum values as division limits
It is allocated as counting the maximum value number that the partial sequence includes for effective spike group maximum value sequence, as effective spike number, and
Result as kurtosis feature is exported；
4.2) spatial distribution characteristic
The time frequency signal matrix of short-wave signal is projected on the frequency axis and obtains one-dimension array Y；Evaluation domain section | V (i) | ∈ [a,
B], if section size is k, k=b-a, k is divided, the scatterplot quantity in each section is counted, calculates in each section and includes
The total scatterplot number of scatterplot number Zhan ratio；It is exported the scatterplot proportionality percentage array in each section as spatial distribution characteristic；
4.3) symmetric characteristics
The time frequency signal matrix of short-wave signal is projected on the frequency axis and obtains one-dimension array Y；Evaluation domain section [a, b], setting
Thresholding C, the part in curve higher than C are effective coverage；Wherein, C=0.2* (b-a)+a；Using curve highest point as starting point, to
Ask starting point at a distance from the end point of two sides respectively as end point in position when two sides look for curve to decline to C or less for the first time
D_right and d_left；The ratio between d_right and d_left are asked, when symmetrical, the result is close to 1；
4.4) Variance feature
The time frequency signal matrix of short-wave signal is projected on the frequency axis and obtains one-dimension array Y, extracts maximum value position in array Y
Corresponding frequency row Y (N) calculates the variance of Y (N) row, exports as Variance feature；
4.5) rectangle morphological feature
For each short-wave signal time-frequency data matrix, even rectangular degree Pr is calculated；Calculation method is Pr=As/Ar, and As is figure
The area of connected domain S as in, Ar are the minimum circumscribed rectangle areas for surrounding the connected domain；
It is corresponding to single short-wave signal time-frequency data matrix 4.1) to be integrated to 4.5) feature, obtain the short-wave signal time-frequency
The eigenmatrix of data matrix；
5) each short-wave signal Input matrix for obtaining step 4) is into preparatory trained SVM classifier,
Classifier output is obtained, is exported according to classifier as a result, short-wave signal type is marked.
2. a kind of short-wave signal detection recognition method based on computer vision according to claim 1, which is characterized in that
It is described 5) in, the SVM classifier trained has been used, based on the extracted kurtosis feature of machine vision, spatial distribution characteristic, right
Title property feature, Variance feature, rectangle morphological feature, identify short-wave signal type.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610917977.9A CN106503649B (en) | 2016-10-20 | 2016-10-20 | A kind of short-wave signal detection recognition method based on computer vision |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610917977.9A CN106503649B (en) | 2016-10-20 | 2016-10-20 | A kind of short-wave signal detection recognition method based on computer vision |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106503649A CN106503649A (en) | 2017-03-15 |
| CN106503649B true CN106503649B (en) | 2019-06-07 |
Family
ID=58319308
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610917977.9A Active CN106503649B (en) | 2016-10-20 | 2016-10-20 | A kind of short-wave signal detection recognition method based on computer vision |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106503649B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019165277A1 (en) * | 2018-02-23 | 2019-08-29 | Boston Scientific Scimed, Inc. | Methods for assessing a vessel with sequential physiological measurements |
| CN110631541B (en) * | 2019-10-31 | 2021-07-23 | 湖北民族大学 | Steel product shape measurement linearity detection and correction method |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101398846A (en) * | 2008-10-23 | 2009-04-01 | 上海交通大学 | Image, semantic and concept detection method based on partial color space characteristic |
| CN101631102A (en) * | 2009-04-10 | 2010-01-20 | 北京理工大学 | Interference pattern recognition technology of frequency hopping system |
| US8369567B1 (en) * | 2010-05-11 | 2013-02-05 | The United States Of America As Represented By The Secretary Of The Navy | Method for detecting and mapping fires using features extracted from overhead imagery |
| CN103968933A (en) * | 2014-04-09 | 2014-08-06 | 西安电子科技大学 | Fuzzy domain characteristics based optical fiber vibration signal identifying method |
| CN105354592A (en) * | 2015-10-22 | 2016-02-24 | 上海无线电设备研究所 | Classification based optimal time-frequency distribution design and target identification method |
| CN105574495A (en) * | 2015-12-14 | 2016-05-11 | 中国电子科技集团公司第二十八研究所 | Interval-type radar radiation source signal feature type recognition capability analyzing method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050232512A1 (en) * | 2004-04-20 | 2005-10-20 | Max-Viz, Inc. | Neural net based processor for synthetic vision fusion |
-
2016
- 2016-10-20 CN CN201610917977.9A patent/CN106503649B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101398846A (en) * | 2008-10-23 | 2009-04-01 | 上海交通大学 | Image, semantic and concept detection method based on partial color space characteristic |
| CN101631102A (en) * | 2009-04-10 | 2010-01-20 | 北京理工大学 | Interference pattern recognition technology of frequency hopping system |
| US8369567B1 (en) * | 2010-05-11 | 2013-02-05 | The United States Of America As Represented By The Secretary Of The Navy | Method for detecting and mapping fires using features extracted from overhead imagery |
| CN103968933A (en) * | 2014-04-09 | 2014-08-06 | 西安电子科技大学 | Fuzzy domain characteristics based optical fiber vibration signal identifying method |
| CN105354592A (en) * | 2015-10-22 | 2016-02-24 | 上海无线电设备研究所 | Classification based optimal time-frequency distribution design and target identification method |
| CN105574495A (en) * | 2015-12-14 | 2016-05-11 | 中国电子科技集团公司第二十八研究所 | Interval-type radar radiation source signal feature type recognition capability analyzing method |
Non-Patent Citations (3)
| Title |
|---|
| "A New Time-Frequency Spectrogram Analysis of FH Signals by Image Enhancement and Mathematical Morphology";Xu M等;《international Conference on Image and Graphics》;20071231;第610-615页 |
| "一种短波宽带综合检测分类算法";董延坤等;《信息工程大学学报》;20150630;第16卷(第3期);第332-338页 |
| "强噪声背景下莫尔斯信号的自动检测与识别";张汝波等;《哈尔滨工程大学学报》;20060228;第27卷(第1期);第112-117页 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106503649A (en) | 2017-03-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106653032B (en) | Based on the animal sounds detection method of multiband Energy distribution under low signal-to-noise ratio environment | |
| CN103577838A (en) | Face recognition method and device | |
| CN108537751B (en) | Thyroid ultrasound image automatic segmentation method based on radial basis function neural network | |
| CN106650605A (en) | Morse signal automatic detection decoding method based on machine learning | |
| CN107527016B (en) | User identity identification method based on motion sequence detection in indoor WiFi environment | |
| CN104680185B (en) | Hyperspectral image classification method based on boundary point reclassification | |
| Bhattacharjee et al. | Detection of Acute Lymphoblastic Leukemia using watershed transformation technique | |
| CN106503649B (en) | A kind of short-wave signal detection recognition method based on computer vision | |
| CN109410184A (en) | Live streaming pornographic image detection method based on dense confrontation network semi-supervised learning | |
| CN110414459A (en) | Establish the associated method and device of people's vehicle | |
| CN104680545A (en) | Method for detecting existence of salient objects in optical images | |
| CN110501741A (en) | A kind of useful signal detection method and system | |
| CN112528774A (en) | Intelligent sorting system and method for unknown radar signals in complex electromagnetic environment | |
| CN110033160A (en) | A kind of performance appraisal system and method | |
| CN107169996A (en) | Dynamic human face recognition methods in a kind of video | |
| CN111046025B (en) | Unmanned aerial vehicle signal detection method and device | |
| CN115294109A (en) | Real wood board production defect identification system based on artificial intelligence, and electronic equipment | |
| CN106951924B (en) | Seismic coherence body image fault automatic identification method and system based on AdaBoost algorithm | |
| CN108712222A (en) | A kind of cooperative frequency spectrum sensing method and relevant apparatus | |
| CN110414369B (en) | Cow face training method and device | |
| Du et al. | Feature extraction and pattern recognition algorithm of power cable partial discharge signal | |
| CN108270495B (en) | Background noise extraction method and system | |
| CN115019390A (en) | Video data processing method and device and electronic equipment | |
| CN110071884A (en) | A kind of Modulation Recognition of Communication Signal method based on improvement entropy cloud feature | |
| EP4167128A1 (en) | Signal analysis method and system based on model for acquiring and identifying noise panoramic distribution |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | 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 | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20221114 Address after: 100012 817, Floor 8, No. 101, Floor 3 to 8, Building 17, Rongchuang Road, Chaoyang District, Beijing Patentee after: Beijing Ge Lei Information Technology Co.,Ltd. Address before: 100124 No. 100 Chaoyang District Ping Tian Park, Beijing Patentee before: Beijing University of Technology |