CN108009525B - A kind of specific objective recognition methods over the ground of the unmanned plane based on convolutional neural networks - Google Patents
A kind of specific objective recognition methods over the ground of the unmanned plane based on convolutional neural networks Download PDFInfo
- Publication number
- CN108009525B CN108009525B CN201711422364.9A CN201711422364A CN108009525B CN 108009525 B CN108009525 B CN 108009525B CN 201711422364 A CN201711422364 A CN 201711422364A CN 108009525 B CN108009525 B CN 108009525B
- Authority
- CN
- China
- Prior art keywords
- convolutional neural
- neural network
- network model
- training
- ground
- 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
- 238000013527 convolutional neural network Methods 0.000 title claims abstract description 84
- 238000000034 method Methods 0.000 title claims abstract description 45
- 238000012549 training Methods 0.000 claims abstract description 61
- 238000012360 testing method Methods 0.000 claims abstract description 28
- 238000012795 verification Methods 0.000 claims abstract description 8
- 230000008859 change Effects 0.000 claims abstract description 7
- 230000006870 function Effects 0.000 claims description 30
- 238000004422 calculation algorithm Methods 0.000 claims description 14
- 238000013135 deep learning Methods 0.000 description 6
- 238000013528 artificial neural network Methods 0.000 description 5
- 238000013507 mapping Methods 0.000 description 4
- 210000002569 neuron Anatomy 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 238000000605 extraction Methods 0.000 description 3
- 230000004913 activation Effects 0.000 description 2
- 238000002790 cross-validation Methods 0.000 description 2
- 230000007423 decrease Effects 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000003062 neural network model Methods 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 238000003909 pattern recognition Methods 0.000 description 2
- 238000012706 support-vector machine Methods 0.000 description 2
- 208000035473 Communicable disease Diseases 0.000 description 1
- 230000003213 activating effect Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 208000015181 infectious disease Diseases 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000001902 propagating effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/13—Satellite images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Astronomy & Astrophysics (AREA)
- Remote Sensing (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The present invention relates to a kind of unmanned plane based on convolutional neural networks, specific objective recognition methods is as follows over the ground：(1) the collected data set of specific objective over the ground of unmanned plane is labeled according to classification, and proportionally divides training set, verification collection, test set；(2) setting training convolutional neural networks model parameter, training parameter, start to train, according to training, obtain the optimal solution of convolutional neural networks model.(3) change convolutional neural networks model depth, restart (2) training, obtain optimal convolutional neural networks model；(4) test set is tested, obtains recognition accuracy；(5) the convolutional neural networks model parameter that accuracy rate is met the requirements is applied to unmanned plane over the ground in specific objective actual scene, unmanned plane the image collected target is identified.
Description
Technical Field
The invention relates to a method for identifying a specific target by an unmanned aerial vehicle based on a convolutional neural network, and belongs to the field of intersection of aviation and computer vision information processing.
Background
Unmanned Aerial Vehicles (UAVs) are unmanned aircraft that are operated by radio remote control devices and self-contained program control devices, or are operated autonomously, either completely or intermittently, by an onboard computer. Unmanned aerial vehicles can be classified into military and civil according to application fields. For military use, unmanned aerial vehicles divide into reconnaissance aircraft and target drone. In the civil aspect, the unmanned aerial vehicle + the industry application is really just needed by the unmanned aerial vehicle; at present, the unmanned aerial vehicle is applied to the fields of aerial photography, agriculture, plant protection, miniature self-timer, express transportation, disaster relief, wild animal observation, infectious disease monitoring, surveying and mapping, news reporting, power inspection, disaster relief, film and television shooting and the like, and the application of the unmanned aerial vehicle is greatly expanded.
In traditional target recognition, a certain amount of artificial features are extracted from an image, then the image is expressed by a mathematical model, and then the image is recognized by a classifier. For example, object recognition uses Scale-invariant feature transform (SIFT), face recognition uses Local texture features (LBP), pedestrian detection uses Histogram of Oriented Gradient features (HOG), and conventional classifiers mainly include Support Vector Machines (SVMs), K-nearest neighbors (KNNs), and the like. However, the recognition of such shallow machine learning methods is low, and the engineering requirements of unmanned aerial vehicles on ground target identification cannot be met generally.
With the development of artificial intelligence and the continuous breakthrough of deep learning, great success is achieved in the fields of voice recognition, natural language processing, computer vision, video analysis, multimedia and the like. The convolutional neural network target identification process based on deep learning mainly comprises the following steps: inputting the image into a neural network, minimizing a loss function by utilizing algorithms such as forward propagation and backward propagation errors of deep learning, updating a weight value to obtain a better identification model, and identifying a new image by utilizing the model. The convolutional neural network model is able to automatically learn features from the image data and is able to be quickly trained from new training data to learn new feature representations. The general pattern recognition system comprises two important parts of feature extraction and classifier, the optimization of the two parts is separated in the traditional method, and under the framework of a neural network, the feature extraction and classifier are optimized in a joint feedback mode, and the performance of joint cooperation of the feature extraction and the classifier can be played as much as possible.
However, the target identification method based on the convolutional neural network has strong dependence on a training data set, and the size of the image data of the ground target acquired by the unmanned aerial vehicle is uncertain, so that the training requirement of the convolutional neural network model cannot be met necessarily, or the information of the data set is wasted to a certain extent due to insufficient network depth. Therefore, the invention combines the characteristics of the unmanned aerial vehicle collected image and the advantages of the convolutional neural network model, adjusts the depth of the network model according to the data set to obtain the optimal network model structure, and provides the convolutional neural network target identification method which can adapt to the size of the unmanned aerial vehicle collected image data set.
Disclosure of Invention
The invention solves the problems: the method for recognizing the specific target of the unmanned aerial vehicle on the ground overcomes the problems that the recognition accuracy of the traditional method for recognizing the target of the unmanned aerial vehicle on the other aspects is not enough and the training data set of the deep learning method is difficult to meet, and provides the method for recognizing the specific target of the unmanned aerial vehicle on the ground based on the convolutional neural network. The method can also be applied to other complex pattern recognition and intelligent system designs.
The technical scheme of the invention is as follows: an unmanned aerial vehicle ground specific target identification method based on a convolutional neural network comprises the following steps:
(1) the method comprises the steps that a ground specific target image data set is acquired by using an unmanned aerial vehicle, is labeled according to categories, is divided into a training set, a verification set and a test set according to the labeled data set, and is finally processed into data types which can be identified by a convolutional neural network model;
(2) constructing an identification convolutional neural network model, setting maximum iteration times, learning rate and test frequency, selecting a back propagation method, starting training according to the setting and selection, and then obtaining the identification accuracy rate of the convolutional neural network model according to the change condition of a training loss function;
(3) on the basis of the structure of the convolutional neural network model, increasing or reducing the number of convolutional layers, restarting the training in the step (2), and when the recognition accuracy of the training convolutional neural network model reaches the highest, indicating that the trained convolutional neural network model adapts to the size of the current data set, and reserving the structure and parameters of the convolutional neural network model obtained by the training at the moment;
(4) testing the test set by using the convolutional neural network model obtained in the step (3) to obtain the identification accuracy, judging the identification accuracy, and if the identification accuracy can meet the actual engineering requirement, applying the convolutional neural network model to the actual task of identifying the specific target of the unmanned aerial vehicle to the ground to execute the step (5); if not, the training set cannot meet the actual engineering requirements, the training set needs to be expanded, and the steps (1), (2) and (3) are restarted until the actual engineering requirements are met;
(5) and applying the parameters of the convolutional neural network model with the accuracy meeting the actual engineering requirements to the actual scene of the specific target of the unmanned aerial vehicle, and identifying the image target acquired by the unmanned aerial vehicle.
In the step (2), the method for obtaining the convolutional neural network model of the parameter optimal solution comprises the following steps: and when the Loss function Loss of the training set does not exceed 0.001 in descending amplitude and the Loss function Loss of the verification set tends to rise at a critical point, obtaining the convolutional neural network model of the parameter optimal solution.
In the step (2), the maximum iteration number, the learning rate and the test frequency are set, and the back propagation method is selected as follows:
maximum number of iterations: 20 ten thousand times;
initial learning rate: 0.001;
testing frequency: 1000 iterations/1;
the back propagation method comprises the following steps: a random gradient descent algorithm.
And (3) the convolutional neural network model in the step (2) is 5 convolutional layers and is added with 3 full-connection layers.
And (4) increasing or reducing the number of layers in the step (3) to be no more than two.
Compared with the prior art, the invention has the advantages that:
(1) according to the method, the convolutional neural network model is adopted to identify the specific ground target of the unmanned aerial vehicle, the unmanned aerial vehicle has various specific ground target types, such as ships, pedestrians and vehicles, the characteristics of the targets are greatly changed, compared with the traditional method, the manually designed characteristics are difficult to completely express target information, the convolutional neural network is utilized to perform feedback learning, characteristic expression with robustness can be learned, and therefore the requirement on identification accuracy is met;
(2) according to the convolutional neural network model training method, a global optimal solution is obtained through training loss function change and testing loss function change conditions, compared with a general target identification method based on a convolutional neural network, the fixed iteration times are set through priori information, the training efficiency can be improved, and meanwhile, the identification accuracy of a convolutional neural network model can also be improved;
(3) the different-depth convolutional neural network model construction method provided by the invention changes the depth of the network model for training and testing aiming at different data sets to obtain the convolutional neural network model with the optimal depth. The general target identification method of the convolutional neural network model has the advantages that the structure of the network model is usually fixed, the existing data set cannot well meet the training requirement of the network model, or the information of the data set is wasted, the method can adapt to different data sets, and therefore the optimization of the structure and the parameters of the convolutional neural network model is achieved.
Drawings
FIG. 1 is a schematic flow chart of the method for identifying a specific target of an unmanned aerial vehicle on the ground based on a convolutional neural network;
FIG. 2 is a diagram of a convolutional neural network model architecture of the present invention;
FIG. 3 is a graph of the loss function variation of the convolutional neural network model during the training process in the present invention;
FIG. 4 is a graph showing the change in the accuracy of the convolutional neural network model during the training process of the present invention.
FIG. 5 is a graph of the test results of the present invention.
The reference numbers and symbols in the figures are as follows:
Jcv(θ) -cross validation set loss function variation curves;
Jtrain(θ) -training set loss function variation curve;
iteration-the number of training iterations;
test accuracy-training is the test accuracy variation curve.
Detailed Description
The present invention will be described in detail below with reference to the accompanying drawings and examples.
Training algorithm
For the supervised identification task of the convolutional neural network, because the categories of all image samples are known in advance, the samples of different categories are distributed on different spatial regions according to the spatial distribution of the same image sample. After a long-time training of the image data set, parameters of the convolutional neural network are continuously updated, and boundary positions for dividing sample space classification are obtained to classify the images. Convolutional neural networks are essentially an input-to-output mapping, learning a function mapping that maps an input image to a k-dimensional feature vector according to certain principles. As long as the convolutional network is trained to obtain the connection weight between networks, the network can learn the mapping capability between input and output pairs by activating the function.
The training algorithm is mainly divided into a forward propagation stage and a backward propagation stage:
first, forward propagation phase: a set of samples (X, Y) is input to the network, where X represents sample data and Y represents a sample label. Through the hierarchical calculation of the network, namely the input is multiplied by the weight matrix of each layer, and the corresponding output value h is obtained after the layer-by-layer operationθ(x) In that respect In the forward propagation stage, the network randomly initializes the networkThe weight of the connection.
Second stage, back propagation stage: calculating the actual output hθ(x) The error from the corresponding ideal output Y, i.e. the value of the cost function. The weight matrix is adjusted by back propagation through a method of minimizing errors.
Second, cost function (LOSS):
for the network output signal hθ(x) And target desired signal y, which is a vector of dimension k. The error between the neural network output and the actual value, i.e. the loss function, is calculated by using the euclidean distance, and the cost function of the neural network can be expressed as:
and (4) calculating one layer by one layer from the first layer in a forward propagation algorithm, and solving the result of each layer obtained by the activation function until the last layer. However, to minimize the loss function value of the entire network, iterations are required to update the parameters and biases in front of the neurons. The minimum loss function of the network model is calculated by using a back propagation algorithm, namely, the error of the last layer is calculated firstly, then the error of each layer is reversely solved layer by layer, and the weight and the bias are updated by using the obtained residual error.
Third, Back Propagation (BP):
most neural network models can solve parameters of the network by using a gradient descent method, and a back propagation algorithm is needed when parameters are updated by network training samples. Firstly, initializing a weight parameter of a network into a group of random values; and secondly, performing iterative training by using the training data. Calculating the error between the output of the neural network model and the expected output, namely the loss function value, propagating the error from the last layer to the previous layer to the input, and updating the weight of each layer of network according to a gradient algorithm until the condition is met or the maximum iteration number is exceeded and stopping.
When the back propagation algorithm is applied to the convolutional neural network, the BP algorithm steps are as follows: assuming that the activation function is sigmoid function, the neural network has m layers in total, and the k layer has skOne neuron for the ith neuron of the k layer, WijRepresenting the weight coefficient associated therewith, s exists in the previous layerk-1A weight coefficient Wi1,Wi2,···,Wij,···,Is connected thereto, biIndicating the bias.
The execution of the BP algorithm comprises the following steps:
(1) for the weight coefficient WijRandomly setting an initial value, inputting a sample (x, y), wherein x represents the sample, and y represents the expected output;
(2) computing the output of each layer, for the output of the ith neuron of the k layerComprises the following steps:
(3) error of learning of each layer is calculated
For the output layer there is k-m,
for other layers, there are
(4) According to the errorCorrection weight coefficient WijAnd bias bi
Wherein,denotes the (k-1) th layerjthe output, t represents the number of iterations, η represents the learning rate, and α represents the random gradient momentum.
(5) And after the weight values of all layers of the convolutional neural network are calculated, judging whether the weights are met according to conditions. If the requirements are met, the algorithm is ended; and if not, returning to the step (2) for execution.
Fourth, the solution
The method of the invention, as shown in figure 1, comprises the following steps:
(1) marking a ground target data set acquired by an unmanned aerial vehicle according to categories, dividing the marked image data set into a training set, a verification set and a test set according to a certain proportion, namely 3:1:1, and then processing the image data set into data types which can be identified by a convolutional neural network model.
(2) And (3) constructing a convolutional neural network model, wherein as shown in fig. 2, the front five layers are convolutional layers, the rear two layers are full-connection layers, and the last layer is a classified output layer, wherein 4 times of downsampling is carried out between the input layer and the first convolutional layer, 2 times of downsampling is carried out between the second convolutional layer and the third convolutional layer, and 2 times of downsampling is carried out between the fifth convolutional layer and the first full-connection layer. Setting training parameters and training methods, for example: maximum iteration times, learning rate, test frequency, back propagation method, etc., to begin training the model. And obtaining the optimal solution at the moment according to the change condition of the training loss function. When the Loss function (Loss) of the training set slowly decreases (the variation amplitude is less than 0.001), and the Loss function (Loss) of the verification set tends to rise at a critical point, the network parameters reach the global optimum, and the convolutional neural network model parameters obtained by training at the moment are stored.
(3) On the structure of the convolutional neural network model, the number of convolutional layers of the convolutional neural network model is properly increased or reduced (the number of the convolutional layers is increased or reduced by no more than two layers), the training is restarted, when the identification accuracy of the convolutional neural network model reaches the highest, the convolutional neural network model is relatively suitable for the size of the current data set, and the structure and the parameters of the convolutional neural network model at the moment are reserved.
(4) And (3) testing the test set by using the convolutional neural network model obtained in the step (3) to obtain the identification accuracy, applying the convolutional neural network model to an actual identification task when the identification accuracy meets the actual engineering requirement, if the identification accuracy does not meet the actual engineering requirement, indicating that the training data set cannot meet the actual engineering requirement, expanding the data set, and restarting the steps (1), (2) and (3).
(5) And applying the convolutional neural network model parameters with the accuracy meeting the requirements to the actual scene of the specific target on the ground of the unmanned aerial vehicle, and identifying the image target acquired by the unmanned aerial vehicle.
The performance of the unmanned aerial vehicle ground-specific target identification method based on deep learning provided by the invention is verified through a specific embodiment, an image data set to be identified is acquired by an unmanned aerial vehicle vision system and is used as an object to be verified, and a deep learning framework based on cafe 1 is adopted.
As shown in fig. 1, the method comprises the following specific steps:
(1) preprocessing the image:
the image data set acquired by the unmanned aerial vehicle is classified into 3 types of targets (ships, pedestrians and vehicles) with 6000 sheets in total, images in the data set are labeled according to the types, and the labeled image data set is divided into a training set, a cross validation set and a test set according to the ratio of 3:1: 1. The image data set is then processed into a data type recognizable by the network model, converted into the lmdb format.
(2) Constructing convolutional neural network model
According to the convolutional neural network model structure, as shown in fig. 2, a prototxt file of a network structure is written, a front 5-layer network is a convolutional layer, and a rear 3-layer network is a full-connection layer. Where Stride of 4 indicates a convolution step of 4, and maxporoling indicates a method using maximum downsampling.
(3) Training a network model:
firstly, setting training parameters: the maximum iteration frequency is set to be 20 ten thousand times, the initial learning rate is 0.001, and a network model begins to be trained by adopting a random gradient descent back propagation algorithm.
And observing the training condition to obtain the optimal solution of the convolutional neural network model. As shown in fig. 3, in the Loss function variation curve of the convolutional neural network model during the training process, when the Loss function (Loss) of the training set decreases slowly, and the Loss function (Loss) of the verification set tends to rise at the critical point, the parameters of the convolutional neural network model reach the global optimal solution. As shown in fig. 4, the test accuracy of the final training model is about 0.99, which indicates that the target recognition accuracy of the convolutional neural network model is high.
(4) Modifying the network model:
adding 1 volume base layer on the structure of the convolutional neural network model shown in fig. 1, restarting the training, and finding that the accuracy of the final target recognition accuracy is slightly reduced by 0.5. 1 volume base layer is reduced on the structure of the convolutional neural network model shown in FIG. 1, the training is restarted, and the final target recognition accuracy rate is found to be slightly reduced by 0.3. The convolutional neural network model of 5 convolutional layers is relatively in accordance with the requirement of the data set, and no modification is needed.
(5) Testing the network model:
and (4) importing the network parameters obtained in the step (3) into a test network, as shown in fig. 5, for three types of targets of the unmanned aerial vehicle to the ground: and identifying the test image according to the identification results of the ship, the vehicle and the pedestrian to obtain correct identification results.
The above examples are provided only for the purpose of describing the present invention, and are not intended to limit the scope of the present invention. The scope of the invention is defined by the appended claims. Various equivalent substitutions and modifications can be made without departing from the spirit and principles of the invention, and are intended to be within the scope of the invention.
Claims (3)
1. An unmanned aerial vehicle ground specific target identification method based on a convolutional neural network is characterized in that: the method comprises the following steps:
(1) the method comprises the steps that a ground specific target image data set is acquired by using an unmanned aerial vehicle, is labeled according to categories, is divided into a training set, a verification set and a test set according to the labeled data set, and is finally processed into data types which can be identified by a convolutional neural network model;
(2) constructing an identification convolutional neural network model, setting maximum iteration times, learning rate and test frequency, selecting a back propagation method, starting training according to the setting and selection, and then obtaining the identification accuracy rate of the convolutional neural network model according to the change condition of a training loss function;
(3) on the basis of the structure of the convolutional neural network model, increasing or reducing the number of convolutional layers, restarting the training in the step (2), and when the recognition accuracy of the training convolutional neural network model reaches the highest, indicating that the trained convolutional neural network model adapts to the size of the current data set, and reserving the structure and parameters of the convolutional neural network model obtained by the training at the moment;
(4) testing the test set by using the convolutional neural network model obtained in the step (3) to obtain the identification accuracy, judging the identification accuracy, and if the identification accuracy can meet the actual engineering requirement, applying the convolutional neural network model to the actual task of identifying the specific target of the unmanned aerial vehicle to the ground to execute the step (5); if not, the training set cannot meet the actual engineering requirements, the training set needs to be expanded, and the steps (1), (2) and (3) are restarted until the actual engineering requirements are met;
(5) applying the parameters of the convolutional neural network model with the accuracy meeting the actual engineering requirements to the actual scene of the specific target of the unmanned aerial vehicle on the ground, and identifying the image target acquired by the unmanned aerial vehicle;
in the step (2), the method for obtaining the convolutional neural network model of the parameter optimal solution comprises the following steps: when the Loss function Loss of the training set does not exceed 0.001 in descending amplitude and the Loss function Loss of the verification set tends to rise at a critical point, a convolutional neural network model of the parameter optimal solution is obtained;
in the step (2), the maximum iteration times, the learning rate and the test frequency are set, and the reverse propagation method is selected as follows:
maximum number of iterations: 20 ten thousand times;
initial learning rate: 0.001;
testing frequency: 1000 iterations/1;
the back propagation method comprises the following steps: a random gradient descent algorithm.
2. The convolutional neural network-based unmanned aerial vehicle ground-specific target identification method of claim 1, wherein: and (3) identifying the convolutional neural network model as 5 convolutional layers and adding 3 full-connection layers in the step (2).
3. The convolutional neural network-based unmanned aerial vehicle ground-specific target identification method of claim 1, wherein: and (4) increasing or reducing the number of layers to be not more than two in the step (3).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711422364.9A CN108009525B (en) | 2017-12-25 | 2017-12-25 | A kind of specific objective recognition methods over the ground of the unmanned plane based on convolutional neural networks |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711422364.9A CN108009525B (en) | 2017-12-25 | 2017-12-25 | A kind of specific objective recognition methods over the ground of the unmanned plane based on convolutional neural networks |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108009525A CN108009525A (en) | 2018-05-08 |
| CN108009525B true CN108009525B (en) | 2018-10-12 |
Family
ID=62061187
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711422364.9A Active CN108009525B (en) | 2017-12-25 | 2017-12-25 | A kind of specific objective recognition methods over the ground of the unmanned plane based on convolutional neural networks |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108009525B (en) |
Families Citing this family (63)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108875593A (en) * | 2018-05-28 | 2018-11-23 | 上海交通大学 | Visible images weather recognition methods based on convolutional neural networks |
| CN108957418A (en) * | 2018-05-30 | 2018-12-07 | 西安电子科技大学 | A kind of radar target identification method based on Recognition with Recurrent Neural Network model |
| CN109086790A (en) * | 2018-06-19 | 2018-12-25 | 歌尔股份有限公司 | A kind of alternative manner of disaggregated model, device and electronic equipment |
| CN109063824B (en) * | 2018-07-25 | 2023-04-07 | 深圳市中悦科技有限公司 | Deep three-dimensional convolutional neural network creation method and device, storage medium and processor |
| CN109033095B (en) * | 2018-08-01 | 2022-10-18 | 苏州科技大学 | Target transformation method based on attention mechanism |
| CN109144099B (en) * | 2018-08-28 | 2020-10-20 | 北京航空航天大学 | Fast evaluation method for unmanned aerial vehicle group action scheme based on convolutional neural network |
| TWI709188B (en) | 2018-09-27 | 2020-11-01 | 財團法人工業技術研究院 | Fusion-based classifier, classification method, and classification system |
| CN110969072B (en) * | 2018-09-30 | 2023-05-02 | 杭州海康威视系统技术有限公司 | Model optimization method, device and image analysis system |
| CN110163236B (en) * | 2018-10-15 | 2023-08-29 | 腾讯科技（深圳）有限公司 | Model training method and device, storage medium and electronic device |
| CN109657541A (en) * | 2018-11-09 | 2019-04-19 | 南京航空航天大学 | A kind of ship detecting method in unmanned plane image based on deep learning |
| CN109657596A (en) * | 2018-12-12 | 2019-04-19 | 天津卡达克数据有限公司 | A kind of vehicle appearance component identification method based on deep learning |
| CN109815798A (en) * | 2018-12-17 | 2019-05-28 | 广东电网有限责任公司 | Unmanned plane image processing method and system |
| CN109685017B (en) * | 2018-12-26 | 2023-02-03 | 中山大学 | Real-time target detection system and detection method based on lightweight neural network |
| CN109726761B (en) * | 2018-12-29 | 2023-03-31 | 青岛海洋科学与技术国家实验室发展中心 | CNN evolution method, CNN-based AUV cluster working method, CNN evolution device and CNN-based AUV cluster working device and storage medium |
| CN109934088A (en) * | 2019-01-10 | 2019-06-25 | 海南大学 | Sea ship discrimination method based on deep learning |
| CN109740560B (en) * | 2019-01-11 | 2023-04-18 | 山东浪潮科学研究院有限公司 | Automatic human body cell protein identification method and system based on convolutional neural network |
| CN109829898B (en) * | 2019-01-17 | 2023-05-16 | 柳州康云互联科技有限公司 | Measurement detection system and method based on neural network for Internet detection |
| CN109886923B (en) * | 2019-01-17 | 2023-05-02 | 柳州康云互联科技有限公司 | Measurement detection system and method based on machine learning for Internet detection |
| CN109886141B (en) * | 2019-01-28 | 2023-06-06 | 同济大学 | Pedestrian re-identification method based on uncertainty optimization |
| CN111507128A (en) * | 2019-01-30 | 2020-08-07 | 北京沃东天骏信息技术有限公司 | Face recognition method and device, electronic equipment and readable medium |
| CN110059538B (en) * | 2019-02-27 | 2021-07-09 | 成都数之联科技有限公司 | Water body identification method based on deep dense neural network |
| CN109934333A (en) * | 2019-02-28 | 2019-06-25 | 天津大学 | Object localization method based on full Connection Neural Network |
| CN109871906B (en) * | 2019-03-15 | 2023-03-28 | 西安获德图像技术有限公司 | Cop appearance defect classification method based on deep convolutional neural network |
| CN110517219B (en) * | 2019-04-01 | 2023-04-07 | 中山大学中山眼科中心 | Corneal topography distinguishing method and system based on deep learning |
| CN109977902A (en) * | 2019-04-03 | 2019-07-05 | 刘西 | A kind of construction vehicle identification method based on deep learning |
| CN110245582A (en) * | 2019-05-25 | 2019-09-17 | 天津大学 | A method of based on classification component single in deep learning for identification bitmap |
| CN110163177B (en) * | 2019-05-28 | 2022-12-09 | 李峥嵘 | Unmanned aerial vehicle automatic sensing and identifying method for wind turbine generator blades |
| CN110347600B (en) * | 2019-07-11 | 2023-04-07 | 中国人民解放军陆军工程大学 | Convolutional neural network-oriented variation coverage testing method and computer storage medium |
| CN110458079A (en) * | 2019-08-05 | 2019-11-15 | 黑龙江电力调度实业有限公司 | A kind of Image Acquisition and target identification method based on FPGA and convolutional neural networks |
| CN110544249A (en) * | 2019-09-06 | 2019-12-06 | 华南理工大学 | Convolutional neural network quality identification method for arbitrary-angle case assembly visual inspection |
| CN110717260A (en) * | 2019-09-26 | 2020-01-21 | 杭州电子科技大学 | Unmanned aerial vehicle maneuvering capability model establishing method |
| CN110782078B (en) * | 2019-10-18 | 2023-04-25 | 武汉科技大学 | Learning method for predicting mud yield of trailing suction hopper dredger |
| CN111008641B (en) * | 2019-10-24 | 2023-04-07 | 云南电网有限责任公司昆明供电局 | Power transmission line tower external force damage detection method based on convolutional neural network |
| CN110749324A (en) * | 2019-10-28 | 2020-02-04 | 深圳市赛为智能股份有限公司 | Unmanned aerial vehicle rescue positioning method and device, computer equipment and storage medium |
| CN110764064A (en) * | 2019-11-08 | 2020-02-07 | 哈尔滨工业大学 | Radar interference signal identification method based on deep convolutional neural network integration |
| CN111046785B (en) * | 2019-12-10 | 2023-07-11 | 长讯通信服务有限公司 | Method for identifying key targets of inspection video of unmanned aerial vehicle based on convolutional neural network |
| CN111160226B (en) * | 2019-12-26 | 2023-03-31 | 华侨大学 | Pedestrian gender identification method based on visual angle adaptive feature learning |
| JP7421363B2 (en) * | 2020-02-14 | 2024-01-24 | 株式会社Ｓｃｒｅｅｎホールディングス | Parameter update device, classification device, parameter update program, and parameter update method |
| CN111325143A (en) * | 2020-02-18 | 2020-06-23 | 西北工业大学 | Underwater target identification method under unbalanced data set condition |
| CN111539178B (en) * | 2020-04-26 | 2023-05-05 | 成都市深思创芯科技有限公司 | Chip layout design method and system based on neural network and manufacturing method |
| CN113642592B (en) * | 2020-04-27 | 2024-07-05 | 武汉Tcl集团工业研究院有限公司 | Training method of training model, scene recognition method and computer equipment |
| CN113743168B (en) * | 2020-05-29 | 2023-10-13 | 北京机械设备研究所 | Urban flyer identification method based on micro-depth neural network search |
| CN111652326B (en) * | 2020-06-30 | 2023-04-28 | 华南农业大学 | Fruit maturity identification method and system based on MobileNet v2 network improvement |
| CN111860278B (en) * | 2020-07-14 | 2024-05-14 | 陕西理工大学 | Human behavior recognition algorithm based on deep learning |
| CN111897333B (en) * | 2020-07-31 | 2022-12-20 | 常州码库数据科技有限公司 | Robot walking path planning method |
| CN111899091B (en) * | 2020-08-06 | 2023-03-24 | 华院计算技术(上海)股份有限公司 | Overdue risk identification method based on robust algorithm |
| CN112153615B (en) * | 2020-09-15 | 2022-07-12 | 南京信息工程大学滨江学院 | Deep learning-based user association method in multi-cell cellular D2D equipment |
| CN112418266A (en) * | 2020-10-15 | 2021-02-26 | 南昌大学 | Pile foundation integrity classification and identification method based on convolutional neural network |
| CN112699928B (en) * | 2020-12-25 | 2022-09-20 | 南京理工大学 | Non-motor vehicle detection and identification method based on deep convolutional network |
| CN113505628A (en) * | 2021-04-02 | 2021-10-15 | 上海师范大学 | Target identification method based on lightweight neural network and application thereof |
| CN113076895B (en) * | 2021-04-09 | 2022-08-02 | 太原理工大学 | Conveyor belt longitudinal damage vibration sensing method based on infrared computer vision |
| CN113361682B (en) * | 2021-05-08 | 2022-08-16 | 南京理工大学 | Reconfigurable neural network training with IP protection and using method |
| CN113313021B (en) * | 2021-05-27 | 2023-05-30 | 云南电网有限责任公司电力科学研究院 | Deep learning model construction method based on low-quality image recognition |
| CN113671161B (en) * | 2021-07-13 | 2023-04-18 | 郑州大学 | Unmanned aerial vehicle pavement disease detection method based on LSTM neural network algorithm |
| CN113505851B (en) * | 2021-07-27 | 2023-01-31 | 电子科技大学 | Multitasking method for intelligent aircraft |
| CN113702719B (en) * | 2021-08-03 | 2022-11-29 | 北京科技大学 | Broadband near-field electromagnetic positioning method and device based on neural network |
| CN113327461B (en) * | 2021-08-03 | 2021-11-23 | 杭州海康威视数字技术股份有限公司 | Cooperative unmanned aerial vehicle detection method, device and equipment |
| CN113627611A (en) * | 2021-08-06 | 2021-11-09 | 苏州科韵激光科技有限公司 | Model training method and device, electronic equipment and storage medium |
| CN113673031B (en) * | 2021-08-11 | 2024-04-12 | 中国科学院力学研究所 | Flexible airship service attack angle identification method integrating strain response and deep learning |
| CN113887636B (en) * | 2021-10-09 | 2023-04-07 | 四川大学 | Selectable data enhancement method and system based on genetic algorithm |
| CN114359546B (en) * | 2021-12-30 | 2024-03-26 | 太原科技大学 | Day lily maturity identification method based on convolutional neural network |
| CN115859056B (en) * | 2022-12-29 | 2023-09-15 | 华诺星空技术股份有限公司 | Unmanned aerial vehicle target detection method based on neural network |
| CN117292283B (en) * | 2023-11-24 | 2024-02-13 | 成都庆龙航空科技有限公司 | Target identification method based on unmanned aerial vehicle |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103955702A (en) * | 2014-04-18 | 2014-07-30 | 西安电子科技大学 | SAR image terrain classification method based on depth RBF network |
| CN105512676A (en) * | 2015-11-30 | 2016-04-20 | 华南理工大学 | Food recognition method at intelligent terminal |
| CN106682569A (en) * | 2016-09-28 | 2017-05-17 | 天津工业大学 | Fast traffic signboard recognition method based on convolution neural network |
| CN107330405A (en) * | 2017-06-30 | 2017-11-07 | 上海海事大学 | Remote sensing images Aircraft Target Recognition based on convolutional neural networks |
| CN107423760A (en) * | 2017-07-21 | 2017-12-01 | 西安电子科技大学 | Based on pre-segmentation and the deep learning object detection method returned |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9940544B2 (en) * | 2016-06-08 | 2018-04-10 | Adobe Systems Incorporated | Event image curation |
-
2017
- 2017-12-25 CN CN201711422364.9A patent/CN108009525B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103955702A (en) * | 2014-04-18 | 2014-07-30 | 西安电子科技大学 | SAR image terrain classification method based on depth RBF network |
| CN105512676A (en) * | 2015-11-30 | 2016-04-20 | 华南理工大学 | Food recognition method at intelligent terminal |
| CN106682569A (en) * | 2016-09-28 | 2017-05-17 | 天津工业大学 | Fast traffic signboard recognition method based on convolution neural network |
| CN107330405A (en) * | 2017-06-30 | 2017-11-07 | 上海海事大学 | Remote sensing images Aircraft Target Recognition based on convolutional neural networks |
| CN107423760A (en) * | 2017-07-21 | 2017-12-01 | 西安电子科技大学 | Based on pre-segmentation and the deep learning object detection method returned |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108009525A (en) | 2018-05-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108009525B (en) | A kind of specific objective recognition methods over the ground of the unmanned plane based on convolutional neural networks | |
| US10332028B2 (en) | Method for improving performance of a trained machine learning model | |
| US10275719B2 (en) | Hyper-parameter selection for deep convolutional networks | |
| CN109241817B (en) | Crop image recognition method shot by unmanned aerial vehicle | |
| EP3918428A1 (en) | Automatic optimization of machine learning algorithms in the presence of target datasets | |
| US8379994B2 (en) | Digital image analysis utilizing multiple human labels | |
| EP3029606A2 (en) | Method and apparatus for image classification with joint feature adaptation and classifier learning | |
| CN113033520B (en) | Tree nematode disease wood identification method and system based on deep learning | |
| CN107423747B (en) | A kind of conspicuousness object detection method based on depth convolutional network | |
| JP7263216B2 (en) | Object Shape Regression Using Wasserstein Distance | |
| CN111507469A (en) | Method and device for optimizing hyper-parameters of automatic labeling device | |
| CN117292283B (en) | Target identification method based on unmanned aerial vehicle | |
| CN105654136B (en) | A kind of extensive remote sensing image Motion parameters method based on deep learning | |
| CN117157678A (en) | Method and system for graph-based panorama segmentation | |
| CN113359843B (en) | Unmanned aerial vehicle autonomous landing method and device, electronic equipment and storage medium | |
| US11270425B2 (en) | Coordinate estimation on n-spheres with spherical regression | |
| CN114329031B (en) | Fine-granularity bird image retrieval method based on graph neural network and deep hash | |
| CN116075820A (en) | Method, non-transitory computer readable storage medium and apparatus for searching image database | |
| CN118230175B (en) | Real estate mapping data processing method and system based on artificial intelligence | |
| KR102210404B1 (en) | Location information extraction device and method | |
| CN110472092B (en) | Geographical positioning method and system of street view picture | |
| CN115482513A (en) | Apparatus and method for adapting a pre-trained machine learning system to target data | |
| CN113553975B (en) | Pedestrian re-identification method, system, equipment and medium based on sample pair relation distillation | |
| CN113496221A (en) | Point supervision remote sensing image semantic segmentation method and system based on depth bilateral filtering | |
| Batool et al. | Ielmnet: An application for traffic sign recognition using cnn and elm |
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 |