CN116343045B - Lightweight SAR image ship target detection method based on YOLO v5 - Google Patents
Lightweight SAR image ship target detection method based on YOLO v5 Download PDFInfo
- Publication number
- CN116343045B CN116343045B CN202310342382.5A CN202310342382A CN116343045B CN 116343045 B CN116343045 B CN 116343045B CN 202310342382 A CN202310342382 A CN 202310342382A CN 116343045 B CN116343045 B CN 116343045B
- Authority
- CN
- China
- Prior art keywords
- yolo
- convolution
- lightweight
- model
- data set
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 73
- 238000012549 training Methods 0.000 claims abstract description 38
- 238000004088 simulation Methods 0.000 claims abstract description 18
- 238000012360 testing method Methods 0.000 claims abstract description 18
- 238000003384 imaging method Methods 0.000 claims abstract description 8
- 238000007781 pre-processing Methods 0.000 claims abstract description 5
- 230000006870 function Effects 0.000 claims description 25
- 230000004913 activation Effects 0.000 claims description 24
- 238000012512 characterization method Methods 0.000 claims description 23
- 238000000034 method Methods 0.000 claims description 20
- 230000007246 mechanism Effects 0.000 claims description 6
- 238000004590 computer program Methods 0.000 claims description 4
- 230000004927 fusion Effects 0.000 claims description 4
- 238000004422 calculation algorithm Methods 0.000 description 9
- 238000000605 extraction Methods 0.000 description 6
- 230000008901 benefit Effects 0.000 description 4
- HPTJABJPZMULFH-UHFFFAOYSA-N 12-[(Cyclohexylcarbamoyl)amino]dodecanoic acid Chemical compound OC(=O)CCCCCCCCCCCNC(=O)NC1CCCCC1 HPTJABJPZMULFH-UHFFFAOYSA-N 0.000 description 2
- 230000001133 acceleration Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 230000001427 coherent effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000007500 overflow downdraw method Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000009467 reduction Effects 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/07—Target detection
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Evolutionary Computation (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a lightweight SAR image ship target detection method based on YOLO v5, which comprises the following steps: acquiring an SAR image dataset, obtaining an SAR image ship target simulation dataset and a public SAR ship detection dataset through simulation imaging, preprocessing the dataset, and dividing the dataset into a training sample set and a test sample set; establishing an improved lightweight YOLO v5 model; inputting the training data set into an improved lightweight YOLO v5 model for training, and obtaining a trained improved lightweight YOLO v5 model; and inputting the test data set into a trained improved lightweight YOLO v5 model to obtain a detection and identification result. The improved lightweight YOLO v5 model provided by the invention can more accurately identify ships in SAR images, greatly reduce the model size and test time, and remarkably improve the detection precision.
Description
Technical Field
The invention belongs to the technical field of radar target detection, and particularly relates to a lightweight SAR image ship target detection method based on YOLO v 5.
Background
Radar image target detection is a field of research and attention of a plurality of students in recent years, and a main stream target detection algorithm based on deep learning has two types, namely a single-stage type and a two-stage type according to the generation stage of a candidate frame. The single-stage target detection algorithm has higher detection speed, but has relatively lower detection accuracy. Recently, with the addition of some improvements, the detection accuracy of the method is greatly improved and even surpassed a two-stage model.
YOLO series is representative of a single-stage detection model, and the model construction can be completed only through end-to-end training. Due to the real-time advantage of YOLO, the method has become a research focus in the field of radar image target detection. The object detection model of the YOLO series becomes more and more powerful along with the introduction of YOLO v5, and YOLO v5 has the highest current reasoning speed and has a very light model size, so that YOLO v5 is selected as a detection framework. Therefore, the method is improved based on the YOLO v5 algorithm and is applied to light SAR image target detection.
In practical application, SAR image ship detection often faces complex sea scenes, ships have different scales, offshore environment interference exists, and coherent noise and background interference are serious. The deep learning network used by the target detection algorithm is mostly very complex, the parameter quantity and the calculation quantity are large, the generated model occupies a large memory, and the difficulty is brought to ocean monitoring. However, the light-weight model often causes a great reduction in detection accuracy, so that a more efficient target detection algorithm needs to be designed to meet the real-time performance.
Disclosure of Invention
The invention aims to provide a lightweight SAR image ship target detection method based on YOLO v5, aiming at the technical defects that the conventional lightweight SAR image detection recognition method has poor characteristic generalization capability, greatly reduced ship recognition rate and insufficient coast and ship characteristic learning, so that coast side information and ship target information cannot be effectively distinguished.
The technical solution for realizing the purpose of the invention is as follows: in a first aspect, the invention provides a YOLO v 5-based lightweight SAR image ship target detection method, which comprises the following steps:
step 1, acquiring SAR image data sets: obtaining a SAR image ship target simulation data set and a public SAR ship detection data set through simulation imaging, preprocessing the SAR image and dividing the SAR image into a training data set and a test data set;
step 2, an improved lightweight YOLO v5 model is built, namely a backbone network of YOLO v5 is replaced by a RepVGG network structure, the model is improved by combining a characterization enhancement module and a feature attention module, the characterization enhancement module is improved based on asymmetric convolution, the feature attention module is formed based on an SE attention mechanism, and an activation function is replaced by a SiLU activation function;
step 3, inputting the training data set into an improved lightweight YOLO v5 model for training, and obtaining a trained improved lightweight YOLO v5 model;
and 4, inputting the test data set into a trained improved lightweight YOLO v5 model to obtain a detection and identification result.
In a second aspect, the present invention provides a YOLO v 5-based lightweight SAR image ship target detection system, comprising:
the SAR image ship target simulation data set and the public SAR ship detection data set are obtained through simulation imaging, and the SAR image is preprocessed and divided into a training data set and a testing data set;
the second module is used for building an improved lightweight YOLO v5 model, namely replacing a backbone network of YOLO v5 with a RepVGG network structure, combining a characterization enhancement module and a feature attention module to improve the model, wherein the characterization enhancement module is improved based on asymmetric convolution, the feature attention module is formed based on an SE attention mechanism, and an activation function is replaced with a SiLU activation function;
the third module is used for inputting the training data set into the improved lightweight YOLO v5 model for training, and obtaining a trained improved lightweight YOLO v5 model;
and the fourth module is used for inputting the test data set into the trained improved lightweight YOLO v5 model to obtain a detection and identification result.
In a third aspect, the present invention provides a computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the steps of the method of the first aspect when the program is executed.
In a fourth aspect, the present invention provides a computer readable storage medium having stored thereon a computer program which when executed by a processor performs the steps of the method of the first aspect.
Compared with the existing lightweight detection technology, the invention has the remarkable advantages that: (1) When the method is applied to identifying ships in a coast scene, the characteristics of the coast and the ships can be effectively extracted, the coast and the ship information can be distinguished, and the method has a good identification rate for the ships in a complex sea scene; (2) The method can accurately identify small, medium and large ships in the SAR image, has better performance compared with the existing algorithm, has higher identification rate on the ships, and improves the identification efficiency on sea detection targets.
Drawings
FIG. 1 is a schematic diagram of an improved lightweight target detection method based on the YOLO v5 model.
FIG. 2 is a schematic diagram of an improved lightweight network that combines a characterization enhancement module with a feature attention module.
Detailed Description
The invention provides a SAR image ship lightweight target detection method under a complex sea scene, which comprises the following steps: acquiring an SAR image data set, obtaining an SAR image ship target simulation data set and a public SAR ship detection data set (actual measurement data set) through simulation imaging, preprocessing the data set, and dividing the data set into a training sample set and a test sample set; an improved lightweight YOLO v5 model is established, namely a backbone network of YOLO v5 is replaced by a RepVGG network structure, the model is improved by combining a characterization enhancement module and a feature attention module, and an activation function is replaced by a SiLU activation function; inputting the training data set into an improved lightweight YOLO v5 model for training, and obtaining a trained improved lightweight YOLO v5 model; and inputting the test data set into a trained improved lightweight YOLO v5 model to obtain a detection and identification result. The improved lightweight YOLO v5 model can more accurately identify ships in SAR images, and has higher detection precision.
The invention is described in further detail below with reference to the accompanying drawings.
Step 1, acquiring SAR image data sets: obtaining a SAR image ship target simulation data set and a public SAR ship detection data set (actual measurement data set) through simulation imaging, preprocessing the SAR image and dividing the SAR image into a training data set and a test data set;
step 2, an improved lightweight YOLO v5 model is built, namely a backbone network of YOLO v5 is replaced by a RepVGG network structure, the improved lightweight YOLO v5 model is improved by combining a characterization enhancement module and a feature attention module, the characterization enhancement module is improved based on asymmetric convolution, the feature attention module is formed based on an SE (sequential-and-specification) attention mechanism, and an activation function is replaced by a SiLU activation function;
step 3, inputting the training data set into an improved lightweight YOLO v5 model for training, and obtaining a trained improved lightweight YOLO v5 model;
step 4, inputting the test data set into a trained improved lightweight YOLO v5 model to obtain a detection and identification result;
further, in step 2, a lightweight YOLO v5 model of a plurality of RepVGG networks based on multi-module fusion is established. The method is characterized in that a backbone network of YOLO v5 is replaced by a RepVGG network structure, the network model is divided into two stages of training and reasoning on the basis of a lightweight model of RepVGG, a new residual branch is added in the training stage by combining the thought of residual branches, asymmetric convolution is integrated in the branch, a characterization enhancement module based on the asymmetric convolution is formed by the asymmetric convolution and conventional convolution, richer characteristic information is fitted in the SAR image characteristic extraction process, better robustness is achieved, a ship target is better focused, and ship lightweight detection precision under a complex sea scene is remarkably improved. In order to improve the attention degree of ship targets, SE channel attention is added to each layer of convolution and asymmetric convolution in a training stage, a plurality of characteristic attention modules based on SE channel attention are formed, the characteristic attention modules have stronger nonlinearity, complex association capacity between different channels can be improved, the ship targets are better focused in a network training process, ship detection results in complex environments can be remarkably improved, better effects are achieved compared with other lightweight networks, ship target detection accuracy can be better than that of other lightweight networks, and the ship targets have strong universality. The reasoning phase contains only two operations, 3 x 3 convolution and the ReLU activation function. This stage converts all network layer convolutions into 3 x 3 convolutions by means of a parameter fusion method. The dual-stage activation function is replaced by the SiLU activation function, and the SiLU activation function has the advantages of no upper bound, low bound, smoothness and non-monotone, and better performance than the ReLU activation function;
furthermore, the conventional convolution and the asymmetric convolution are mixed in a residual mode, so that the characterization enhancement module based on the asymmetric convolution is realized. The asymmetric convolution is to transform an n×n convolution kernel into three parallel convolution kernels of n×n, 1×n, and n×1. Meanwhile, 1×n and n×1 are spread out through the center of the n×n convolution kernel. And after model training is finished, the n multiplied by n convolution kernel is directly fused with the 1 multiplied by n and the n multiplied by 1 convolution kernel, and a nonlinear activation function is added between the 1 multiplied by n and the n multiplied by 1 convolution kernel, so that the nonlinearity of the model is improved. The asymmetric convolution can improve the expression capacity of the conventional convolution without additional time consumption, can improve the robustness of the model to the overturning SAR image, better extracts the characteristics of a ship target in the SAR image sea surface ship detection under the complex environment, reduces the influence of the near-shore complex scene on the ship detection, and obviously improves the detection precision. The characteristic enhancement module based on the asymmetric convolution is formed by constructing a conventional convolution and the asymmetric convolution in a residual form, and can enhance the feature extraction of ships and promote the robustness of the model. In the training and reasoning double stage, the reasoning stage still consists of a single branch, a characterization enhancement module is applied to the training stage, and the detection precision is improved and the generalization of the model is improved by improving the ship feature extraction performance;
further, SE channel attention is added to each layer of conventional convolution and asymmetric convolution, and a characteristic attention module based on the channel attention is constructed. In SAR image sea surface ship target detection under a complex environment, SE channel attention can learn SAR image features of different channels, and the feature extraction network can be enabled to pay more attention to ship features effectively, so that detection accuracy is improved. In SAR images in complex sea scenes, not all areas contribute equally to the detection task, only areas relevant to the ship need to be concerned. The spatial attention is used for searching the most important part in the network for processing, and in a complex near-shore scene, the ship is often interfered with the near-shore, and the spatial attention can ignore part of important information in the SAR image, so that ship detection is reduced. The feature attention module based on SE channel attention is to add the SE channel attention to all conventional convolution and asymmetric convolution, so that the attention degree of different ship features is enhanced by each operation, the complexity and the calculation burden of a model are only slightly increased, the influence of a near-shore complex environment on ship detection is effectively reduced, and the accuracy of model detection is improved.
As shown in fig. 1, the SAR image is preprocessed and then input into a detection network, feature extraction is performed based on an improved lightweight network combining a characterization enhancement module and a feature attention module, multi-scale feature fusion is realized through a feature aggregation network, and finally a loss function is calculated to predict the final result of the ship.
As shown in fig. 2, the lightweight network combines a characterization enhancement module based on asymmetric convolution and a feature attention module based on channel attention, forms a new improved lightweight network, and adds the new improved lightweight network as a feature extraction module of YOLO v5 to a detection flow.
Based on the same inventive concept, the invention also provides a lightweight SAR image ship target detection system based on YOLO v5, which comprises the following steps:
the SAR image ship target simulation data set and the public SAR ship detection data set are obtained through simulation imaging, and the SAR image is preprocessed and divided into a training data set and a testing data set;
the second module is used for building an improved lightweight YOLO v5 model, namely replacing a backbone network of YOLO v5 with a RepVGG network structure, combining a characterization enhancement module and a feature attention module to improve the model, wherein the characterization enhancement module is improved based on asymmetric convolution, the feature attention module is formed based on an SE attention mechanism, and an activation function is replaced with a SiLU activation function;
the third module is used for inputting the training data set into the improved lightweight YOLO v5 model for training, and obtaining a trained improved lightweight YOLO v5 model;
and the fourth module is used for inputting the test data set into the trained improved lightweight YOLO v5 model to obtain a detection and identification result.
The specific implementation method of each module is the same as the light SAR image ship target detection method, and is not repeated here.
The method for detecting the ship target of the lightweight SAR image based on the YOLO v5 is described in detail below with reference to the accompanying drawings and embodiments.
Example 1
The embodiment performs the ship target detection of the lightweight SAR image based on the YOLO v5, and the platform implemented by the embodiment is a CPU: intel (R) Core (TM) i7-8700CPU@3.20GHz,GPU: NVIDIA GeForce RTX 2080, 32g of memory; an operating system window10; acceleration was performed using CUDA 10.1. The data set is a common SAR vessel survey data set. The pair of light SAR image target detection results under the complex sea scene based on the YOLO v5 algorithm is shown in table 1.
Table 1 comparison of the method with other lightweight methods on a common SAR ship survey dataset
According to the improved lightweight model based on YOLO v5, which is disclosed by the table 1, the detection of sea surface ship targets in a complex environment has good performance, obvious advantages are achieved on indexes, SAR image offshore ship targets are easily affected by background clutter, coastal buildings and the like, the problems of poor detection effect, high false alarm rate and omission rate and the like of SAR image offshore ship targets are caused, the precision in the lightweight model is seriously reduced, and experiments prove that the lightweight structure disclosed by the chapter has higher detection rate under complex background and offshore scenes, and the SAR image offshore ship targets are superior to other lightweight structures.
Example 2
In the embodiment, lightweight SAR image target detection under a complex sea scene based on a YOLO v5 algorithm is performed, and a platform implemented by the embodiment is a CPU: intel (R) Core (TM) i7-8700CPU@3.20GHz,GPU: NVIDIA GeForce RTX 2080, 32g of memory; an operating system window10; acceleration was performed using CUDA 10.1. The data set is SAR image ship target simulation data set. The pair of target detection results of the lightweight SAR image under the complex sea field based on the YOLO v5 algorithm is shown in table 2.
Table 2 comparison of the method with other lightweight methods on SAR image ship target simulation data set
As can be seen from Table 2, compared with the original YOLO v5 model, the proposed model and the compared lightweight model are improved by about three times in test duration, and can better meet the real-time performance of sea surface ship target detection. The novel lightweight model can be obtained from the results, and the model has higher accuracy in detection of multi-class ships under the condition that the accuracy of other lightweight models is greatly reduced although the size of the model is slightly increased compared with that of other lightweight models, and the model has higher recognition rate in the complex environment of offshore ships and ships under different sea conditions, and has various indexes superior to that of the original model.
Claims (4)
1. A method for detecting a ship target of a lightweight SAR image based on YOLO v5 is characterized by comprising the following steps:
step 1, acquiring SAR image data sets: obtaining a SAR image ship target simulation data set and a public SAR ship detection data set through simulation imaging, preprocessing the SAR image and dividing the SAR image into a training data set and a test data set;
step 2, an improved lightweight YOLO v5 model is built, namely a backbone network of YOLO v5 is replaced by a RepVGG network structure, the model is improved by combining a characterization enhancement module and a feature attention module, the characterization enhancement module is improved based on asymmetric convolution, the feature attention module is formed based on an SE attention mechanism, and an activation function is replaced by a SiLU activation function;
a lightweight YOLO v5 model of a plurality of RepVGG networks based on multi-module fusion is established, namely a backbone network of the YOLO v5 is replaced by a RepVGG network structure, a new residual branch is newly added in a training stage on the basis of training reasoning double-stage lightweight of the RepVGG, and an asymmetric convolution-based characterization enhancement module, a SE channel attention-based feature attention module and a SiLU activation function are fused;
mixing conventional convolution and asymmetric convolution in a residual mode to realize a characterization enhancement module based on the asymmetric convolution; the asymmetric convolution is to transform an n×n convolution kernel into three parallel convolution kernels of n×n, 1×n, and n×1; and unwrapping 1 xn and n 1 through the center of the n xn convolution kernel; after training, directly fusing the n×n convolution kernels with the 1×n and n×1 convolution kernels, and adding a nonlinear activation function between the 1×n and n×1 convolution kernels;
adding SE channel attention to each layer of conventional convolution and asymmetric convolution to construct a characteristic attention module based on the channel attention;
step 3, inputting the training data set into an improved lightweight YOLO v5 model for training, and obtaining a trained improved lightweight YOLO v5 model;
and 4, inputting the test data set into a trained improved lightweight YOLO v5 model to obtain a detection and identification result.
2. A lightweight SAR image ship target detection system based on YOLO v5 is characterized by comprising:
the SAR image ship target simulation data set and the public SAR ship detection data set are obtained through simulation imaging, and the SAR image is preprocessed and divided into a training data set and a testing data set;
the second module is used for building an improved lightweight YOLO v5 model, namely replacing a backbone network of YOLO v5 with a RepVGG network structure, combining a characterization enhancement module and a feature attention module to improve the model, wherein the characterization enhancement module is improved based on asymmetric convolution, the feature attention module is formed based on an SE attention mechanism, and an activation function is replaced with a SiLU activation function;
a lightweight YOLO v5 model of a plurality of RepVGG networks based on multi-module fusion is established, namely a backbone network of the YOLO v5 is replaced by a RepVGG network structure, a new residual branch is newly added in a training stage on the basis of training reasoning double-stage lightweight of the RepVGG, and an asymmetric convolution-based characterization enhancement module, a SE channel attention-based feature attention module and a SiLU activation function are fused;
mixing conventional convolution and asymmetric convolution in a residual mode to realize a characterization enhancement module based on the asymmetric convolution; the asymmetric convolution is to transform an n×n convolution kernel into three parallel convolution kernels of n×n, 1×n, and n×1; and unwrapping 1 xn and n 1 through the center of the n xn convolution kernel; after training, directly fusing the n×n convolution kernels with the 1×n and n×1 convolution kernels, and adding a nonlinear activation function between the 1×n and n×1 convolution kernels;
adding SE channel attention to each layer of conventional convolution and asymmetric convolution to construct a characteristic attention module based on the channel attention;
the third module is used for inputting the training data set into the improved lightweight YOLO v5 model for training, and obtaining a trained improved lightweight YOLO v5 model;
and the fourth module is used for inputting the test data set into the trained improved lightweight YOLO v5 model to obtain a detection and identification result.
3. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the steps of the method of claim 1 when the program is executed by the processor.
4. A computer readable storage medium, on which a computer program is stored, characterized in that the program, when being executed by a processor, implements the steps of the method according to claim 1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310342382.5A CN116343045B (en) | 2023-03-30 | 2023-03-30 | Lightweight SAR image ship target detection method based on YOLO v5 |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310342382.5A CN116343045B (en) | 2023-03-30 | 2023-03-30 | Lightweight SAR image ship target detection method based on YOLO v5 |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116343045A CN116343045A (en) | 2023-06-27 |
| CN116343045B true CN116343045B (en) | 2024-03-19 |
Family
ID=86883871
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310342382.5A Active CN116343045B (en) | 2023-03-30 | 2023-03-30 | Lightweight SAR image ship target detection method based on YOLO v5 |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116343045B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116958907B (en) * | 2023-09-18 | 2023-12-26 | 四川泓宝润业工程技术有限公司 | Method and system for inspecting surrounding hidden danger targets of gas pipeline |
| CN117274192A (en) * | 2023-09-20 | 2023-12-22 | 重庆市荣冠科技有限公司 | Pipeline magnetic flux leakage defect detection method based on improved YOLOv5 |
| CN117115170B (en) * | 2023-10-25 | 2024-01-12 | 安徽大学 | Self-adaptive SAR ship detection method and system in unsupervised domain |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| GB202113615D0 (en) * | 2021-09-23 | 2021-11-10 | Ucl Business Ltd | Computer-implemented method of enhancing object detection in a digital image of known underlying structure,& corresponding module, data processing apparatus |
| CN114155445A (en) * | 2021-11-23 | 2022-03-08 | 北京理工大学 | SAR image target detection method based on improved YOLOv3 |
| CN115047455A (en) * | 2022-05-27 | 2022-09-13 | 南京理工大学 | Lightweight SAR image ship target detection method |
| CN115063665A (en) * | 2022-07-05 | 2022-09-16 | 西安邮电大学 | Lightweight fire detection method based on weighted bidirectional feature pyramid network |
| CN115100457A (en) * | 2022-06-02 | 2022-09-23 | 西安电子科技大学 | SAR image target detection method combining deep learning and CFAR |
| CN115205667A (en) * | 2022-08-02 | 2022-10-18 | 江苏大学 | Dense target detection method based on YOLOv5s |
| CN115311194A (en) * | 2022-05-23 | 2022-11-08 | 湘潭大学 | Automatic CT liver image segmentation method based on transformer and SE block |
| CN115546259A (en) * | 2022-09-22 | 2022-12-30 | 成都思晗科技股份有限公司 | Multi-target tracking method used in complex scene of transformer substation |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113269237B (en) * | 2021-05-10 | 2022-12-27 | 青岛理工大学 | Assembly change detection method, device and medium based on attention mechanism |
-
2023
- 2023-03-30 CN CN202310342382.5A patent/CN116343045B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| GB202113615D0 (en) * | 2021-09-23 | 2021-11-10 | Ucl Business Ltd | Computer-implemented method of enhancing object detection in a digital image of known underlying structure,& corresponding module, data processing apparatus |
| CN114155445A (en) * | 2021-11-23 | 2022-03-08 | 北京理工大学 | SAR image target detection method based on improved YOLOv3 |
| CN115311194A (en) * | 2022-05-23 | 2022-11-08 | 湘潭大学 | Automatic CT liver image segmentation method based on transformer and SE block |
| CN115047455A (en) * | 2022-05-27 | 2022-09-13 | 南京理工大学 | Lightweight SAR image ship target detection method |
| CN115100457A (en) * | 2022-06-02 | 2022-09-23 | 西安电子科技大学 | SAR image target detection method combining deep learning and CFAR |
| CN115063665A (en) * | 2022-07-05 | 2022-09-16 | 西安邮电大学 | Lightweight fire detection method based on weighted bidirectional feature pyramid network |
| CN115205667A (en) * | 2022-08-02 | 2022-10-18 | 江苏大学 | Dense target detection method based on YOLOv5s |
| CN115546259A (en) * | 2022-09-22 | 2022-12-30 | 成都思晗科技股份有限公司 | Multi-target tracking method used in complex scene of transformer substation |
Non-Patent Citations (5)
| Title |
|---|
| High Precision Feature Fast Extraction Strategy for Aircraft Attitude Sensor Fault Based on RepVGG and SENet Attention Mechanism;Zhen Jia 等;《sensors》;1-12 * |
| RepVGG: Making VGG-style ConvNets Great Again;Xiaohan Ding 等;《2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition》;13728-13737 * |
| 基于RB-YOLO生活垃圾实时检测算法;董霄霄 等;《智能计算机与应用》;第11卷(第10期);5-11、19 * |
| 基于改进RepVGG网络的车道线检测算法;杨鹏强 等;《交通信息与安全》;73-81 * |
| 多变环境下基于多尺度卷积网络的猪个体识别;王荣 等;《江西农业大学学报》;第42卷(第2期);391-400 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116343045A (en) | 2023-06-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116343045B (en) | Lightweight SAR image ship target detection method based on YOLO v5 | |
| Kong et al. | YOLOv3-DPFIN: A dual-path feature fusion neural network for robust real-time sonar target detection | |
| CN112766087A (en) | Optical remote sensing image ship detection method based on knowledge distillation | |
| CN111079739B (en) | Multi-scale attention feature detection method | |
| CN107844743A (en) | A kind of image multi-subtitle automatic generation method based on multiple dimensioned layering residual error network | |
| CN115223017B (en) | Multi-scale feature fusion bridge detection method based on depth separable convolution | |
| Liu et al. | Two-stage underwater object detection network using swin transformer | |
| CN112037225A (en) | Marine ship image segmentation method based on convolutional nerves | |
| CN116935332A (en) | Fishing boat target detection and tracking method based on dynamic video | |
| CN115410078A (en) | Low-quality underwater image fish target detection method | |
| CN114565824A (en) | Single-stage rotating ship detection method based on full convolution network | |
| Liu et al. | A Multi-scale Feature Pyramid SAR Ship Detection Network with Robust Background Interference | |
| Zhao et al. | ST-YOLOA: a Swin-transformer-based YOLO model with an attention mechanism for SAR ship detection under complex background | |
| CN116168240A (en) | Arbitrary-direction dense ship target detection method based on attention enhancement | |
| CN116206214A (en) | Automatic landslide recognition method, system, equipment and medium based on lightweight convolutional neural network and double attention | |
| CN116402761A (en) | Photovoltaic panel crack detection method based on double-channel multi-scale attention mechanism | |
| CN116468980A (en) | Infrared small target detection method and device for deep fusion of edge details and deep features | |
| CN113343924B (en) | Modulation signal identification method based on cyclic spectrum characteristics and generation countermeasure network | |
| CN115100428A (en) | Target detection method using context sensing | |
| Zhang et al. | SSRDet: Small Object Detection Based on Feature Pyramid Network | |
| CN114782983A (en) | Road scene pedestrian detection method based on improved feature pyramid and boundary loss | |
| Yang et al. | An Effective and Lightweight Hybrid Network for Object Detection in Remote Sensing Images | |
| Chen et al. | HFPNet: Super Feature Aggregation Pyramid Network for Maritime Remote Sensing Small Object Detection | |
| Shi et al. | Adaptive Feature Fusion With Attention-Guided Small Target Detection in Remote Sensing Images | |
| Zhou et al. | Weakly perceived object detection based on an improved CenterNet |
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 |