CN114936577A - Mixed image blind separation method based on improved lion group algorithm - Google Patents
Mixed image blind separation method based on improved lion group algorithm Download PDFInfo
- Publication number
- CN114936577A CN114936577A CN202210560535.9A CN202210560535A CN114936577A CN 114936577 A CN114936577 A CN 114936577A CN 202210560535 A CN202210560535 A CN 202210560535A CN 114936577 A CN114936577 A CN 114936577A
- Authority
- CN
- China
- Prior art keywords
- lion
- image
- population
- separation
- 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.)
- Granted
Links
- 238000000926 separation method Methods 0.000 title claims abstract description 87
- 241000282320 Panthera leo Species 0.000 title claims abstract description 64
- 239000011159 matrix material Substances 0.000 claims abstract description 44
- 238000012545 processing Methods 0.000 claims abstract description 12
- 238000000034 method Methods 0.000 claims abstract description 11
- 230000002087 whitening effect Effects 0.000 claims abstract description 4
- 230000036039 immunity Effects 0.000 claims description 4
- 238000005070 sampling Methods 0.000 claims description 3
- 239000000126 substance Substances 0.000 claims description 2
- 238000010586 diagram Methods 0.000 description 5
- 230000015556 catabolic process Effects 0.000 description 2
- 238000006731 degradation reaction Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 241000283153 Cetacea Species 0.000 description 1
- LFQSCWFLJHTTHZ-UHFFFAOYSA-N Ethanol Chemical compound CCO LFQSCWFLJHTTHZ-UHFFFAOYSA-N 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Analysis (AREA)
- Computational Mathematics (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Mathematical Optimization (AREA)
- Artificial Intelligence (AREA)
- Pure & Applied Mathematics (AREA)
- Databases & Information Systems (AREA)
- Signal Processing (AREA)
- Algebra (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a mixed image blind separation method based on an improved lion group algorithm, which comprises the following steps: processing the image source signal by a nonsingular mixed matrix A to obtain an image observation signal; -centering and whitening said image observation signal x (t); obtaining an initial population x, wherein each individual x i Corresponds to a signal separation matrix, each individual x i The fitness value of (a) corresponds to the kurtosis value of (b); acquiring signal kurtosis corresponding to population individuals according to a fitness function of the image blind separation problem, and updating the highest kurtosis value of the current population and the highest kurtosis value of the individuals; in the later stage of iteration, an immune concentration selection mechanism is introduced to inhibit individuals with low affinity and high concentration and keep individuals with high affinity and low concentration; outputting an image separation signal according to Y (t) ═ WX (t), and realizing blind separation of mixed images(ii) a The method has high searching precision and high convergence speed, can effectively separate the noise-containing mixed images, and has wide application prospect in the fields of modern signal processing, image processing and the like.
Description
Technical Field
The invention relates to the technical field of image signal processing, in particular to a mixed image blind separation method based on an improved lion group algorithm.
Background
Blind Source Separation (BSS), also known as Blind signal processing, can recover a Source signal using an obtained observation signal in the absence of important information such as a Source and a channel, and is widely applied in the fields of image processing and the like. The image recovery and the image separation are to eliminate or minimize the image degradation caused by interference, speckle, dirt, noise and the like through the prior knowledge of the image degradation, and the image blind separation is a process of estimating or separating an original source image from blurred image features.
Researchers at home and abroad have many researches on image blind source separation methods, but the traditional natural gradient algorithm excessively depends on gradient information, and the fast fixed point algorithm is sensitive to initial solution. Therefore, how to improve the speed and the precision of solving the separation matrix and obtain a separation signal with higher quality has important practical significance.
Disclosure of Invention
The invention aims to provide a hybrid image blind separation method based on an improved lion group algorithm, which overcomes the limitations that the traditional blind source separation method excessively depends on gradient information and is sensitive to an initial solution, and effectively improves the separation performance.
In order to achieve the above object, the present application provides a hybrid image blind separation method based on an improved lion group algorithm, including:
step 1: image source signal S (t) ═ S 1 (t),S 2 (t),...,S n (t)]Processing the image by a nonsingular mixed matrix A to obtain an image observation signal X (t) ═ X 1 (t),X 2 (t),...,X m (t)]In which S is n (t) is the nth component, X, of the image source signal S (t) m (t) is the mth component of the image observation signal x (t), t is the sampling time, m and n are positive integers, a is the mxn order mixing matrix;
and 2, step: -centering and whitening said image observation signal x (t);
and step 3: setting the maximum iteration number T, the dimension space D and the scale factor of adult lion in lion group to obtain the initial population x, wherein each individual x i Corresponds to a signal separation matrix, each individual x i The fitness value of (a) corresponds to its kurtosis value;
and 4, step 4: acquiring the signal kurtosis corresponding to population individuals according to a fitness function of the blind image separation problem, and updating the highest kurtosis value of the current population and the highest kurtosis value of the individuals;
and 5: taking the individual with the highest kurtosis value in the population as the lion king, and updating the separation matrix corresponding to the lion king according to the formula (1):
wherein the content of the first and second substances,
a historical optimal separation matrix of the kth generation of the ith individual in the lion group; g is a radical of formula k Representing the optimal separation matrix of the kth generation population; gamma is [0,1 ]]A random value in between;
step 6: taking the adult lion with the second kurtosis value in the population as the parent lion, setting the individual with the worst kurtosis value as the child lion, and updating the separation matrix corresponding to the parent lion and the child lion according to the following formulas (2) and (3):
wherein q and rand are random numbers between (0, 1);
a separation matrix corresponding to the randomly selected female lion;
a historical optimal separation matrix of the kth generation of the ith individual in the lion group; g k Representing the optimal separation matrix of the kth generation population;
is the kth generation historical optimal separation matrix of the young lion following the mother lion; inertial weight ω and historical optimal separation matrix corresponding to ith young lion driven within hunting range
The updated formula is as follows:
wherein, ω is max And ω min 0.95 and 0.35, respectively; e and T represent the current and maximum number of iterations, respectively; l is max Represents the maximum of all individual vector distances of the search interval; l is a radical of an alcohol ik Vector distance between the global historical optimal position and the historical optimal position of each lion;
and
maximum and minimum mean values for each dimension, respectively;
and
respectively representing the component of the ith lion history optimal separation matrix in the d dimension and the component of the population history optimal separation matrix in the d dimension.
And 7: in the later stage of iteration, an immunity concentration selection mechanism is introduced to inhibit individuals with low affinity and high concentration, retain individuals with high affinity and low concentration, and improve the diversity of corresponding kurtosis values of population individuals. The definition of affinity and concentration is as in formulae (7), (8):
wherein fit is the kurtosis value of population individuals; d is a dimension space; n is the number of the population; l (x) i ) Vector distance between individuals in the population;
the formula chosen for the concentrations of immunity is:
wherein alpha is a balance coefficient and takes a value between (0, 1).
And 8: judging whether the maximum iteration frequency is reached, if so, outputting a separation matrix W corresponding to the individual with the highest population kurtosis, and outputting an image separation signal according to Y (t) ═ WX (t) to realize blind separation of the mixed images; otherwise, returning to the step 4.
Compared with the prior art, the technical scheme adopted by the invention has the advantages that: according to the method, the sum of absolute values of the kurtosis of the signals is used as a fitness function, an optimal solution, namely an optimal separation matrix, is obtained through the fitness function, and blind separation of the mixed signals is achieved; the method has high search precision and high convergence speed, can effectively separate the noise-containing mixed image, and has wide application prospect in the fields of modern signal processing, image processing and the like.
Drawings
FIG. 1 is a diagram of a linear hybrid blind source separation method model according to the present invention;
FIG. 2 is a flow chart of the hybrid image blind separation method based on the improved lion group algorithm;
fig. 3 is a schematic diagram of image source signals according to the present invention;
FIG. 4 is a schematic diagram of an image observation signal according to the present invention;
FIG. 5 is a diagram of an image separation signal according to the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of and not restrictive on the broad application, i.e., the embodiments described are only a few examples and not all examples.
Example 1
As can be seen from the model diagram of the linear hybrid blind source separation method given in fig. 1, the key of blind source separation is to determine a fitness function, and then perform iterative optimization on the fitness function by using an improved lion group algorithm to obtain an optimal separation matrix W, so that each component of an output vector has the strongest independence. Wherein, the relation between the separation signal Y (t) and the observation signal X (t) is as follows:
Y(t)＝WX(t)＝WAS(t)+WN(t) (10)
wherein A is an mxn-order mixing matrix (m is larger than or equal to n); x (t) is a signal consisting of m signals X i (t) an observation signal of the formula, X (t) ([ X ]) 1 (t),X 2 (t),...,X m (t)](ii) a S (t) is a signal consisting of n signals S i (t) the source signal, S (t) S 1 (t),S 2 (t),...,S n (t)](ii) a N (t) is formed by m signals N i (t) a noise signal of [ N ], [ N: [, ] 1 (T)，N 2 (T),...,N m (t)](ii) a Y (t) is a signal consisting of n signals Y i (t) separation signal Y (t) ═ Y 1 (t),Y 2 (t),...,Y n (t)]。
Without the source signal and channel information, the signal amplitude and sequence after blind source separation is difficult to determine, called ambiguity. Although blind source separation is ambiguous, in most scientific research and production practices, the ambiguity of blind source separation does not have much impact on the results.
In order to prove the effectiveness of the method provided by the invention, 3 256 × 256 gray images and random noise images are selected as image source signals, image observation signals are obtained after mixing, the sum of absolute values of signal kurtosis is used as a fitness function, the method provided by the invention is used for solving the fitness function to obtain an optimal solution, namely an optimal separation matrix, so that the image separation signals with the strongest independence are obtained, and finally the blind separation of the mixed images is realized.
As shown in fig. 2, the embodiment provides a hybrid image blind separation method based on an improved lion group algorithm, which specifically includes:
step 1: a plurality of image source signals S (t) as shown in fig. 3 [ S ] 1 (t),S 2 (t),...,S n (t)]Processing the image by the nonsingular mixing matrix a to obtain an image observation signal X (t) ═ X shown in fig. 4 1 (t),X 2 (t),...,X m (t)]In which S is n (t) is the nth component, X, of the image source signal S (t) m (t) is the mth component of the image observation signal x (t), t is the sampling time, m and n are positive integers, a is the mxn order mixing matrix;
and 2, step: and (3) carrying out centering and whitening processing on the image observation signal X (t) obtained in the step (1).
And 3, step 3: setting the maximum iteration number T, the dimension space D and the scale factor of adult lion in lion group to obtain the initial population x, wherein each individual x i Corresponds to a signal separation matrix, each individual x i The fitness value of (a) corresponds to its kurtosis value;
and 4, step 4: acquiring the signal kurtosis corresponding to population individuals according to a fitness function of the blind image separation problem, and updating the highest kurtosis value of the current population and the highest kurtosis value of the individuals; the fitness function is shown in formula (11):
wherein, kurtosis
The higher the kurtosis of the signal, the greater the independence.
And 5: taking the individual with the highest kurtosis value in the population as a lion king, and updating a separation matrix corresponding to the lion king;
step 6: taking the adult lion with the second kurtosis value in the population as a parent lion, setting the individual with the worst kurtosis value as a child lion, and updating the separation matrix corresponding to the parent lion and the child lion;
and 7: in the later stage of algorithm iteration, an immunity concentration selection mechanism is introduced, and the diversity of corresponding kurtosis values of population individuals is improved.
And 8: judging whether the maximum iteration times is reached, if so, outputting a separation matrix W corresponding to the individual with the highest population kurtosis, and outputting an image separation signal according to Y (t) -WX (t) to realize blind separation of the mixed images; otherwise, returning to the step 4
The parameters in the method provided by the invention can be as follows: the maximum iteration time T is 50, the population number is 30, the dimension space D is 10, the ratio of adult lions in the lions is 0.2, and the maximum inertia weight w max Is 0.95, the minimum inertial weight w min 0.35, and the balance coefficient alpha is a random number between 0 and 1.
As can be seen from fig. 5, the image separated by the method of the present invention can better restore the original image.
To compare the separation performance of the algorithms, the similarity coefficient, crosstalk error, and computation time were taken as evaluation indexes. The similarity coefficient is an index for measuring the similarity degree of the source signal and the separation signal, the separation effect is better if the similarity coefficient is larger, the similarity coefficient is a 4 multiplied by 4 matrix in the invention, and the maximum value of each channel is taken as experimental data; the closer the crosstalk error is to 0, the better the separation performance. The comparison algorithm includes BOAIALSO, whale algorithm (WOA), lion group algorithm (LSO) and genetic lion Group Algorithm (GALSO). Each algorithm was run independently 30 times to improve the accuracy of the experiment.
The comparative results are shown in Table 1. Where the best similarity factor, crosstalk error and computation time for all algorithms will be bolded.
Table 1. several algorithms solve the performance evaluation index data of the blind separation problem of the mixed image.
As can be seen from table 1, the overall separation performance of BOAIALSO is better than that of other algorithms for comparison, the obtained separation signal has the highest similarity coefficient, and the crosstalk error and the calculation time are the smallest. Therefore, the method provided by the invention has good separation performance.
The foregoing description of specific exemplary embodiments of the invention has been presented for the purposes of illustration and description. It is not intended to limit the invention to the precise form disclosed, and obviously many modifications and variations are possible in light of the above teaching. The exemplary embodiments were chosen and described in order to explain certain principles of the invention and its practical application to enable one skilled in the art to make and use various exemplary embodiments of the invention and various alternatives and modifications. It is intended that the scope of the invention be defined by the claims and their equivalents.
Claims (5)
1. A hybrid image blind separation method based on an improved lion group algorithm is characterized by comprising the following steps:
step 1: image source signal S (t) ═ S 1 (t),S 2 (t),...,S n (t)]Processing the image by a nonsingular mixed matrix A to obtain an image observation signal X (t) ═ X 1 (t),X 2 (t),...,X m (t)]In which S is n (t) is the image source signal SThe nth component of (t), X m (t) is the mth component of the image observation signal x (t), t is the sampling time, m and n are positive integers, a is the mxn order mixing matrix;
step 2: -centering and whitening said image observation signal x (t);
and step 3: setting the maximum iteration number T, the dimension space D and the scale factor of adult lion in lion group to obtain the initial population x, wherein each individual x i Corresponds to a signal separation matrix, each individual x i The fitness value of (a) corresponds to its kurtosis value;
and 4, step 4: acquiring signal kurtosis corresponding to population individuals according to a fitness function of the image blind separation problem, and updating the highest kurtosis value of the current population and the highest kurtosis value of the individuals;
and 5: taking the individual with the highest kurtosis value in the population as a lion king, and updating a separation matrix corresponding to the lion king;
step 6: taking the adult lion with the second kurtosis value in the population as a parent lion, setting the individual with the worst kurtosis value as a child lion, and updating the separation matrix corresponding to the parent lion and the child lion;
and 7: in the later stage of iteration, an immune concentration selection mechanism is introduced to inhibit individuals with low affinity and high concentration and keep the individuals with high affinity and low concentration;
and 8: judging whether the maximum iteration frequency is reached, if so, outputting a separation matrix W corresponding to the individual with the highest population kurtosis, and outputting an image separation signal according to Y (t) ═ WX (t) to realize blind separation of the mixed images; otherwise, returning to the step 4.
2. The method for blind separation of mixed images based on an improved lion group algorithm as claimed in claim 1, wherein the kurtosis value of lion king is updated in a manner shown in formula (1):
wherein the content of the first and second substances,
a historical optimal separation matrix of the kth generation of the ith individual in the lion group; g k Representing the optimal separation matrix of the kth generation population; gamma is [0,1 ]]A random value in between.
3. The blind separation method of mixed images based on the improved lion group algorithm as claimed in claim 1, wherein the kurtosis values of the female lion and the young lion are updated according to the following formulas (2) and (3):
wherein q and rand are random numbers between (0, 1);
a separation matrix corresponding to the randomly selected female lion;
a historical optimal separation matrix of the kth generation of the ith individual in the lion group; g k Representing the optimal separation matrix of the kth generation population;
is the kth generation historical optimal separation matrix of the young lion following the mother lion; omega is the weight of the inertia,
the historical optimal separation matrix corresponding to the ith young lion being driven within the hunting zone.
4. The hybrid image blind separation method based on the improved lion group algorithm as claimed in claim 3, wherein the method comprisesIs characterized in that the inertia weight omega and the kurtosis value corresponding to the ith young lion driven in the hunting range
The formula is updated as follows:
wherein, ω is max And omega min 0.95 and 0.35, respectively; e and T represent the current and maximum number of iterations, respectively; l is max Represents the maximum of all individual vector distances of the search interval; l is ik Vector distance between the global historical optimal position and the historical optimal position of each lion;
and
maximum and minimum mean values for each dimension, respectively;
and
respectively representing the component of the ith lion history optimal separation matrix in the d dimension and the component of the population history optimal separation matrix in the d dimension.
5. The blind separation method for mixed images based on the improved lion-group algorithm as claimed in claim 1, wherein the definition of said affinity and concentration is as shown in formulas (7), (8):
degree of affinity
Concentration of
Wherein fit is the kurtosis value of individual population; d is a dimension space; n is the number of the population; l (x) i ) Vector distance between individuals in the population;
the formula chosen for the concentrations of immunity is:
wherein alpha is a balance coefficient and takes a value between (0, 1).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210560535.9A CN114936577B (en) | 2022-05-23 | 2022-05-23 | Mixed image blind separation method based on improved lion group algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210560535.9A CN114936577B (en) | 2022-05-23 | 2022-05-23 | Mixed image blind separation method based on improved lion group algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114936577A true CN114936577A (en) | 2022-08-23 |
| CN114936577B CN114936577B (en) | 2024-03-26 |
Family
ID=82864851
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210560535.9A Active CN114936577B (en) | 2022-05-23 | 2022-05-23 | Mixed image blind separation method based on improved lion group algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114936577B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106778809A (en) * | 2016-11-23 | 2017-05-31 | 安徽理工大学 | A kind of blind source separation method based on improvement chicken group's algorithm |
| WO2018072351A1 (en) * | 2016-10-20 | 2018-04-26 | 北京工业大学 | Method for optimizing support vector machine on basis of particle swarm optimization algorithm |
| CN109712160A (en) * | 2018-12-26 | 2019-05-03 | 桂林电子科技大学 | Improved lion group algorithm is combined to realize carrying out image threshold segmentation method based on generalized entropy |
| CN109872330A (en) * | 2019-01-25 | 2019-06-11 | 安徽理工大学 | A kind of two-dimentional Otsu Fast image segmentation method for improving lion group's optimization |
-
2022
- 2022-05-23 CN CN202210560535.9A patent/CN114936577B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018072351A1 (en) * | 2016-10-20 | 2018-04-26 | 北京工业大学 | Method for optimizing support vector machine on basis of particle swarm optimization algorithm |
| CN106778809A (en) * | 2016-11-23 | 2017-05-31 | 安徽理工大学 | A kind of blind source separation method based on improvement chicken group's algorithm |
| CN109712160A (en) * | 2018-12-26 | 2019-05-03 | 桂林电子科技大学 | Improved lion group algorithm is combined to realize carrying out image threshold segmentation method based on generalized entropy |
| CN109872330A (en) * | 2019-01-25 | 2019-06-11 | 安徽理工大学 | A kind of two-dimentional Otsu Fast image segmentation method for improving lion group's optimization |
Non-Patent Citations (2)
| Title |
|---|
| 刘生建;杨艳;周永权;: "一种群体智能算法――狮群算法", 模式识别与人工智能, no. 05, 15 May 2018 (2018-05-15), pages 45 - 55 * |
| 岳克强;赵知劲;沈雷;: "基于负熵和智能优化算法的盲源分离方法", 计算机工程, no. 04, 20 February 2010 (2010-02-20), pages 256 - 258 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114936577B (en) | 2024-03-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111814719B (en) | Skeleton behavior recognition method based on 3D space-time diagram convolution | |
| CN110188228B (en) | Cross-modal retrieval method based on sketch retrieval three-dimensional model | |
| Jennings et al. | Partial elimination | |
| Kalluri et al. | Fast algorithms for weighted myriad computation by fixed-point search | |
| CN111832228B (en) | Vibration transmission system based on CNN-LSTM | |
| CN111008991B (en) | Background-aware related filtering target tracking method | |
| CN108399268B (en) | Incremental heterogeneous graph clustering method based on game theory | |
| CN112967210B (en) | Unmanned aerial vehicle image denoising method based on full convolution twin network | |
| CN111178514A (en) | Neural network quantification method and system | |
| CN110857998B (en) | Elastic reverse time migration method and system based on lowrank finite difference | |
| CN111242003B (en) | Video salient object detection method based on multi-scale constrained self-attention mechanism | |
| CN114037743A (en) | Three-dimensional point cloud robust registration method for Qinhong warriors based on dynamic graph attention mechanism | |
| CN114936577A (en) | Mixed image blind separation method based on improved lion group algorithm | |
| CN114624646A (en) | DOA estimation method based on model-driven complex neural network | |
| CN108520205B (en) | motion-KNN-based human body motion recognition method | |
| CN110415281B (en) | Loam curvature weighting-based point set rigid registration method | |
| CN110138479B (en) | Spectrum sensing method based on dictionary learning under extremely low signal-to-noise ratio environment | |
| CN111951181A (en) | Hyperspectral image denoising method based on non-local similarity and weighted truncation kernel norm | |
| CN115014313B (en) | Polarized light compass heading error processing method based on parallel multi-scale | |
| CN113869289B (en) | Multi-channel ship radiation noise feature extraction method based on entropy | |
| CN109520496A (en) | A kind of inertial navigation sensors data de-noising method based on blind source separation method | |
| CN109727219A (en) | A kind of image de-noising method and system based on image sparse expression | |
| CN115631771A (en) | Sound event detection and positioning method based on combined convolutional neural network | |
| CN114023350A (en) | Sound source separation method based on shallow feature reactivation and multi-stage mixed attention | |
| CN112666520A (en) | Method and system for positioning time-frequency spectrum sound source with adjustable response |
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 |