CN109002848B - Weak and small target detection method based on feature mapping neural network - Google Patents

Weak and small target detection method based on feature mapping neural network Download PDF

Info

Publication number
CN109002848B
CN109002848B CN201810729648.0A CN201810729648A CN109002848B CN 109002848 B CN109002848 B CN 109002848B CN 201810729648 A CN201810729648 A CN 201810729648A CN 109002848 B CN109002848 B CN 109002848B
Authority
CN
China
Prior art keywords
weak
neural network
target
representing
small
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
Application number
CN201810729648.0A
Other languages
Chinese (zh)
Other versions
CN109002848A (en
Inventor
谢春芝
高志升
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Jiangxi Chengan Technology Co.,Ltd.
Shenzhen Wanzhida Technology Co ltd
Original Assignee
Xihua University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Xihua University filed Critical Xihua University
Priority to CN201810729648.0A priority Critical patent/CN109002848B/en
Publication of CN109002848A publication Critical patent/CN109002848A/en
Application granted granted Critical
Publication of CN109002848B publication Critical patent/CN109002848B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • 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
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
    • 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
    • G06V2201/00—Indexing scheme relating to image or video recognition or understanding
    • G06V2201/07—Target detection

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Physics & Mathematics (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • General Engineering & Computer Science (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Multimedia (AREA)
  • Image Analysis (AREA)

Abstract

The invention discloses a weak and small target detection method based on a feature mapping neural network, relating to the field of weak and small target detection; which comprises the following steps of 1: constructing and training a spindle-type deep neural network; step 2: inputting the collected images of the weak and small targets into a trained spindle-type deep neural network to obtain an amplitude map of target enhancement and background suppression; and step 3: and the amplitude diagram completes the detection of the weak and small targets by adopting a constant false alarm rate method. The invention adopts the spindle-shaped network structure, improves the strong representation capability of the network, solves the problem of low detection precision caused by the influence of noise and interference on the existing weak and small targets, and achieves the effects of improving the strong representation capability of the network and improving the detection precision of the weak and small targets in a high-noise environment.

Description

Weak and small target detection method based on feature mapping neural network
Technical Field
The invention relates to the field of weak and small target detection, in particular to a weak and small target detection method based on a feature mapping neural network.
Background
The passive millimeter wave PMMW and infrared imaging have the excellent characteristics of no radiation and strong penetration capability, and the application of the passive millimeter wave PMMW and infrared imaging in the military field is increasingly concerned, so that the passive millimeter wave PMMW and infrared imaging has very important significance in the research of the detection of weak and small targets under the millimeter wave and infrared imaging. The technology for detecting the weak and small targets is rapidly developed in recent years, the diameter of the weak and small target is 3-5 pixels, but the high-precision detection of the weak and small target under the conditions of millimeter wave and infrared imaging still faces great difficulty: firstly, the imaging distance of a target is generally longer, the detected target area is smaller, the signal-to-noise ratio is lower, and no texture feature can be extracted; second, target imaging is usually disturbed by complex background, a lot of clutter, noise, and some marginal information such as: the presence of cloud edges, sea-sky baselines, building edges, etc., causes the target to be submerged in the background.
Aiming at the detection of weak and small targets, a series of detection methods are proposed in the academic world in recent years; the background suppression method is the most common method in weak and small target detection, and the method carries out target detection on the basis of estimating the background of an image to be detected. It is mainly divided into three detection methods: the first type is a filtering-based method, which estimates the background by image filtering, and finally enhances the target; the effect of inhibiting the background is better under the condition of simpler background, the false alarm probability is increased under the condition of more complex background and lower signal-to-noise ratio, and the detection precision is reduced; the second category is regression-based methods, which can be further classified into linear regression and nonlinear regression, the classical linear regression method relies on a specific background clutter model and parameter estimation seeking a hypothesis model; whereas non-linear regression methods rely solely on the data itself to estimate the regression function; the nuclear regression algorithm NRRKR is a typical nonlinear regression algorithm; in practical application, due to lack of prior knowledge of background clutter, the nonlinear regression method is more suitable for detecting weak and small targets under complex background conditions, but has obvious disadvantages: each local area needs to be subjected to multiple regression iterations, and the overall algorithm efficiency is extremely low; the third method is to suppress the background according to the local contrast difference, enhance the target and complete the detection of the target. The method has a good detection effect under the condition of a simpler background, and is easy to increase the number of false alarm targets and be influenced by noise under the condition of a complex background.
Besides a background suppression method, a detection method based on machine learning is provided, which solves the target detection problem by using the idea of pattern classification, respectively trains and models a target and a background, and then judges whether sub-image blocks of a test image contain the target according to a judgment rule, such as NLPCA, SPCA, FLD and the like. Later, with the advent of sparse representation theory, new methods were brought to solve the problem of weak and small target detection. According to the infrared weak and small target detection algorithm SR based on image sparse representation, a target dictionary is generated by adopting a binary Gaussian model, and then the position of a target is judged through the difference of sparse coefficients of a background sub-block and a target sub-block in the target dictionary. The Gaussian dictionary is only suitable for weak and small targets with Gaussian distribution as a typical structured overcomplete dictionary, and for non-structural targets, sparse representation coefficients of the Gaussian dictionary are not enough to distinguish the target from background clutter. Later, Wang et al propose a multi-scale self-adaptive form sparse dictionary to detect weak and small targets, and describe different components of an image by adopting atoms with different sizes, so as to capture more subtle local features of the image and improve the detection precision; then, chen et al propose a sparsity-based method that improves the difference of sparse representation by manually constructing discriminative dual dictionaries in an offline learning manner; later, new methods were proposed: the method comprises the steps of firstly constructing a self-adaptive type over-complete space-time dictionary through the content of a learning sequence image, then extracting a target space-time dictionary and a background space-time dictionary from the over-complete dictionary by using a multivariate Gaussian model, carrying out sparse reconstruction on a plurality of frames of images in the target space-time dictionary and the background space-time dictionary respectively, and distinguishing a target and a background by using reconstruction differences. Aiming at the defects of the method in the aspect of dictionary learning, an improved ISR method is provided, and the method provides a remarkable background and target double-dictionary construction method which has better target and background modeling capability. The methods improve the detection precision to a certain extent, but the methods have the following defects: on one hand, the method is easily interfered by noise, on the other hand, the target and the background sparse feature have insignificant difference and are easily mixed together, and the detection difficulty is increased.
Therefore, a method for detecting a small target is needed, which can overcome the noise factor of the small target and improve the detection accuracy.
Disclosure of Invention
The invention aims to: the invention provides a method for detecting a small and weak target based on a feature mapping neural network, which solves the problem of low detection precision caused by the influence of noise and interference on the existing small and weak target.
The technical scheme adopted by the invention is as follows:
a weak and small target detection method based on a feature mapping neural network comprises the following steps:
step 1: constructing and training a spindle-type deep neural network;
step 2: inputting the collected images of the weak and small targets into a trained spindle-type deep neural network to obtain an amplitude map of target enhancement and background suppression;
and step 3: and the amplitude diagram completes the detection of the weak and small targets by adopting a constant false alarm rate method.
Preferably, the step 1 comprises the steps of:
step 1.1: constructing a structure of a spindle-type deep neural network comprising an input layer, a decoding layer, an encoding layer and a softmax output layer;
step 1.2: determining the hyper-parameters of the spindle-type deep neural network by adopting a cross validation method to obtain the spindle-type deep neural network;
step 1.3: constructing a training data set;
step 1.4: inputting the training data set into the spindle-type deep neural network, and training in an unsupervised mode to obtain the weight of the initialized network to finish training.
Preferably, the decoding layer and the encoding layer are trained and calculated as follows:
hk＝σ(WkX+bk)
wherein, WkRepresenting a weight matrix, bkRepresenting a bias vector, sigma representing an activation function, X ═ X1,x2,...,xmDenotes the input of the current layer, hkRepresenting the output of the current layer.
Preferably, the step 2 comprises the steps of:
step 2.1: inputting the collected small target image into a trained spindle-type deep neural network, setting a small target sample label as 1 and setting a background sample label as 0;
step 2.2: judging the weak and small target image by adopting a sliding window method to obtain the probability value of the weak and small target in the image;
step 2.3: and obtaining target enhancement and background suppression amplitude maps corresponding to a plurality of weak and small targets by using the output layer logistic regression result, namely a plurality of probability values as window coordinate point response amplitude values.
Preferably, the step 3 comprises the steps of:
step 3.1: extracting sub-blocks of each value in the amplitude map by adopting a sliding window
Figure GDA0003269489500000031
Detecting, inputting the spindle type deep neural network to obtain an amplitude value;
step 3.2: counting each detected amplitude value by adopting a constant false alarm rate to obtain a false alarm probability;
Figure GDA0003269489500000032
wherein T represents likelihood ratio detectionThe threshold value is set to a value that is,
Figure GDA0003269489500000033
representing the mean of the sliding window sub-blocks, P representing the number of points within a window sub-block, PfaIndicating the false alarm probability, tau, of a constant false alarm rate detection settingCFARDenotes a detection threshold value, F1,p-1(τCFAR) A cumulative distribution function representing a central F random variable;
step 3.3: the total number of candidate targets is calculated according to the false alarm probability, the false alarm probability is ranked from high to low, and the bit target is determined from a plurality of amplitude values.
Preferably, the constructing of the training data set in step 1.3 comprises the following steps:
step 1.3.1: randomly generating coordinate points in the image without the weak and small targets as simulation targets, and extracting an N x N window area as a background sample;
step 1.3.2: in the background sample, a two-dimensional Gaussian intensity model and a simulation target are used as target samples, wherein the two-dimensional Gaussian model is as follows:
Figure GDA0003269489500000034
wherein (x)0,y0) Representing the center position of the target image, s (i, j) representing the pixel value of the target image at position (i, j), sERepresenting the intensity of the generated object, and having a value of (0, 1)]Random number between, σxAnd σyRespectively representing horizontal and vertical dispersion parameters, having values between 0, 2]To (c) to (d);
step 1.3.3: and adjusting different parameters of the target sample to generate weak targets with different signal-to-noise ratios to complete the construction of the training data set.
In summary, due to the adoption of the technical scheme, the invention has the beneficial effects that:
1. according to the method, a spindle-shaped network structure is adopted, firstly, low-dimensional small target block features are mapped to a high-dimensional space, then, high-resolution features are extracted through a coding neural network, background and target discrimination is completed, an image with enhanced background suppression targets is obtained according to the strength output by the network discrimination, and finally, detection of the small targets is completed by a detection method based on a constant false alarm rate, so that the problem that the existing small targets are low in detection precision due to the influence of noise and interference is solved, the strong representation capability of the network is improved, and the detection precision of the small targets is improved under a high-noise environment;
2. the method detects the millimeter waves and the infrared images under different scenes, realizes decoding operation at the front end of the network, enables the network to have stronger expression capability, firstly maps the pixel characteristics of a weak and small target area to a high-dimensional characteristic space, and then maps the high-dimensional characteristics to a low-dimensional characteristic space which is easy to distinguish through coding, thereby achieving the effects of lower false alarm rate, higher detection precision and stronger robustness;
3. the network is pre-trained in an unsupervised learning mode, a deeper structure is constructed, the stability of the network can be improved through the initialized network weight obtained by unsupervised learning, so that the local minimum problem in the process of directly training the deep neural network is avoided, meanwhile, the internal characteristics of a series of related data sets are obtained by unsupervised learning, the redundant components of input data are removed, the high discriminable characteristics can be obtained, and the discrimination precision of the network is further improved;
4. the constant false alarm rate method adopts the slider to detect, and adopts the slider to detect the constant false alarm rate when the distance between two targets is assumed to be a certain distance according to the actual situation, so that the detection speed can be increased on one hand, and the detection precision can be improved on the other hand.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained according to the drawings without inventive efforts.
FIG. 1 is a schematic diagram of a spindle-type deep neural network structure according to the present invention;
FIG. 2 is a schematic diagram of an embodiment of a spindle-type deep neural network according to the present invention;
FIG. 3 is a two-dimensional graph of training data according to the present invention;
FIG. 4 is a three-dimensional map of training data of the present invention;
FIG. 5 is a two-dimensional map of training data of the present invention;
FIG. 6 is a graph of training data for the present invention;
FIG. 7 is a data diagram of a simulation test data set according to the present invention;
FIG. 8 is a diagram illustrating a detection result after background suppression of a simulation image according to the present invention;
FIG. 9 shows SNR of the present invention<Probability of 10-hour detection PdAnd false alarm rate PfaA graph of relationships between;
FIG. 10 is a drawing 10 of the present invention<SNR<Probability of detection P at 20 hoursdAnd false alarm rate PfaA graph of relationships between;
FIG. 11 shows a schematic view of 20 of the present invention<SNR<Probability of detection P at 30 hoursdAnd false alarm rate PfaA graph of relationships between;
FIG. 12 shows a schematic view of a device 30 according to the present invention<SNR<Probability of detection P of 40 hoursdAnd false alarm rate PfaA graph of relationships between;
FIG. 13 shows P of the present inventionfaUnder 1e-4, PdDetecting the effect along with the change of SNR;
FIG. 14 shows P of the present inventionfa＝10e-4，PdDetecting the effect along with the change of SNR;
FIG. 15 shows P of the present inventionfa＝20e-4，PdDetecting the effect along with the change of SNR;
FIG. 16 shows P of the present inventionfa＝30e-4，PdDetecting the effect along with the change of SNR;
fig. 17 is a schematic enlarged view of the effect of the DL algorithm of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the detailed description and specific examples, while indicating the preferred embodiment of the invention, are intended for purposes of illustration only and are not intended to limit the scope of the invention. The components of embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations.
Thus, the following detailed description of the embodiments of the present invention, presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the present invention without making any creative effort, shall fall within the protection scope of the present invention.
It is noted that relational terms such as "first" and "second," and the like, may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
The technical problem is as follows: solves the problem that the prior weak and small target is influenced by noise and interference to cause low detection precision
The technical means is as follows:
a weak and small target detection method based on a feature mapping neural network comprises the following steps:
step 1: constructing and training a spindle-type deep neural network;
step 2: inputting the collected images of the weak and small targets into a trained spindle-type deep neural network to obtain an amplitude map of target enhancement and background suppression;
and step 3: and the amplitude diagram completes the detection of the weak and small targets by adopting a constant false alarm rate method.
The step 1 comprises the following steps:
step 1.1: constructing a structure of a spindle-type deep neural network comprising an input layer, a decoding layer, an encoding layer and a softmax output layer;
step 1.2: determining the hyper-parameters of the spindle-type deep neural network by adopting a cross validation method to obtain the spindle-type deep neural network;
step 1.3: constructing a training data set;
step 1.4: inputting the training data set into the spindle-type deep neural network, and training in an unsupervised mode to obtain the weight of the initialized network to finish training.
The decoding layer and the coding layer are trained and calculated as follows:
hk＝σ(WkX+bk)
wherein, WkRepresenting a weight matrix, bkRepresenting a bias vector, sigma representing an activation function, X ═ X1,x2,...,xmDenotes the input of the current layer, hkRepresenting the output of the current layer.
The step 2 comprises the following steps:
step 2.1: inputting the collected small target image into a trained spindle-type deep neural network, setting a small target sample label as 1 and setting a background sample label as 0;
step 2.2: judging the weak and small target image by adopting a sliding window method to obtain the probability value of the weak and small target in the image;
step 2.3: and obtaining target enhancement and background suppression amplitude maps corresponding to a plurality of weak and small targets by using the output layer logistic regression result, namely a plurality of probability values as window coordinate point response amplitude values.
The step 3 comprises the following steps:
step 3.1: extracting sub-blocks of each value in the amplitude map by adopting a sliding window
Figure GDA0003269489500000061
Detecting, inputting the spindle type deep neural network to obtain an amplitude value;
step 3.2: counting each detected amplitude value by adopting a constant false alarm rate to obtain a false alarm probability;
Figure GDA0003269489500000062
wherein T represents a threshold value of likelihood ratio detection,
Figure GDA0003269489500000063
representing the mean of the sliding window sub-blocks, P representing the number of points within a window sub-block, PfaIndicating the false alarm probability, tau, of a constant false alarm rate detection settingCFARDenotes a detection threshold value, F1,p-1(τCFAR) A cumulative distribution function representing a central F random variable;
step 3.3: the total number of candidate targets is calculated according to the false alarm probability, the false alarm probability is ranked from high to low, and the bit target is determined from a plurality of amplitude values.
The construction of the training data set in step 1.3 comprises the following steps:
step 1.3.1: randomly generating coordinate points in the image without the weak and small targets as simulation targets, and extracting an N x N window area as a background sample;
step 1.3.2: in the background sample, a two-dimensional Gaussian intensity model and a simulation target are used as target samples, wherein the two-dimensional Gaussian model is as follows:
Figure GDA0003269489500000071
wherein (x)0,y0) Representing the center position of the target image, s (i, j) representing the pixel value of the target image at position (i, j), sERepresenting the intensity of the generated object, and having a value of (0, 1)]Random number between, σxAnd σyRespectively representing horizontal and vertical dispersion parameters, having values between 0, 2]To (c) to (d);
step 1.3.3: and adjusting different parameters of the target sample to generate weak targets with different signal-to-noise ratios to complete the construction of the training data set.
The technical effects are as follows:
the method adopts a spindle-shaped network structure, firstly maps the low-dimensional weak and small target block characteristics to a high-dimensional space, then extracts high-resolution characteristics through a coding neural network to complete background and target discrimination, obtains an image with enhanced background suppression targets according to the strength output by the network discrimination, and finally completes the detection of the weak and small targets by adopting a detection method based on the constant false alarm rate, thereby solving the problem of low detection precision caused by the influence of noise and interference on the existing weak and small targets, achieving the effects of improving the strong representation capability of the network and improving the detection precision of the weak and small targets in a high-noise environment.
The features and properties of the present invention are described in further detail below with reference to examples.
Example 1
The training data of the network is shown in fig. 3, a small circle represents one type of project, an x point represents another type of sample, and the training samples are linear inseparable and have high discrimination difficulty; after the deep network is trained through a training sample, the features of the second layer and the third layer are output as shown in fig. 4, after the two-dimensional features which are not linearly separable are mapped to three dimensions through a neural network, the three-dimensional features have linear discriminability in a three-dimensional feature space, and after the three-dimensional features are recoded to two dimensions, as shown in fig. 5, the original low-dimensional features which are not linearly separable are mapped to a linear discriminability feature space, and blue x sample samples are distributed in a compact area. As shown in fig. 2, the whole network contains 6 layers, which can be divided into 2 parts, the input layer to the 2 nd layer are decoding layers, and the mapping from the ascending dimension of the features to the high dimension is completed; 2, a typical sparse self-encoder is arranged between the 2 layer and the 5 layer, the encoding extraction of abstract high-level features is realized, and the last layer is a softmax output layer; the network model and the typical deep neural network model provided by the invention mainly differ from a first layer to a second layer of decoding layers, and the traditional deep neural network mainly processes high-dimensional data and extracts abstract dimension reduction characteristics; based on the characteristics of small detection window and low dimensionality of a weak target, decoding operation is realized at the front end of the network firstly, so that the network has stronger representation capability; in a high-noise environment, the detection precision of the weak and small targets is higher; constructing and training a spindle type deep neural network: step 1.1: constructing a structure of a spindle-type deep neural network comprising an input layer, a decoding layer, an encoding layer and a softmax output layer; step 1.2: determining the hyper-parameters of the spindle-type deep neural network by adopting a cross validation method; step 1.3: constructing a training data set; step 1.4: inputting the training data set into the spindle-type deep neural network, and training in an unsupervised mode to obtain the weight of the initialized network to finish training. The sizes of the layers of the trained network model are [81, 512, 256, 121, 81, 1], wherein the input layers { I1, I2, …, IN } are linear arrangements of weak and small target detection window pixels IN the image; feature transformation, we adopt a sparse autoencoder, and the calculation formula is as follows:
Figure GDA0003269489500000081
Figure GDA0003269489500000082
wherein x is(i)Indicating that, given an input of x, the self-coding neural network hides the degree of activation of neuron j,
Figure GDA0003269489500000083
representing the average liveness of the hidden neuron j; ρ represents a sparsity parameter and β represents a hyperparameter.
Rho is set to be 0.5, beta is set to be 3, the network parameters are learned by adopting a gradient descent method, and the learning rate is set to be 0.01; during training, the label of a small target sample is set to be 1, the label of a background sample is set to be 0, all area input networks of an image are judged by adopting a sliding window method, the result of output layer logistic regression is adopted as a window coordinate point response amplitude, the more values, the higher the probability that the detection window contains the small target is, and the lower the probability is, and the lower the probability is, and the lower the probability is, the lower the probability is, the lower the probability is, the probability isThe amplitude map of the target enhancement and the background suppression obtained by the deep neural network is Io(ii) a The amplitude diagram adopts a constant false alarm rate method to complete the detection of the weak and small targets; acquiring a training data set by adopting a simulation method, artificially adding weak and small targets in 220 images which do not contain the weak and small targets, and simulating to construct a training set; randomly generating coordinate points in each image, extracting a 9 x 9 region as a background sample, and adding a simulation target in the background sample by using a two-dimensional Gaussian intensity model as a target sample, wherein the two-dimensional Gaussian model is as follows:
Figure GDA0003269489500000084
wherein (x)0,y0) Representing the center position of the target image, s (i, j) representing the pixel value of the target image at position (i, j), sERepresenting the intensity of the generated object, and having a value of (0, 1)]Random number between, σxAnd σyRespectively representing horizontal and vertical dispersion parameters, having values between 0, 2]In the meantime.
Weak targets with different signal-to-noise ratios are generated by adjusting different parameters, the signal-to-noise ratio of the weak targets generated in the method is between [0 and 120], all samples are basically and uniformly distributed in the range of the signal-to-noise ratio interval, 26400 positive samples and 26400 negative samples are included, a part of generated training sample images are shown in fig. 6, and part of data of a simulation test data set is shown in fig. 7.
The test set comprises a simulation test set and a real data test set, the image data comprises millimeter wave images and infrared images, the infrared images are from a plurality of data sets, the simulation test data set adopts a weak and small target simulation method with the same training data, weak and small targets with different signal-to-noise ratios are added into a background image at random, 1920 weak and small targets are added into 32 images in total, and the signal-to-noise ratios of the 1920 weak and small targets are approximately and uniformly distributed between [0-120] Db; inputting test set data into a trained spindle-shaped neural network to complete testing, firstly mapping low-dimensional weak and small target block features to a high-dimensional space through the spindle-shaped neural network, then extracting high-resolution features through a coding neural network to complete background and target discrimination, obtaining an image with enhanced background suppression targets according to the strength output by the network discrimination, and finally completing detection of the weak and small targets by adopting a detection method based on constant false alarm rate, thereby effectively improving the detection precision of the weak and small targets.
Example 2
The step 1 comprises the following steps:
step 1.1: constructing a structure of a spindle-type deep neural network comprising an input layer, a decoding layer, an encoding layer and a softmax output layer;
step 1.2: determining the hyper-parameters of the spindle-type deep neural network by adopting a cross validation method to obtain the spindle-type deep neural network;
step 1.3: constructing a training data set;
step 1.4: inputting the training data set into the spindle-type deep neural network, and training in an unsupervised mode to obtain the weight of the initialized network to finish training.
The decoding layer and the coding layer are trained and calculated as follows:
hk＝σ(WkX+bk)
wherein, WkRepresenting a weight matrix, bkRepresenting a bias vector, sigma representing an activation function, X ═ X1,x2,...,xmDenotes the input of the current layer, hkRepresenting the output of the current layer.
The step 2 comprises the following steps:
step 2.1: after the acquired weak and small target image is input into the trained spindle-type deep neural network, setting the label of the weak and small target sample as 1 and the label of the background sample as 0, wherein the image is as follows:
Figure GDA0003269489500000091
where s denotes the sensor acquired image, stRepresenting the target signal, sbRepresenting the background signal and n representing noise.
Step 2.2: judging the weak and small target image by adopting a sliding window method to obtain the probability value of the weak and small target in the image;
step 2.3: and obtaining target enhancement and background suppression amplitude maps corresponding to a plurality of weak and small targets by using the output layer logistic regression result, namely a plurality of probability values as window coordinate point response amplitude values.
The step 3 comprises the following steps:
step 3.1: extracting sub-blocks of each value in the amplitude map by adopting a sliding window
Figure GDA0003269489500000101
Detecting, inputting the amplitude value obtained by the network;
step 3.2: counting each detected amplitude value by adopting a constant false alarm rate to obtain a false alarm probability;
Figure GDA0003269489500000102
wherein T represents a threshold value of likelihood ratio detection,
Figure GDA0003269489500000103
representing the mean of the sliding window sub-blocks, P representing the number of points within a window sub-block, PfaIndicating the false alarm probability, tau, of a constant false alarm rate detection settingCFARDenotes a detection threshold value, F1,p-1(τCFAR) A cumulative distribution function representing a central F random variable;
step 3.3: the total number of candidate targets is calculated according to the false alarm probability, the false alarm probability is ranked from high to low, and the bit target is determined from a plurality of amplitude values.
The construction of the training data set in step 1.3 comprises the following steps:
step 1.3.1: randomly generating coordinate points in the image without the weak and small targets as simulation targets, and extracting an N x N window area as a background sample;
step 1.3.2: in the background sample, a two-dimensional Gaussian intensity model and a simulation target are used as target samples, wherein the two-dimensional Gaussian model is as follows:
Figure GDA0003269489500000104
wherein (x)0,y0) Representing the center position of the target image, s (i, j) representing the pixel value of the target image at position (i, j), sERepresenting the intensity of the generated object, and having a value of (0, 1)]Random number between, σxAnd σyRespectively representing horizontal and vertical dispersion parameters, having values between 0, 2]To (c) to (d);
step 1.3.3: and adjusting different parameters of the target sample to generate weak targets with different signal-to-noise ratios to complete the construction of the training data set.
The precision of the method is embodied by comparing and analyzing the detection effects of other mainstream algorithms and the detection effect of the method. Two types of curves are adopted as evaluation indexes, the first type of curve is an ROC curve, and the ROC curve reflects the detection probability P in target detectiondAnd false alarm rate PfaThe larger the area under the ROC curve is, the better the detection performance is, PdAnd PfaThe calculation formula of (a) is as follows:
Figure GDA0003269489500000111
wherein N istIndicating the number of correctly detected targets, NaDenotes the total number of targets, NfRepresenting the number of false objects detected and N representing the number of all pixel points in the image.
The second type of curve is the detection probability PdVariation with SNR of the signal-to-noise ratio, P as the SNR value increasesdWill gradually become larger, and finally approach to 1, the SNR calculation formula adopted is:
Figure GDA0003269489500000112
wherein, gtMean value, g, representing pixels of the target local areabAnd σbRespectively representing pixels of local areas of the backgroundMean and standard deviation.
Several mainstream algorithms detect effects including: ACSDM, CSCD, SR, ISTCR, ISTCSR-CSCD, the application represents DL algorithm; as shown in fig. 8: from left to right represent ACSDM, CSCD, SR, ISTCR, ISTCSR-CSCD and DL algorithms, respectively; the real position of the target is represented by a green solid line frame, the output of the detector is represented by a red frame, if the output position of the detector is superposed with the real position of the target, the green frame is covered by the red frame, the target point is only a real red frame, only the target of the green frame shows that the detector has missed detection under the specified false alarm rate, and the DL algorithm is shown to have the largest number of detected real targets and the smallest number of missed targets relative to other algorithms;
the results of the quantitative analysis of the 6 algorithms are shown in FIGS. 9-12, and FIGS. 9-12 depict the SNR (SNR) under various signal to noise ratios<10，10<SNR<20，20<SNR<30，30<SNR<40) Probability of detection PdAnd false alarm rate PfaThe solid line marked as star is the result of the DL algorithm proposed herein, and it can be seen that the algorithm of the present application is generally superior to the other 5 algorithms in 4 different SNR intervals; at SNR < 10, at Pfa＝1×10-4In time, the ISTCR algorithm is superior to the method, the best result is achieved, and in other cases, the method Dl has better detection precision; at 10 < SNR < 20, the detection rate of the algorithm herein is much higher at each false alarm rate than the rest of the methods, DL being about 20% higher than the second ranked method, CSCD; at 20 < SNR < 30, about 12% higher than the best of the same class; when SNR is more than 30 and less than 40, the method is also obviously superior to the similar method, and is about 8 percent higher;
FIGS. 13-16 are shown at the same PfaUnder P, PdThe solid line labeled star represents the DL algorithm proposed in this application as a function of SNR, and it can be seen that the DL algorithm achieves the best results under 4 different constant false alarm rates, especially under Pfa>10×10-4While, the process herein is on average 20% higher than the best of the same class; at Pfa＝1×10-4In time, the performance of the detection methods of the three algorithms of DL, ISTCR and ISTCR-CSCD is relatively close, butIs obviously superior to the other three algorithms. Fig. 17 is an enlarged schematic view of the detection effect of the DL algorithm, because the detection uses colors to distinguish the real position frame from the detection output frame, the effect may not be obvious after performing the black-and-white process according to the patent law, and a color image can be provided if necessary.
The method comprises the steps of learning the characteristics of a weak and small target under a complex background by constructing a spindle neural network structure, judging an image block by a deep neural network to output a target probability, wherein the target region has higher probability and the background region has lower probability, constructing a target intensity map according to the probability, and detecting and positioning the target by using a Constant False Alarm Rate (CFAR) method.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents and improvements made within the spirit and principle of the present invention are intended to be included within the scope of the present invention.

Claims (4)

1. A weak and small target detection method based on a feature mapping neural network is characterized by comprising the following steps: the method comprises the following steps:
step 1: constructing and training a spindle-type deep neural network;
step 2: inputting the collected images of the weak and small targets into a trained spindle-type deep neural network to obtain an amplitude map of target enhancement and background suppression;
and step 3: the amplitude diagram adopts a constant false alarm rate method to complete the detection of the weak and small targets;
wherein, the step 2 comprises the following steps:
step 2.1: inputting the collected small target image into a trained spindle-type deep neural network, setting a small target sample label as 1 and setting a background sample label as 0;
step 2.2: judging the weak and small target image by adopting a sliding window method to obtain the probability value of the weak and small target in the image;
step 2.3: obtaining target enhancement and background suppression amplitude maps corresponding to a plurality of weak and small targets by using the output layer logistic regression result, namely a plurality of probability values as window coordinate point response amplitude values;
wherein, the step 3 comprises the following steps:
step 3.1: extracting sub-blocks of each value in the amplitude map by adopting a sliding window
Figure FDA0003269489490000013
Detecting, inputting the spindle type deep neural network to obtain an amplitude value;
step 3.2: counting each detected amplitude value by adopting a constant false alarm rate to obtain a false alarm probability;
Figure FDA0003269489490000011
wherein T represents a threshold value of likelihood ratio detection,
Figure FDA0003269489490000012
representing the mean of the sliding window sub-blocks, P representing the number of points within a window sub-block, PfaIndicating the false alarm probability, tau, of a constant false alarm rate detection settingCFARDenotes a detection threshold value, F1,p-1(τCFAR) A cumulative distribution function representing a central F random variable;
step 3.3: the total number of candidate targets is calculated according to the false alarm probability, the false alarm probability is ranked from high to low, and the bit target is determined from a plurality of amplitude values.
2. The method for detecting the weak and small targets based on the feature mapping neural network as claimed in claim 1, wherein: the step 1 comprises the following steps:
step 1.1: constructing a structure of a spindle-type deep neural network comprising an input layer, a decoding layer, an encoding layer and a softmax output layer;
step 1.2: determining the hyper-parameters of the spindle-type deep neural network by adopting a cross validation method to obtain the spindle-type deep neural network;
step 1.3: constructing a training data set;
step 1.4: inputting the training data set into the spindle-type deep neural network, and training in an unsupervised mode to obtain the weight of the initialized network to finish training.
3. The method for detecting the weak and small targets based on the feature mapping neural network as claimed in claim 2, wherein: the decoding layer and the coding layer are trained and calculated as follows:
hk＝σ(WkX+bk)
wherein, WkRepresenting a weight matrix, bkRepresenting a bias vector, sigma representing an activation function, X ═ X1,x2,...,xmDenotes the input of the current layer, hkRepresenting the output of the current layer.
4. The method for detecting the weak and small targets based on the feature mapping neural network as claimed in claim 2, wherein: the training data set construction in the step 1.3 comprises the following steps:
step 1.3.1: randomly generating coordinate points in the image without the weak and small targets as simulation targets, and extracting an N x N window area as a background sample;
step 1.3.2: in the background sample, a two-dimensional Gaussian intensity model and a simulation target are used as target samples, wherein the two-dimensional Gaussian model is as follows:
Figure FDA0003269489490000021
wherein (x)0,y0) Representing the center position of the target image, s (i, j) representing the pixel value of the target image at position (i, j), sERepresenting the intensity of the generated object, and having a value of (0, 1)]Random number between, σxAnd σyRespectively representing horizontal and vertical dispersion parameters, having values between 0, 2]To (c) to (d);
step 1.3.3: and adjusting different parameters of the target sample to generate weak targets with different signal-to-noise ratios to complete the construction of the training data set.
CN201810729648.0A 2018-07-05 2018-07-05 Weak and small target detection method based on feature mapping neural network Active CN109002848B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810729648.0A CN109002848B (en) 2018-07-05 2018-07-05 Weak and small target detection method based on feature mapping neural network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810729648.0A CN109002848B (en) 2018-07-05 2018-07-05 Weak and small target detection method based on feature mapping neural network

Publications (2)

Publication Number Publication Date
CN109002848A CN109002848A (en) 2018-12-14
CN109002848B true CN109002848B (en) 2021-11-05

Family

ID=64598687

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810729648.0A Active CN109002848B (en) 2018-07-05 2018-07-05 Weak and small target detection method based on feature mapping neural network

Country Status (1)

Country Link
CN (1) CN109002848B (en)

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109002848B (en) * 2018-07-05 2021-11-05 西华大学 Weak and small target detection method based on feature mapping neural network
CN109767434B (en) * 2019-01-07 2023-04-07 西安电子科技大学 Time domain weak and small target detection method based on neural network
CN109902715B (en) * 2019-01-18 2022-09-06 南京理工大学 Infrared dim target detection method based on context aggregation network
CN110210518B (en) * 2019-05-08 2021-05-28 北京互金新融科技有限公司 Method and device for extracting dimension reduction features
CN112288778B (en) * 2020-10-29 2022-07-01 电子科技大学 Infrared small target detection method based on multi-frame regression depth network
CN113327253B (en) * 2021-05-24 2024-05-24 北京市遥感信息研究所 Weak and small target detection method based on satellite-borne infrared remote sensing image
CN115222775B (en) * 2022-09-15 2022-12-06 中国科学院长春光学精密机械与物理研究所 Weak and small target detection tracking device and detection tracking method thereof

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2005103929A1 (en) * 2004-04-20 2005-11-03 Pluck Corporation Method, system, and computer program product for sharing information within a global computer network
CN103812577A (en) * 2012-11-06 2014-05-21 西南交通大学 Method for automatically identifying and learning abnormal radio signal type
CN105844627A (en) * 2016-03-21 2016-08-10 华中科技大学 Sea surface object image background inhibition method based on convolution nerve network
CN106056097A (en) * 2016-08-17 2016-10-26 西华大学 Millimeter wave weak small target detection method
CN107067413A (en) * 2016-12-27 2017-08-18 南京理工大学 A kind of moving target detecting method of time-space domain statistical match local feature
CN108122003A (en) * 2017-12-19 2018-06-05 西北工业大学 A kind of Weak target recognition methods based on deep neural network
CN109002848A (en) * 2018-07-05 2018-12-14 西华大学 A kind of detection method of small target based on Feature Mapping neural network
CN112507840A (en) * 2020-12-02 2021-03-16 中国船舶重工集团公司第七一六研究所 Man-machine hybrid enhanced small target detection and tracking method and system
CN113065558A (en) * 2021-04-21 2021-07-02 浙江工业大学 Lightweight small target detection method combined with attention mechanism

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2005103929A1 (en) * 2004-04-20 2005-11-03 Pluck Corporation Method, system, and computer program product for sharing information within a global computer network
CN103812577A (en) * 2012-11-06 2014-05-21 西南交通大学 Method for automatically identifying and learning abnormal radio signal type
CN105844627A (en) * 2016-03-21 2016-08-10 华中科技大学 Sea surface object image background inhibition method based on convolution nerve network
CN106056097A (en) * 2016-08-17 2016-10-26 西华大学 Millimeter wave weak small target detection method
CN107067413A (en) * 2016-12-27 2017-08-18 南京理工大学 A kind of moving target detecting method of time-space domain statistical match local feature
CN108122003A (en) * 2017-12-19 2018-06-05 西北工业大学 A kind of Weak target recognition methods based on deep neural network
CN109002848A (en) * 2018-07-05 2018-12-14 西华大学 A kind of detection method of small target based on Feature Mapping neural network
CN112507840A (en) * 2020-12-02 2021-03-16 中国船舶重工集团公司第七一六研究所 Man-machine hybrid enhanced small target detection and tracking method and system
CN113065558A (en) * 2021-04-21 2021-07-02 浙江工业大学 Lightweight small target detection method combined with attention mechanism

Non-Patent Citations (10)

* Cited by examiner, † Cited by third party
Title
Dim and small target detection based on feature mapping neural networks;Gao, Zhisheng等;《Journal of visual communication & image representation》;20190722;第62卷;第206-216页 *
Dim small target detection in single frame complex background;Long Geng等;《2016 2nd IEEE International Conference on Computer and Communications (ICCC)》;20170511;第391-395页 *
Space debris detection methods utilizing hyperspectral sequence analysis based on Hilbert-Huang transformation;Miao Zhang等;《2012 IEEE International Instrumentation and Measurement Technology Conference Proceedings》;20120702;第1-5页 *
具有高斯噪声不变性的特征描述算子;高志升等;《计算机工程与应用》;20111101;第47卷(第31期);第4-6页 *
基于卷积神经网络与高斯混合建模的红外弱小目标检测方法;娄康等;《第三十八届中国控制会议论文集（7）》;20190727;第181-186页 *
基于卷积神经网络的红外弱小目标检测算法;于周吉;《光学与光电技术》;20201010;第18卷(第5期);第63-67页 *
基于自适应背景估计的复杂红外背景抑制技术;蔡智富;《中国优秀博硕士学位论文全文数据库(硕士)信息科技辑》;20060315(第03期);第I136-20页 *
红外弱小目标检测算法综述;流浪的Mook;《https://blog.csdn.net/m0_51625082/article/details/118963429》;20210721;第1页 *
融合全卷积神经网络和视觉显著性的红外小目标检测;刘俊明等;《光子学报》;20200904;第49卷(第7期);第40-50页 *
采用目标背景建模的毫米波弱小目标检测;高志升等;《光学精密工程》;20161015;第24卷(第10期);第2601-2611页 *

Also Published As

Publication number Publication date
CN109002848A (en) 2018-12-14

Similar Documents

Publication Publication Date Title
CN109002848B (en) Weak and small target detection method based on feature mapping neural network
CN109522857B (en) People number estimation method based on generation type confrontation network model
CN106778595B (en) Method for detecting abnormal behaviors in crowd based on Gaussian mixture model
CN108764085B (en) Crowd counting method based on generation of confrontation network
CN113050042B (en) Radar signal modulation type identification method based on improved UNet3+ network
CN110188774B (en) Eddy current scanning image classification and identification method based on deep learning
CN112435221A (en) Image anomaly detection method based on generative confrontation network model
CN110348399B (en) Hyperspectral intelligent classification method based on prototype learning mechanism and multidimensional residual error network
CN106295124B (en) The method of a variety of image detecting technique comprehensive analysis gene subgraph likelihood probability amounts
CN105608446B (en) A kind of detection method and device of video flowing anomalous event
CN108229458A (en) A kind of intelligent flame recognition methods based on motion detection and multi-feature extraction
CN107563355A (en) Hyperspectral abnormity detection method based on generation confrontation network
CN112580590A (en) Finger vein identification method based on multi-semantic feature fusion network
CN109902564B (en) Abnormal event detection method based on structural similarity sparse self-coding network
CN113011357A (en) Depth fake face video positioning method based on space-time fusion
CN108171119B (en) SAR image change detection method based on residual error network
CN108960142B (en) Pedestrian re-identification method based on global feature loss function
CN104298992A (en) Self-adaptive scale pedestrian re-identification method based on data driving
CN115994907B (en) Intelligent processing system and method for comprehensive information of food detection mechanism
CN115953666B (en) Substation site progress identification method based on improved Mask-RCNN
CN106557740A (en) The recognition methods of oil depot target in a kind of remote sensing images
CN114266982A (en) SAR image change detection method based on network parameter and structure collaborative optimization
CN110458219B (en) phi-OTDR vibration signal identification algorithm based on STFT-CNN-RVFL
Kaur et al. Classification of biological species based on leaf architecture
CN111222576B (en) High-resolution remote sensing image classification method

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
TR01 Transfer of patent right
TR01 Transfer of patent right

Effective date of registration: 20230413

Address after: 1002, Building A, Zhiyun Industrial Park, No. 13 Huaxing Road, Henglang Community, Dalang Street, Longhua District, Shenzhen City, Guangdong Province, 518110

Patentee after: Shenzhen Wanzhida Technology Co.,Ltd.

Address before: 610039, No. 999, Jin Zhou road, Jinniu District, Sichuan, Chengdu

Patentee before: XIHUA University

Effective date of registration: 20230413

Address after: Room 3303-3306, Building 2, Binjiang Business Center, No. 1888 Ganjiang South Avenue, Honggutan District, Nanchang City, Jiangxi Province, 330000

Patentee after: Jiangxi Chengan Technology Co.,Ltd.

Address before: 1002, Building A, Zhiyun Industrial Park, No. 13 Huaxing Road, Henglang Community, Dalang Street, Longhua District, Shenzhen City, Guangdong Province, 518110

Patentee before: Shenzhen Wanzhida Technology Co.,Ltd.