CN108764082A - A kind of Aircraft Targets detection method, electronic equipment, storage medium and system - Google Patents
A kind of Aircraft Targets detection method, electronic equipment, storage medium and system Download PDFInfo
- Publication number
- CN108764082A CN108764082A CN201810474112.9A CN201810474112A CN108764082A CN 108764082 A CN108764082 A CN 108764082A CN 201810474112 A CN201810474112 A CN 201810474112A CN 108764082 A CN108764082 A CN 108764082A
- Authority
- CN
- China
- Prior art keywords
- image
- sample image
- aircraft targets
- aircraft
- positive sample
- 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.)
- Withdrawn
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/13—Satellite images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Multimedia (AREA)
- Astronomy & Astrophysics (AREA)
- Remote Sensing (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
The present invention provides a kind of Aircraft Targets detection method, it is included in preset data set and obtains the training image containing aircraft brake disc and obtain Aircraft Targets image to be detected in test side, first conspicuousness pre-detection is carried out to training image and handles and obtain sample image, carrying out classification to sample image handles to obtain original positive sample image and training negative sample image；Data enhancing is carried out to original positive sample image to handle to obtain trained positive sample image；It will be trained in training positive sample image and training negative sample image input convolutional neural networks model and obtain target detection model, Aircraft Targets image to be detected is input in target detection model, target detection model is detected to obtain Aircraft Targets testing result to Aircraft Targets image to be detected.A kind of Aircraft Targets detection method of the present invention realizes the precise positioning for aircraft, and the Aircraft Targets image to be detected of a variety of different application scenarios can be used by a little target detection models of depth.
Description
Technical field
The present invention relates to Airplane detection field more particularly to a kind of Aircraft Targets detection method, electronic equipment, storage mediums
And system.
Background technology
The detection of Aircraft Targets is detected using synthetic aperture radar mostly at present, because synthetic aperture radar can
It is influenced with conditions such as not climate, weather, illumination, obtains high-resolution radar image；Compared with optical sensor, synthetic aperture
Radar has more advantage in military fields such as scouting, monitoring and tracking, however SAR image imaging mechanism is complex, target by compared with
Few scattering point composition is without clear profile, and there are speckle noises for image, this makes synthetic aperture radar carry out Airplane detection
When can encounter many problems.
Traditional diameter radar image object detection method can be divided into 3 classes.The 1st kind of method based on single feature is logical
Often using the brighter part of Radar Cross Section (Radar Cross Section, RCS) information singles contrast as candidate
Target.Wherein most of detection methods do image using constant false alarm rate (Constant False Alarm Rate, CFAR) algorithm
Segmentation and candidate target positioning.2nd kind is the method based on multiple features, and various features such as geometry extends and divides shape, wavelet systems
Number etc. can merge detection target；3rd kind is the method based on priori, priori such as imaging parameters, latitude and longitude information
Testing process is added in equal needs collaboration, and such methods are more complex and apply in practice less.Current object detection method exists
The problem of precise positioning is unable to target, all application scenarios can not be suitble to.
Invention content
For overcome the deficiencies in the prior art, one of the objects of the present invention is to provide a kind of Aircraft Targets detection method,
It can solve current object detection method in the presence of precise positioning is unable to target, can not be suitble to asking for all application scenarios
Topic.
The second object of the present invention is to provide a kind of electronic equipment, can solve current object detection method and exist pair
Target is unable to precise positioning, the problem of can not being suitble to all application scenarios.
The third object of the present invention is to provide a kind of storage medium, can solve current object detection method and exist pair
Target is unable to precise positioning, the problem of can not being suitble to all application scenarios.
The fourth object of the present invention is to provide a kind of Aircraft Targets detecting system, can solve current target detection side
There is the problem of precise positioning is unable to target, all application scenarios can not be suitble in method.
An object of the present invention is realized using following technical scheme：
A kind of Aircraft Targets detection method, including:
Image obtains, and the training image containing aircraft brake disc is obtained in preset data set and is obtained in test side and is waited for
Detect Aircraft Targets image；
First pre-detection carries out the first conspicuousness pre-detection to the training image and handles and obtain sample image, to institute
It states sample image and carries out classification and handle to obtain original positive sample image and training negative sample image；
Data enhancing is handled, and carrying out data enhancing to the original positive sample image handles to obtain trained positive sample image；
Target detection model is built, the trained positive sample image and the trained negative sample image are inputted into convolutional Neural
It is trained in network model and obtains target detection model；
The Aircraft Targets image to be detected is input in the target detection model by image recognition, the target inspection
It surveys model the Aircraft Targets image to be detected is detected to obtain Aircraft Targets testing result.
Further, first pre-detection is specially：Extract the trained atlas significant characteristics obtain it is several not
With the Saliency maps picture of scale, image co-registration is carried out to several Saliency maps pictures using non-maxima suppression algorithm and is obtained
Including around Aircraft Targets and aircraft scene sample image, classification processing is carried out to the sample image, is obtained containing aircraft
The original positive sample image of target and training negative sample image containing scene around aircraft.
Further, the data enhancing processing is specially to carry out translation processing to the original positive sample image plus make an uproar
Processing, contrast enhancement processing and angle rotation processing simultaneously obtain training positive sample image.
Further, further include that the second pre-detection is carried out to the Aircraft Targets image to be detected before described image identification
Processing.
Further, second pre-detection is specially and extracts the significant characteristics of the Aircraft Targets image to be detected to obtain
To the Saliency maps picture of several different scales, image is carried out to several Saliency maps pictures using non-maxima suppression algorithm and is melted
It closes.
The second object of the present invention is realized using following technical scheme：
A kind of electronic equipment, including：Processor；
Memory；And program, wherein described program is stored in the memory, and is configured to by processor
It executes, described program includes a kind of Aircraft Targets detection method for executing the present invention.
The third object of the present invention is realized using following technical scheme：
A kind of computer readable storage medium, is stored thereon with computer program, it is characterised in that：The computer program
It is executed by processor a kind of Aircraft Targets detection method of the present invention.
The fourth object of the present invention is realized using following technical scheme：
A kind of Aircraft Targets detecting system, including：
Image collection module, described image acquisition module are used to obtain the instruction containing aircraft brake disc in preset data set
Practice image and obtains Aircraft Targets image to be detected in test side；
First pre-detection module, the first pre-detection module are used to carry out the first conspicuousness preliminary examination to the training image
It surveys and handles and obtain sample image, carrying out classification to the sample image handles to obtain original positive sample image and training negative sample
Image；
Data enhance module, and the data enhancing module is used to carry out data enhancing processing to the original positive sample image
It obtains training positive sample image；
Target detection model module is built, the structure target detection model module is used for the trained positive sample image
And it is trained in the trained negative sample image input convolutional neural networks model and obtains target detection model；
The Aircraft Targets image to be detected is input in the target detection model, the mesh by picture recognition module
Mark detection model is detected to obtain Aircraft Targets testing result to the Aircraft Targets image to be detected.
Further, further include the second pre-detection module, the second pre-detection module is used to extract described to be detected winged
The significant characteristics of machine target image obtain the Saliency maps picture of several different scales, using non-maxima suppression algorithm to several
The Saliency maps picture carries out image co-registration.
Further, described that data enhancing processing is carried out specially to the original positive sample to the original positive sample image
This image carries out translation processing, adds and make an uproar processing, contrast enhancement processing and angle rotation processing and obtain trained positive sample figure
Picture, the data enhancing module includes translation unit, adds make an uproar unit, contrast enhancement unit and angle rotary unit, described
Translation unit is used to carry out translation processing to the trained positive sample image, and described plus unit of making an uproar is for the trained positive sample
Image carries out that processing of making an uproar, the contrast enhancement unit is added to be used to carry out at contrast enhancing the trained positive sample image
Reason, the angle rotary unit are used to carry out angle rotation processing to the trained positive sample image.
Compared with prior art, the beneficial effects of the present invention are：The present invention a kind of Aircraft Targets detection method, by
The training image containing aircraft brake disc is obtained in preset data set and obtains Aircraft Targets image to be detected in test side；It is right
The training image carries out the first conspicuousness pre-detection and handles and obtain sample image, and classification processing is carried out to the sample image
Obtain original positive sample image and training negative sample image；Original positive sample image progress data enhancing is handled and is instructed
Practice positive sample image；By in the trained positive sample image and trained negative sample image input convolutional neural networks model into
Row training simultaneously obtains target detection model；The Aircraft Targets image to be detected is input in the target detection model, institute
It states target detection model the Aircraft Targets image to be detected is detected to obtain Aircraft Targets testing result；In detection process
By being first input to convolution net to the training image progress pre-detection of data concentration and data enhancing processing, then by training image
Network model is trained to obtain the convolutional network model by deep learning and is detected to Aircraft Targets image to be detected,
Therefore the testing result finally obtained is more accurate, the precise positioning for aircraft is realized, by a little target detections of depth
Model can use the Aircraft Targets image to be detected of a variety of different application scenarios.
Above description is only the general introduction of technical solution of the present invention, in order to better understand the technical means of the present invention,
And can be implemented in accordance with the contents of the specification, below with presently preferred embodiments of the present invention and after coordinating attached drawing to be described in detail such as.
The specific implementation mode of the present invention is shown in detail by following embodiment and its attached drawing.
Description of the drawings
Attached drawing described herein is used to provide further understanding of the present invention, and is constituted part of this application, this hair
Bright illustrative embodiments and their description are not constituted improper limitations of the present invention for explaining the present invention.In the accompanying drawings：
Fig. 1 is a kind of flow chart of Aircraft Targets detection method of the present invention；
Fig. 2 is a kind of module frame chart of Aircraft Targets detecting system of the present invention.
Specific implementation mode
In the following, in conjunction with attached drawing and specific implementation mode, the present invention is described further, it should be noted that not
Under the premise of conflicting, new implementation can be formed between various embodiments described below or between each technical characteristic in any combination
Example.
As shown in Figure 1, a kind of Aircraft Targets detection method of the present invention is used for for variety classes and in different scenes
Under aircraft be detected, Aircraft Targets detection method of the invention be used for military field, especially military field is for borderland
And at important military affairs deploy troops on garrison duty, a dynamic is laid for monitoring in real time；Specifically include following steps：
Image obtains, and the training image containing aircraft brake disc is obtained in preset data set and is obtained in test side and is waited for
Detect Aircraft Targets image；In embodiments of the present invention, include the aircraft of variety classes different scale in the data set of use
And the image of different scenes is as training image, almost all of military civil aircraft, the image in training set includes pair
The aircraft of type, a variety of directions；Aircraft Targets image to be detected is the image that the needs containing aircraft detect in scene.
First pre-detection carries out the first conspicuousness pre-detection to training image and handles and obtain sample image, to sample graph
It handles to obtain original positive sample image and training negative sample image as carrying out classification.Specially：The conspicuousness of extraction training atlas
Feature obtains the Saliency maps picture of several different scales, that is, extracts the different Saliency maps picture of several scale sizes, and use is non-
The image of different windows comprising same target aircraft is carried out the image that permeates by maximum restrainable algorithms, to all scales
Identical Saliency maps picture carries out image co-registration and finally obtains the sample image for including scene around Aircraft Targets and aircraft again,
Aircraft Targets are a kind of and to contain only scene around aircraft be that one kind carries out classification processing to sample image according to containing only, obtain
Training negative sample image to the original positive sample image containing Aircraft Targets and containing scene around aircraft,
Pair data enhancing is handled, and carries out data enhancing to original positive sample image and handle to obtain to train positive sample image, i.e.,
Original positive sample image carries out translation processing plus makes an uproar processing, contrast enhancement processing and angle rotation processing and to be trained
Positive sample image.Specially：Data enhancement methods are used in the present embodiment to carry out at data enhancing original positive sample image
Reason carries out translation processing under conditions of Aircraft Targets are no more than original positive sample image boundary to original positive sample image, then
Carry out plus make an uproar to original sample image processing, because the picture obtained in same place in original positive sample image may go out
Existing different brightness, therefore data enhancing is carried out using pixel contrast information, the contrast enhancing in the present embodiment is using non-
Linear transformation is realized；In the present embodiment to carrying out position rotation, this reality by the enhanced original sample image of contrast
The rotation for applying progress low-angle in example finally obtains trained positive sample image.
Target detection model is built, training positive sample image and training negative sample image are inputted into convolutional neural networks model
In be trained and obtain target detection model；It is designed firstly for convolutional neural networks in the present embodiment, the present embodiment
In convolutional neural networks include 3 convolutional layers and 3 pooling layer (pond layer) composition, the convolution kernel of the 1st convolutional layer is big
Small is 5 × 5, and has 32 output figures.Similarly, the convolution kernel size of the 2nd convolutional layer is also 5 × 5 and has 64 output figures.
It is 6 × 6 that the last one convolutional layer, which has 128 output figures, convolution kernel size,.In succession after each convolutional layer 2 × 2 Max-
Pooling layers.The size of input picture slice is 120 × 120.They become 116 × 116 after the 1st convolutional layer, the 1st
Become 58 × 58 after a pooling layers；It moves in circles, the characteristic pattern that two sizes of output are 11 × 11 before Softmax layers.
Second pre-detection processing is carried out to Aircraft Targets image to be detected, extracts the conspicuousness of Aircraft Targets image to be detected
Feature obtains the Saliency maps picture of several different scales, and image is carried out to several Saliency maps pictures using non-maxima suppression algorithm
Fusion；Specific steps are identical as step in the processing of the first pre-detection, but are not classified finally.
Aircraft Targets image to be detected is input in target detection model by image recognition, and target detection model is to be checked
Aircraft Targets image is surveyed to be detected to obtain Aircraft Targets testing result.
The present invention provides a kind of electronic equipment, including：Processor；
Memory；And program, wherein described program is stored in the memory, and is configured to by processor
It executes, described program includes a kind of Aircraft Targets detection method for executing the present invention.
A kind of computer readable storage medium, is stored thereon with computer program, and the computer program is held by processor
A kind of Aircraft Targets detection method of the row present invention.
As shown in Fig. 2, the present invention also provides a kind of Aircraft Targets detecting systems, including：
Image collection module, image collection module in preset data set for obtaining the training figure containing aircraft brake disc
As and test side obtains Aircraft Targets image to be detected；First pre-detection module, the first pre-detection module are used for training
Image carries out the first conspicuousness pre-detection and handles and obtain sample image, and carrying out classification to sample image handles to obtain original positive sample
This image and training negative sample image；Data enhance module, and data enhance module and are used to carry out data to original positive sample image
Enhancing, which handles to obtain, trains positive sample image；Target detection model module is built, structure target detection model module will be for that will instruct
Practice and is trained in positive sample image and training negative sample image input convolutional neural networks model and obtains target detection model；
Aircraft Targets image to be detected is input in target detection model by picture recognition module, and target detection model flies to be detected
Machine target image is detected to obtain Aircraft Targets testing result.
In the present embodiment, further include the second pre-detection module, the second pre-detection module is for extracting aircraft mesh to be detected
The significant characteristics of logo image obtain the Saliency maps picture of several different scales, using non-maxima suppression algorithm to several notable
Property image carry out image co-registration.It is specially to be carried out to original positive sample image to carry out data enhancing processing to original positive sample image
Translation processing adds and makes an uproar processing, contrast enhancement processing and angle rotation processing and obtain trained positive sample image, data enhancing
Module includes translation unit plus make an uproar unit, contrast enhancement unit and angle rotary unit, and translation unit is used for training just
Sample image carries out translation processing, adds unit of making an uproar for carrying out adding processing of making an uproar, contrast enhancement unit to training positive sample image
For carrying out contrast enhancement processing to training positive sample image, angle rotary unit is used to carry out angle to training positive sample image
Spend rotation processing.
A kind of Aircraft Targets detection method of the present invention, by obtaining the instruction containing aircraft brake disc in preset data set
Practice image and obtains Aircraft Targets image to be detected in test side；The training image is carried out at the first conspicuousness pre-detection
Sample image is managed and obtains, carrying out classification to the sample image handles to obtain original positive sample image and training negative sample figure
Picture；Data enhancing is carried out to the original positive sample image to handle to obtain trained positive sample image；By the trained positive sample figure
It is trained in picture and the trained negative sample image input convolutional neural networks model and obtains target detection model；It will be described
Aircraft Targets image to be detected is input in the target detection model, and the target detection model is to the aircraft mesh to be detected
Logo image is detected to obtain Aircraft Targets testing result；It is pre- by first being carried out to the training image that data are concentrated in detection process
Detection and data enhancing processing, then training image is input to convolutional network model and is trained to obtain by depth
The convolutional network model of habit is detected Aircraft Targets image to be detected, therefore the testing result finally obtained is more accurate,
The precise positioning for aircraft is realized, a variety of different application scenarios can be used by a little target detection models of depth
Aircraft Targets image to be detected；Military and other relevant departments are allow to monitor the dynamic of each military key area in real time.
More than, only presently preferred embodiments of the present invention is not intended to limit the present invention in any form；All one's own professions
The those of ordinary skill of industry can be shown in by specification attached drawing and above and swimmingly implement the present invention；But all to be familiar with sheet special
The technical staff of industry without departing from the scope of the present invention, is made a little using disclosed above technology contents
The equivalent variations of variation, modification and evolution are the equivalent embodiment of the present invention；Meanwhile all substantial technologicals according to the present invention
To the variation, modification and evolution etc. of any equivalent variations made by above example, technical scheme of the present invention is still fallen within
Within protection domain.
Claims (10)
1. a kind of Aircraft Targets detection method, it is characterised in that including:
Image obtains, and the training image containing aircraft brake disc is obtained in preset data set and is obtained in test side to be detected
Aircraft Targets image；
First pre-detection carries out the first conspicuousness pre-detection to the training image and handles and obtain sample image, to the sample
This image carries out classification and handles to obtain original positive sample image and training negative sample image；
Data enhancing is handled, and carrying out data enhancing to the original positive sample image handles to obtain trained positive sample image；
Target detection model is built, the trained positive sample image and the trained negative sample image are inputted into convolutional neural networks
It is trained in model and obtains target detection model；
The Aircraft Targets image to be detected is input in the target detection model, the target detection mould by image recognition
Type is detected to obtain Aircraft Targets testing result to the Aircraft Targets image to be detected.
2. a kind of Aircraft Targets detection method as described in claim 1, it is characterised in that：First pre-detection is specially：
The significant characteristics for extracting the trained atlas obtain the Saliency maps picture of several different scales, using non-maxima suppression algorithm
Image co-registration is carried out to several Saliency maps pictures and obtains including the sample image of scene around Aircraft Targets and aircraft, it is right
The sample image carries out classification processing, obtains the original positive sample image containing Aircraft Targets and containing scene around aircraft
Training negative sample image.
3. a kind of Aircraft Targets detection method as described in claim 1, it is characterised in that：The data enhancing is handled
To the original positive sample image carry out translation processing, plus make an uproar processing, contrast enhancement processing and angle rotation processing and
To training positive sample image.
4. a kind of Aircraft Targets detection method as described in claim 1, it is characterised in that：Further include before described image identification
Second pre-detection processing is carried out to the Aircraft Targets image to be detected.
5. a kind of Aircraft Targets detection method as claimed in claim 4, it is characterised in that：Second pre-detection is specially to carry
The significant characteristics of the Aircraft Targets image to be detected are taken to obtain the Saliency maps picture of several different scales, using non-maximum
Restrainable algorithms carry out image co-registration to several Saliency maps pictures.
6. a kind of electronic equipment, it is characterised in that including：Processor；
Memory；And program, wherein described program is stored in the memory, and is configured to be held by processor
Row, described program include the method required for perform claim described in 1-5 any one.
7. a kind of computer readable storage medium, is stored thereon with computer program, it is characterised in that：The computer program quilt
Processor executes the method as described in claim 1-5 any one.
8. a kind of Aircraft Targets detecting system, it is characterised in that including：
Image collection module, described image acquisition module in preset data set for obtaining the training figure containing aircraft brake disc
As and test side obtains Aircraft Targets image to be detected；
First pre-detection module, the first pre-detection module are used to carry out at the first conspicuousness pre-detection the training image
Sample image is managed and obtains, carrying out classification to the sample image handles to obtain original positive sample image and training negative sample figure
Picture；
Data enhance module, and the data enhancing module is used to carry out data enhancing to the original positive sample image to handle to obtain
Training positive sample image；
Target detection model module is built, the structure target detection model module is used for the trained positive sample image and institute
It states and is trained in trained negative sample image input convolutional neural networks model and obtains target detection model；
The Aircraft Targets image to be detected is input in the target detection model by picture recognition module, the target inspection
It surveys model the Aircraft Targets image to be detected is detected to obtain Aircraft Targets testing result.
9. such as a kind of Aircraft Targets detecting system of claim 8, it is characterised in that：Further include the second pre-detection module, described
The significant characteristics that two pre-detection modules are used to extract the Aircraft Targets image to be detected obtain the notable of several different scales
Property image, image co-registrations are carried out to several Saliency maps pictures using non-maxima suppression algorithm.
10. such as a kind of Aircraft Targets detecting system of claim 8, it is characterised in that：It is described to the original positive sample image into
The enhancing processing of row data is specially to carry out translation processing to the original positive sample image plus make an uproar processing, contrast enhancement processing
And angle rotation processing and obtaining trains positive sample image, the data enhancing module includes translation unit plus unit of making an uproar, right
Than degree enhancement unit and angle rotary unit, the translation unit is for carrying out at translation the trained positive sample image
Reason, for the trained positive sample image to be carried out plus made an uproar processing, the contrast enhancement unit is used for pair described plus unit of making an uproar
The trained positive sample image carries out contrast enhancement processing, and the angle rotary unit is used for the trained positive sample image
Carry out angle rotation processing.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810474112.9A CN108764082A (en) | 2018-05-17 | 2018-05-17 | A kind of Aircraft Targets detection method, electronic equipment, storage medium and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810474112.9A CN108764082A (en) | 2018-05-17 | 2018-05-17 | A kind of Aircraft Targets detection method, electronic equipment, storage medium and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108764082A true CN108764082A (en) | 2018-11-06 |
Family
ID=64006896
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810474112.9A Withdrawn CN108764082A (en) | 2018-05-17 | 2018-05-17 | A kind of Aircraft Targets detection method, electronic equipment, storage medium and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108764082A (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110210513A (en) * | 2019-04-23 | 2019-09-06 | 深圳信息职业技术学院 | Data classification method, device and terminal device |
| CN110245711A (en) * | 2019-06-18 | 2019-09-17 | 西安电子科技大学 | The SAR target identification method for generating network is rotated based on angle |
| CN110569721A (en) * | 2019-08-01 | 2019-12-13 | 平安科技（深圳）有限公司 | Recognition model training method, image recognition method, device, equipment and medium |
| CN110991531A (en) * | 2019-12-02 | 2020-04-10 | 中电科特种飞机系统工程有限公司 | Training sample library construction method, device and medium based on air-to-ground small and slow target |
| CN111160123A (en) * | 2019-12-11 | 2020-05-15 | 桂林长海发展有限责任公司 | Airplane target identification method and device and storage medium |
| CN112084945A (en) * | 2020-09-09 | 2020-12-15 | 深圳市赛为智能股份有限公司 | Active bird repelling method and device, computer equipment and storage medium |
| CN113159209A (en) * | 2021-04-29 | 2021-07-23 | 深圳市商汤科技有限公司 | Target detection method, device, equipment and computer readable storage medium |
| CN113283409A (en) * | 2021-07-23 | 2021-08-20 | 中国人民解放军国防科技大学 | Airplane detection method in aerial image based on EfficientDet and Transformer |
| CN113673541A (en) * | 2021-10-21 | 2021-11-19 | 广州微林软件有限公司 | Image sample generation method for target detection and application |
| CN113705489A (en) * | 2021-08-31 | 2021-11-26 | 中国电子科技集团公司第二十八研究所 | Remote sensing image fine-grained airplane identification method based on priori regional knowledge guidance |
| CN113705489B (en) * | 2021-08-31 | 2024-06-07 | 中国电子科技集团公司第二十八研究所 | Remote sensing image fine-granularity airplane identification method based on priori regional knowledge guidance |
-
2018
- 2018-05-17 CN CN201810474112.9A patent/CN108764082A/en not_active Withdrawn
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110210513A (en) * | 2019-04-23 | 2019-09-06 | 深圳信息职业技术学院 | Data classification method, device and terminal device |
| CN110245711B (en) * | 2019-06-18 | 2022-12-02 | 西安电子科技大学 | SAR target identification method based on angle rotation generation network |
| CN110245711A (en) * | 2019-06-18 | 2019-09-17 | 西安电子科技大学 | The SAR target identification method for generating network is rotated based on angle |
| CN110569721A (en) * | 2019-08-01 | 2019-12-13 | 平安科技（深圳）有限公司 | Recognition model training method, image recognition method, device, equipment and medium |
| CN110569721B (en) * | 2019-08-01 | 2023-08-29 | 平安科技（深圳）有限公司 | Recognition model training method, image recognition method, device, equipment and medium |
| CN110991531A (en) * | 2019-12-02 | 2020-04-10 | 中电科特种飞机系统工程有限公司 | Training sample library construction method, device and medium based on air-to-ground small and slow target |
| CN111160123A (en) * | 2019-12-11 | 2020-05-15 | 桂林长海发展有限责任公司 | Airplane target identification method and device and storage medium |
| CN111160123B (en) * | 2019-12-11 | 2023-06-09 | 桂林长海发展有限责任公司 | Aircraft target identification method, device and storage medium |
| CN112084945A (en) * | 2020-09-09 | 2020-12-15 | 深圳市赛为智能股份有限公司 | Active bird repelling method and device, computer equipment and storage medium |
| CN113159209A (en) * | 2021-04-29 | 2021-07-23 | 深圳市商汤科技有限公司 | Target detection method, device, equipment and computer readable storage medium |
| CN113159209B (en) * | 2021-04-29 | 2024-05-24 | 深圳市商汤科技有限公司 | Object detection method, device, equipment and computer readable storage medium |
| CN113283409A (en) * | 2021-07-23 | 2021-08-20 | 中国人民解放军国防科技大学 | Airplane detection method in aerial image based on EfficientDet and Transformer |
| CN113705489A (en) * | 2021-08-31 | 2021-11-26 | 中国电子科技集团公司第二十八研究所 | Remote sensing image fine-grained airplane identification method based on priori regional knowledge guidance |
| CN113705489B (en) * | 2021-08-31 | 2024-06-07 | 中国电子科技集团公司第二十八研究所 | Remote sensing image fine-granularity airplane identification method based on priori regional knowledge guidance |
| CN113673541A (en) * | 2021-10-21 | 2021-11-19 | 广州微林软件有限公司 | Image sample generation method for target detection and application |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108764082A (en) | A kind of Aircraft Targets detection method, electronic equipment, storage medium and system | |
| CN108510467B (en) | SAR image target identification method based on depth deformable convolution neural network | |
| CN111145177B (en) | Image sample generation method, specific scene target detection method and system thereof | |
| CN109583369B (en) | Target identification method and device based on target area segmentation network | |
| CN110310264A (en) | A kind of large scale object detection method, device based on DCNN | |
| CN107609483B (en) | Dangerous target detection method and device for driving assistance system | |
| CN104408482B (en) | A kind of High Resolution SAR Images object detection method | |
| CN104463249B (en) | A kind of remote sensing images airfield detection method based on Weakly supervised learning framework | |
| CN109284704A (en) | Complex background SAR vehicle target detection method based on CNN | |
| CN106874889A (en) | Multiple features fusion SAR target discrimination methods based on convolutional neural networks | |
| CN105574550A (en) | Vehicle identification method and device | |
| Qu et al. | A pedestrian detection method based on yolov3 model and image enhanced by retinex | |
| CN104881865A (en) | Forest disease and pest monitoring and early warning method and system based on unmanned plane image analysis | |
| CN103761526B (en) | Urban area detecting method based on feature position optimization and integration | |
| CN110008900B (en) | Method for extracting candidate target from visible light remote sensing image from region to target | |
| CN113989662A (en) | Remote sensing image fine-grained target identification method based on self-supervision mechanism | |
| Nayak et al. | Machine learning techniques for AUV side-scan sonar data feature extraction as applied to intelligent search for underwater archaeological sites | |
| CN110910341A (en) | Transmission line corrosion area defect detection method and device | |
| CN108257179B (en) | Image processing method | |
| Yang et al. | Toward country scale building detection with convolutional neural network using aerial images | |
| CN116157867A (en) | Neural network analysis of LFA test strips | |
| Wang et al. | A camouflaged object detection model based on deep learning | |
| CN111881984A (en) | Target detection method and device based on deep learning | |
| Mejias et al. | Automated marine mammal detection from aerial imagery | |
| CN111553184A (en) | Small target detection method and device based on electronic purse net and electronic equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20181106 |
|
| WW01 | Invention patent application withdrawn after publication |