CN115953704A - Unmanned aerial vehicle detection method - Google Patents
Unmanned aerial vehicle detection method Download PDFInfo
- Publication number
- CN115953704A CN115953704A CN202310063092.7A CN202310063092A CN115953704A CN 115953704 A CN115953704 A CN 115953704A CN 202310063092 A CN202310063092 A CN 202310063092A CN 115953704 A CN115953704 A CN 115953704A
- Authority
- CN
- China
- Prior art keywords
- data
- flying object
- video data
- carrying
- detection
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses an unmanned aerial vehicle detection method, which comprises the following steps: the method comprises the steps of obtaining video data, carrying out gray value analysis and connectivity analysis on the video data to obtain image data containing a target, carrying out pixel point analysis and judgment on the image data, obtaining video data containing a flying object according to an analysis and judgment result, carrying out distance judgment on an unknown flying object in the video data containing the flying object to obtain a distance judgment result, carrying out distance detection on the video data containing the flying object based on the distance judgment result, wherein the distance detection comprises short-distance detection and long-distance detection to obtain flying object track data, and carrying out real-time recording and/or early warning on the flying object based on the flying object track data. The method and the device can observe the unknown flyer and perform real-time recording and early warning.
Description
Technical Field
The invention belongs to the field of anti-unmanned aerial vehicles, and particularly relates to an unmanned aerial vehicle detection method.
Background
There are many detection systems for unmanned aerial vehicles, including radar, acoustic sensing, passive thermal imaging technology, passive optical imaging technology, radio frequency technology, etc., however, unmanned aerial vehicle detection has a difficult problem of "low speed and small noise". The low-speed small-impurity unmanned aerial vehicle mainly has the characteristics of low flying height, low flying speed, small reflection sectional area and multiple types of unmanned aerial vehicles, and the characteristics of low speed and small impurity provide difficult problems for real-time detection, tracking and detailed classification of targets of an unmanned aerial vehicle detection system.
Disclosure of Invention
The invention aims to provide an unmanned aerial vehicle detection method to solve the problems in the prior art.
In order to achieve the purpose, the invention provides an unmanned aerial vehicle detection method, which comprises the following steps:
the method comprises the steps of obtaining video data, carrying out gray value analysis and connectivity analysis on the video data to obtain image data containing a target, carrying out pixel point analysis and judgment on the image data, obtaining video data containing a flying object according to an analysis and judgment result, carrying out distance judgment on an unknown flying object in the video data containing the flying object to obtain a distance judgment result, carrying out distance detection on the video data containing the flying object based on the distance judgment result, wherein the distance detection comprises short-distance detection and long-distance detection to obtain flying object track data, and carrying out real-time recording and/or early warning on the flying object based on the flying object track data.
Optionally, the process of acquiring image data including an object includes:
and performing gray value analysis on the video data to obtain difference image data, acquiring a gray threshold, and performing connectivity analysis on the difference image data based on the gray threshold to obtain image data containing a target.
Optionally, the process of analyzing and judging the pixel points of the image data includes:
and acquiring a pixel threshold, and performing pixel point analysis and judgment on the image data based on the pixel threshold, wherein when the judgment result is that the pixel point data is greater than the pixel threshold, video data containing a flying object is obtained, otherwise, the video data is considered not to contain the flying object, and gray value analysis and connectivity analysis are performed on the video data at the next moment.
Optionally, the process of obtaining the distance determination result includes:
analyzing the flying objects in the video data containing the flying objects to obtain proportion data of the flying objects in the video images, obtaining proportion threshold values, and comparing the proportion data with the proportion threshold values to obtain distance judgment results.
Optionally, when the proportion data is smaller than the proportion threshold, performing long-distance detection on the video data containing the flying object;
and when the proportion data is larger than or equal to the proportion threshold value, performing close-range detection on the video data containing the flying object.
Optionally, the process of remotely detecting the video data containing the flying object includes:
preprocessing video data containing the flyer to obtain similarity measurement data, and obtaining unknown flyer trajectory clustering data based on the similarity measurement data;
constructing a hidden Markov model and acquiring clustering data of different flyer tracks, and training the hidden Markov model through a Baum-Welch algorithm based on the clustering data of the different flyer tracks to obtain a track detection model and different flyer track models;
processing the unknown flyer trajectory clustering data through a trajectory detection model to obtain an observation sequence, performing probability calculation based on the observation sequence and different flyer models to obtain conditional probability data of the flyers, and analyzing and calculating the conditional probability data through a Viterbi algorithm and a maximum likelihood formula to obtain the flyer trajectory data.
Optionally, the process of performing close-range detection on the video data containing the flying object includes:
preprocessing video data containing the flyer to obtain preprocessed data, and analyzing and calculating the preprocessed data through a single-stage target detection algorithm to obtain primary screening data of the flyer, wherein the primary screening data comprises large-variety information data, apparent characteristic data and position information data of the flyer;
and carrying out category judgment on the primary screening data, carrying out small sample target refinement processing on the large category information data based on a category judgment result to obtain small category data, carrying out small category judgment on the small category data, carrying out target tracking on the flying object according to the apparent characteristic data and the position information data based on a small category judgment result, and uploading the trajectory data of the flying object.
Optionally, the small sample target refinement processing is performed on the large-category information data through a deep neural network.
Optionally, the target tracking is performed on the flyer through a kalman filter, a hungarian algorithm and a cascade matching algorithm.
The invention has the technical effects that:
according to the unmanned aerial vehicle detection method under the condition of the small sample, the unmanned aerial vehicle is analyzed through long-distance observation and short-distance observation, the invasion unmanned aerial vehicle is accurately detected and tracked, and the 'black flight problem' is solved, so that the safety of important facilities, the national defense safety and the privacy of citizens are guaranteed not to be invaded, and meanwhile, the unmanned aerial vehicle detection method has the advantages of high robustness, strong mobility, good accuracy and the like.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this application, illustrate embodiments of the application and, together with the description, serve to explain the application and are not intended to limit the application. In the drawings:
FIG. 1 is a flow chart of a video pre-processing process according to an embodiment of the present invention;
FIG. 2 is a flowchart of a method according to an embodiment of the present invention.
Detailed Description
It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict. The present application will be described in detail below with reference to the embodiments with reference to the attached drawings.
Example one
As shown in fig. 1-2, the present embodiment provides a method for detecting an unmanned aerial vehicle, including:
the method comprises the steps of obtaining video data, carrying out gray value analysis and connectivity analysis on the video data to obtain image data containing a target, carrying out pixel point analysis and judgment on the image data, obtaining video data containing a flying object according to an analysis and judgment result, carrying out distance judgment on an unknown flying object in the video data containing the flying object to obtain a distance judgment result, carrying out distance detection on the video data containing the flying object based on the distance judgment result, wherein the distance detection comprises short-distance detection and long-distance detection to obtain flying object track data, and carrying out real-time recording and/or early warning on the flying object based on the flying object track data.
At present, the principle, advantages and disadvantages and detection distance of the detection method aiming at the low speed and small impurities of the unmanned aerial vehicle are shown in the table, wherein the table 1 shows the advantages and disadvantages in the prior detection technology;
TABLE 1
According to the unmanned aerial vehicle detection method under the condition of the small sample, the unmanned aerial vehicle is analyzed through long-distance observation and short-distance observation, the invasion unmanned aerial vehicle is accurately detected and tracked, and the 'black flight problem' is solved, so that the safety of important facilities, the national defense safety and the privacy of citizens are guaranteed not to be invaded, and meanwhile, the unmanned aerial vehicle detection method has the advantages of being high in robustness, strong in mobility, good in accuracy and the like.
Step one, a high-definition camera on the unmanned aerial vehicle detection system is responsible for 24h uninterrupted video recording, and the recorded video is transmitted into a moving object detection module of the unmanned aerial vehicle detection system.
Step two, a motion detection module of the unmanned aerial vehicle detection system reads a background image frame B of a frame B (a background image frame B obtained by performing Gaussian mixture filtering operation on an image of a previous n frames of a current background) of the acquired video, and records a video image of the current frame as f n The gray values of the corresponding pixels of the background frame and the current frame are respectively marked as B (x, y) and f n (x, y) gray of pixel points corresponding to two frames of imagesThe values are subtracted, and the absolute value is taken to obtain a difference image D n ：
D n (x,y)＝|f n (x,y)-B(x,y)|
Setting a threshold T 1 Carrying out binarization processing on the pixel points one by one to obtain a binarization image R n . Wherein, the point with the gray value of 255 is the foreground (moving object) point, and the point with the gray value of 0 is the background point; to image R' n And performing connectivity analysis to finally obtain an image Rn containing the complete moving target.
And calculating pixel points of the image Rn containing the complete moving target, if the number of the pixel points is less than n, repeating the step two, if the number of the pixel points is more than n, determining that the section of the collected video contains the video of the unknown flying object, immediately entering a real-time recording mode by the system, and transmitting the image to the distance detection module. As shown in fig. 1.
And step three, the distance detector compares the occupancy of the unknown flying object in the video with a threshold value T, if the occupancy is higher than or equal to the threshold value, the distance detector processes the occupancy, and if the occupancy is lower than the threshold value, the distance detector processes the occupancy. Namely, the step two and the step three are respectively carried out.
Step four, when the distance is far, the collected video V is processed by a long-distance detector, and a preprocessing part P 1 The method comprises the steps of Fourier transform, data noise filtering, edge detection, K-means clustering algorithm and the like;
pretreatment of P 1 Firstly, data noise filtering is carried out on the part, clutter interference in a video is filtered, then the collected video resources are converted into space domain map and frequency domain map information of a track by utilizing Fourier transform on the remaining part, hausdorff distance is extracted from the space domain map and the frequency domain map, the Hausdorff distance is set as similarity measurement between samples, and the track is clustered by utilizing a K mean algorithm;
trajectory detectionModule M 1 The subject structure of (1) is a hidden Markov model, M 1 Composed of 4-6 state numbers which do not jump from left to right, and the state transition probability P of each state number n Is between 0 and 1. In the training process, various flyer tracks clustered by K means based on Hausdorff distance are used as output, the Baum-Welch algorithm is utilized to continuously obtain the estimated value of the model, so that the convergence of the model is ensured, and the trained model is obtained
And different flight object models λ = { λ = { [ λ ] 0 ,λ 1 ,λ 2 …λ n In which λ is 0 For a model of the trajectory of the unmanned aerial vehicle, λ 1 -λ n Inputting unknown flier tracks which are subjected to K-means clustering based on Hausdorff distance into a model->
Obtaining an observed sequence of the test sample O = O 1 ,O 2 …O i Calculating conditional probability of the test sample relative to the model of the flying object
P(O|λ i )＝P(O,Q * |λ i )
Wherein Q * In order to test the optimal state sequence in the sample observation value sequence, a Viterbi algorithm is utilized, and a track model with the maximum conditional probability is used as the track sequence of the unknown flyer according to a maximum likelihood formula.
Step five, when the distance is close, the collected video V is processed by a close distance detector, and noise points existing in some videos are removed through a preprocessing part of a preprocessing module by using the technologies of image enhancement, image filtering, edge detection and the like to obtain the video V 2 The real-time flyer primary screening target detection module is formed by a single-stage target detection algorithm and can read non-screened targets from the acquired video in real timeThe large category information, the apparent characteristic and the position information of the flying object are known, and the large category Y of the unknown flying object is output 2 And if the unknown flying object is not the unmanned aerial vehicle type, exiting the real-time recording mode and entering the early warning mode, and if the unknown flying object is the unmanned aerial vehicle type, entering the step six.
Step six, delivering the unmanned aerial vehicle picture frame extracted by the real-time flyer prescreening target detection module to a small sample target refining module, wherein the small sample target refining module of the unmanned aerial vehicle is mainly composed of a deep learning prototype network M 2 The composition comprises an input layer, C hidden layers, a prototype representation layer and an output layer, wherein the deep neural network M 2 The input of the data is model pictures of different unmanned aerial vehicles under the condition of small samples, and the deep neural network M 2 The output of the method is small type of the flying object, and a deep neural network is obtained after training
Trained deep learning prototype network &>
Outputting the subclass of the unmanned aerial vehicle, entering a seventh step if the unmanned aerial vehicle is a dangerous unmanned aerial vehicle, exiting the real-time recording mode and entering an early warning mode if the unmanned aerial vehicle is not a dangerous unmanned aerial vehicle (step one).
And step seven, delivering the apparent information and the position information of the unmanned aerial vehicle extracted by the real-time flyer prescreening target detection module to a target tracking module, wherein the target tracking module is composed of a Kalman filter, a Hungarian algorithm and a cascade matching algorithm, the Kalman filter is mainly used for generating a motion prediction frame at the next moment based on the motion variable at the previous moment, the Hungarian algorithm is used for matching the frame of the Kalman filter with the detection frame to realize the real-time tracking effect, the cascade matching algorithm is combined with apparent characteristics, the flyer can still track the flyer in real time after receiving shielding and losing signals, the track information of the dangerous unmanned aerial vehicle is recorded and uploaded in real time, and the unmanned aerial vehicle exits from the real-time recording mode and enters an early warning mode (step one) after the dangerous unmanned aerial vehicle disappears.
The above description is only for the preferred embodiment of the present application, but the scope of the present application is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present application should be covered within the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
Claims (9)
1. An unmanned aerial vehicle detection method is characterized by comprising the following steps:
the method comprises the steps of obtaining video data, carrying out gray value analysis and connectivity analysis on the video data to obtain image data containing a target, carrying out pixel point analysis and judgment on the image data, obtaining video data containing a flying object according to an analysis and judgment result, carrying out distance judgment on an unknown flying object in the video data containing the flying object to obtain a distance judgment result, carrying out distance detection on the video data containing the flying object based on the distance judgment result, wherein the distance detection comprises short-distance detection and long-distance detection to obtain flying object track data, and carrying out real-time recording and/or early warning on the flying object based on the flying object track data.
2. A method for unmanned aerial vehicle detection as defined in claim 1,
the process of acquiring image data containing an object includes:
and performing gray value analysis on the video data to obtain difference image data, acquiring a gray threshold, and performing connectivity analysis on the difference image data based on the gray threshold to obtain image data containing a target.
3. A method for unmanned aerial vehicle detection as claimed in claim 1,
the process of analyzing and judging the pixel points of the image data comprises the following steps:
and acquiring a pixel threshold, and performing pixel point analysis and judgment on the image data based on the pixel threshold, wherein when the judgment result is that the pixel point data is greater than the pixel threshold, video data containing a flying object is obtained, otherwise, the video data is considered not to contain the flying object, and gray value analysis and connectivity analysis are performed on the video data at the next moment.
4. A method for unmanned aerial vehicle detection as defined in claim 1,
the process of obtaining the distance judgment result comprises the following steps:
analyzing the flying objects in the video data containing the flying objects to obtain proportion data of the flying objects in the video images, acquiring proportion threshold values, and comparing the proportion data with the proportion threshold values to obtain distance judgment results.
5. A method for unmanned aerial vehicle detection as defined in claim 4,
when the proportion data is smaller than the proportion threshold value, carrying out remote detection on the video data containing the flyer;
and when the proportion data is larger than or equal to the proportion threshold value, performing close-range detection on the video data containing the flying object.
6. A method for unmanned aerial vehicle detection as defined in claim 5,
the process of remotely detecting video data containing a flying object comprises the following steps:
preprocessing video data containing the flyer to obtain similarity measurement data, and obtaining unknown flyer trajectory clustering data based on the similarity measurement data;
constructing a hidden Markov model and acquiring clustering data of different flyer tracks, and training the hidden Markov model through a Baum-Welch algorithm based on the clustering data of the different flyer tracks to obtain a track detection model and different flyer track models;
processing the unknown flyer trajectory clustering data through a trajectory detection model to obtain an observation sequence, performing probability calculation based on the observation sequence and different flyer models to obtain conditional probability data of the flyers, and analyzing and calculating the conditional probability data through a Viterbi algorithm and a maximum likelihood formula to obtain the flyer trajectory data.
7. A method for unmanned aerial vehicle detection as claimed in claim 5,
the process of performing close-range detection on video data containing a flying object comprises the following steps:
preprocessing video data containing the flyer to obtain preprocessed data, and analyzing and calculating the preprocessed data through a single-stage target detection algorithm to obtain primary screening data of the flyer, wherein the primary screening data comprises large-variety information data, apparent characteristic data and position information data of the flyer;
and carrying out category judgment on the primary screening data, carrying out small sample target refinement processing on the large category information data based on a category judgment result to obtain small category data, carrying out small category judgment on the small category data, carrying out target tracking on the flying object according to the apparent characteristic data and the position information data based on a small category judgment result, and uploading the trajectory data of the flying object.
8. A drone detecting method according to claim 7,
and carrying out small sample target refinement processing on the large-class information data through a deep neural network.
9. The unmanned aerial vehicle detection method of claim 7, wherein the flyers are subjected to target tracking through a Kalman filter, a Hungarian algorithm and a cascade matching algorithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310063092.7A CN115953704B (en) | 2023-01-18 | 2023-01-18 | Unmanned aerial vehicle detection method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310063092.7A CN115953704B (en) | 2023-01-18 | 2023-01-18 | Unmanned aerial vehicle detection method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115953704A true CN115953704A (en) | 2023-04-11 |
| CN115953704B CN115953704B (en) | 2023-10-03 |
Family
ID=87282578
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310063092.7A Active CN115953704B (en) | 2023-01-18 | 2023-01-18 | Unmanned aerial vehicle detection method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115953704B (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108957445A (en) * | 2018-07-30 | 2018-12-07 | 四川九洲空管科技有限责任公司 | A kind of low-altitude low-velocity small targets detection system and its detection method |
| WO2019006769A1 (en) * | 2017-07-06 | 2019-01-10 | 杨顺伟 | Following-photographing method and device for unmanned aerial vehicle |
| KR20200140723A (en) * | 2019-06-07 | 2020-12-16 | 한국전자통신연구원 | Apparatus and method for tracking unmanned aerial vehicle |
| CN112698665A (en) * | 2020-12-28 | 2021-04-23 | 同济大学 | Unmanned aerial vehicle detection positioning method |
| CN112799426A (en) * | 2020-12-25 | 2021-05-14 | 陈南方 | Unmanned aerial vehicle navigation control system and method based on big data analysis |
| KR20220019377A (en) * | 2020-08-10 | 2022-02-17 | 세종대학교산학협력단 | Drone detection apparatus using deep learning and method thereof |
| CN114757974A (en) * | 2022-04-24 | 2022-07-15 | 汕头大学 | Trajectory tracking method and system for multi-rotor unmanned aerial vehicle |
| CN115311580A (en) * | 2022-08-02 | 2022-11-08 | 深圳市栢迪科技有限公司 | Unmanned aerial vehicle threat determination method based on image recognition and related equipment |
| CN115424072A (en) * | 2022-09-06 | 2022-12-02 | 鄂尔多斯市视达科技有限公司 | Unmanned aerial vehicle defense method based on detection technology |
-
2023
- 2023-01-18 CN CN202310063092.7A patent/CN115953704B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019006769A1 (en) * | 2017-07-06 | 2019-01-10 | 杨顺伟 | Following-photographing method and device for unmanned aerial vehicle |
| CN108957445A (en) * | 2018-07-30 | 2018-12-07 | 四川九洲空管科技有限责任公司 | A kind of low-altitude low-velocity small targets detection system and its detection method |
| KR20200140723A (en) * | 2019-06-07 | 2020-12-16 | 한국전자통신연구원 | Apparatus and method for tracking unmanned aerial vehicle |
| KR20220019377A (en) * | 2020-08-10 | 2022-02-17 | 세종대학교산학협력단 | Drone detection apparatus using deep learning and method thereof |
| CN112799426A (en) * | 2020-12-25 | 2021-05-14 | 陈南方 | Unmanned aerial vehicle navigation control system and method based on big data analysis |
| CN112698665A (en) * | 2020-12-28 | 2021-04-23 | 同济大学 | Unmanned aerial vehicle detection positioning method |
| CN114757974A (en) * | 2022-04-24 | 2022-07-15 | 汕头大学 | Trajectory tracking method and system for multi-rotor unmanned aerial vehicle |
| CN115311580A (en) * | 2022-08-02 | 2022-11-08 | 深圳市栢迪科技有限公司 | Unmanned aerial vehicle threat determination method based on image recognition and related equipment |
| CN115424072A (en) * | 2022-09-06 | 2022-12-02 | 鄂尔多斯市视达科技有限公司 | Unmanned aerial vehicle defense method based on detection technology |
Non-Patent Citations (1)
| Title |
|---|
| 张寅;蔡旭阳;闫钧华;苏恺;张琨;: "联合时空信息和轨迹关联的空中多目标检测", 武汉大学学报(信息科学版), no. 10 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115953704B (en) | 2023-10-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107862705B (en) | Unmanned aerial vehicle small target detection method based on motion characteristics and deep learning characteristics | |
| Ao et al. | Needles in a haystack: Tracking city-scale moving vehicles from continuously moving satellite | |
| CN109389609B (en) | Interactive self-feedback infrared target detection method based on FART neural network | |
| Zhou et al. | Detecting and tracking small moving objects in wide area motion imagery (wami) using convolutional neural networks (cnns) | |
| CN112818905B (en) | Finite pixel vehicle target detection method based on attention and spatio-temporal information | |
| CN111179318B (en) | Double-flow method-based complex background motion small target detection method | |
| Cruz et al. | Learning temporal features for detection on maritime airborne video sequences using convolutional LSTM | |
| Chen et al. | Advanced automatic target recognition (ATR) with infrared (IR) sensors | |
| CN116109950A (en) | Low-airspace anti-unmanned aerial vehicle visual detection, identification and tracking method | |
| Lin et al. | Deep learning based classification of radar spectral maps | |
| Khoshboresh-Masouleh et al. | Robust building footprint extraction from big multi-sensor data using deep competition network | |
| CN115953704B (en) | Unmanned aerial vehicle detection method | |
| CN111311640A (en) | Unmanned aerial vehicle identification and tracking method based on motion estimation | |
| CN109558771B (en) | Behavior state identification method, device and equipment of marine ship and storage medium | |
| Ao et al. | Detecting tiny moving vehicles in satellite videos | |
| Kaliyev et al. | Computer Vision-based Fire Detection using Enhanced Chromatic Segmentation and Optical Flow Model | |
| Brosch et al. | Automatic target recognition on high resolution sar images with deep learning domain adaptation | |
| CN109859200B (en) | Low-altitude slow-speed unmanned aerial vehicle rapid detection method based on background analysis | |
| CN113223081A (en) | High-altitude parabolic detection method and system based on background modeling and deep learning | |
| CN114663916A (en) | Thermal infrared human body target identification method based on depth abstract features | |
| CN114494359A (en) | Small sample moving object detection method based on abnormal optical flow | |
| CN113963178A (en) | Method, device, equipment and medium for detecting infrared dim and small target under ground-air background | |
| Bhagat et al. | Moving camera-based automated system for drone identification using focus measures | |
| Zhao et al. | Research on vehicle detection and vehicle type recognition under cloud computer vision | |
| Yin et al. | Dynamic Vehicle Detection in Satellite Video with Multi-frame Brightness Gradient |
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 |