CN112989898A - Image processing method, system, computer device, readable storage medium and ship - Google Patents
Image processing method, system, computer device, readable storage medium and ship Download PDFInfo
- Publication number
- CN112989898A CN112989898A CN201911310620.4A CN201911310620A CN112989898A CN 112989898 A CN112989898 A CN 112989898A CN 201911310620 A CN201911310620 A CN 201911310620A CN 112989898 A CN112989898 A CN 112989898A
- Authority
- CN
- China
- Prior art keywords
- characteristic diagram
- feature map
- matrix representing
- standard
- neural network
- 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.)
- Pending
Links
Images
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
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
Abstract
The invention discloses an image processing method, an image processing system, computer equipment, a readable storage medium and a ship. The image processing method comprises the following steps: processing the image to be processed by a convolutional neural network method to determine a standard feature map; processing the standard feature map by a deformable convolutional neural network method to determine a deformed feature map; and determining a result characteristic diagram according to the standard characteristic diagram and the deformation characteristic diagram. Therefore, the matrix representing the image to be processed is processed through the convolutional neural network method to determine the matrix representing the standard characteristic diagram, the matrix representing the standard characteristic diagram is processed through the deformable convolutional neural network method to determine the matrix representing the deformed characteristic diagram, the matrix representing the result characteristic diagram is determined according to the matrix representing the standard characteristic diagram and the matrix representing the deformed characteristic diagram, and the matrix representing the result characteristic diagram is high in robustness and accuracy.
Description
Technical Field
The invention relates to the field of ships, in particular to an image processing method, an image processing system, a computer device, a readable storage medium and a ship.
Background
Surface target detection plays a very important role in the navigation of a ship. Object detection is an important issue in computer vision applications, not only to be able to identify the kind of object, but also to mark the position of the object in the image using a rectangular frame. In recent years, with the development of deep learning technology and the application of the technology in target detection, the target detection technology is rapidly developed, and a plurality of effective methods are appeared. The prior ship usually adopts a convolution neural network method to process the collected image, thereby realizing target detection.
However, the target detection method cannot adapt to the field of view of the ship, the attitude of the target to be detected and the deformation of the target to be detected, so that the target detection method has insufficient robustness and accuracy.
Therefore, the invention provides an image processing method, an image processing system, a computer device, a readable storage medium and a ship, which are used for solving the problems in the prior art.
Disclosure of Invention
In this summary, concepts in a simplified form are introduced that are further described in the detailed description. This summary of the invention is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used as an aid in determining the scope of the claimed subject matter.
The invention provides an image processing method, which comprises the following steps:
processing the image to be processed by a convolutional neural network method to determine a standard feature map;
processing the standard feature map by a deformable convolutional neural network method to determine a deformed feature map;
and determining a result characteristic diagram according to the standard characteristic diagram and the deformation characteristic diagram.
According to the image processing method, the matrix representing the image to be processed is processed through the convolutional neural network method to determine the matrix representing the standard characteristic diagram, the matrix representing the standard characteristic diagram is processed through the deformable convolutional neural network method to determine the matrix representing the deformed characteristic diagram, the matrix representing the result characteristic diagram is determined according to the matrix representing the standard characteristic diagram and the matrix representing the deformed characteristic diagram, the robustness and the accuracy of the matrix representing the result characteristic diagram are high, therefore, the matrix representing the result characteristic diagram can be used for replacing the matrix representing the standard characteristic diagram to complete the detection of the target, the detection accuracy is high, and the influence of factors such as sea waves and weather is reduced.
Optionally, the step of determining the result feature map according to the standard feature map and the deformation feature map includes:
the matrix of the standard profile and the matrix of the deformed profile are added to determine a matrix of the resulting profile.
Optionally, the step of determining the result feature map according to the standard feature map and the deformation feature map includes:
connecting the matrix of the standard characteristic diagram and the matrix of the deformed characteristic diagram to determine a matrix of the middle characteristic diagram;
the matrix of intermediate signatures is processed through a convolutional neural network to determine a resultant signature having dimensions that are the same as the dimensions of the standard signature.
Optionally, the convolutional neural network method is a fast RCNN method.
Optionally, the network structure of the deformable convolutional neural network method and/or the network structure of the convolutional neural network method is ResNet 101.
The present invention also provides an image processing system comprising:
the first determining device is used for processing the image to be processed by a convolutional neural network method to determine a standard feature map;
second determining means for processing the standard feature map by a deformable convolutional neural network method to determine a deformed feature map;
and the third determining device is used for determining the result characteristic diagram according to the standard characteristic diagram and the deformation characteristic diagram.
The image processing system processes the matrix representing the image to be processed by the convolutional neural network method to determine the matrix representing the standard characteristic diagram, processes the matrix representing the standard characteristic diagram by the deformable convolutional neural network method to determine the matrix representing the deformed characteristic diagram, and determines the matrix representing the result characteristic diagram according to the matrix representing the standard characteristic diagram and the matrix representing the deformed characteristic diagram.
The invention also provides a computer device comprising a memory, a processor and a computer program stored on the memory and operable on the processor, wherein the processor executes the program to implement the steps of:
processing the image to be processed by a convolutional neural network method to determine a standard feature map;
processing the standard feature map by a deformable convolutional neural network method to determine a deformed feature map;
and determining a result characteristic diagram according to the standard characteristic diagram and the deformation characteristic diagram.
The computer equipment processes the matrix representing the image to be processed by the convolutional neural network method to determine the matrix representing the standard characteristic diagram, processes the matrix representing the standard characteristic diagram by the deformable convolutional neural network method to determine the matrix representing the deformed characteristic diagram, and determines the matrix representing the result characteristic diagram according to the matrix representing the standard characteristic diagram and the matrix representing the deformed characteristic diagram.
The invention also provides a computer-readable storage medium, on which a computer program is stored which, when executed by a processor, carries out the steps of:
processing the image to be processed by a convolutional neural network method to determine a standard feature map;
processing the standard feature map by a deformable convolutional neural network method to determine a deformed feature map;
and determining a result characteristic diagram according to the standard characteristic diagram and the deformation characteristic diagram.
The computer-readable storage medium of the embodiment processes the matrix representing the image to be processed by the convolutional neural network method to determine the matrix representing the standard characteristic diagram, processes the matrix representing the standard characteristic diagram by the deformable convolutional neural network method to determine the matrix representing the deformed characteristic diagram, and determines the matrix representing the result characteristic diagram according to the matrix representing the standard characteristic diagram and the matrix representing the deformed characteristic diagram.
The invention also provides a ship, and the ship is controlled by the image processing method.
According to the ship of the embodiment, the matrix representing the image to be processed is processed through the convolutional neural network method to determine the matrix representing the standard characteristic diagram, the matrix representing the standard characteristic diagram is processed through the deformable convolutional neural network method to determine the matrix representing the deformed characteristic diagram, the matrix representing the result characteristic diagram is determined according to the matrix representing the standard characteristic diagram and the matrix representing the deformed characteristic diagram, and the matrix representing the result characteristic diagram is high in robustness and accuracy, so that the matrix representing the result characteristic diagram can be used for replacing the matrix representing the standard characteristic diagram to complete detection of the target, the detection accuracy is high, and influences of factors such as sea waves and weather are reduced.
Optionally, the vessel is an unmanned ship.
Drawings
The following drawings of the invention are included to provide a further understanding of the invention. The drawings illustrate embodiments of the invention and, together with the description, serve to explain the principles of the invention.
In the drawings:
fig. 1 is a block diagram of a structure of an image processing method according to an embodiment of the present invention;
FIG. 2 is a schematic flow chart of fast RCNN;
FIG. 3 is a schematic flow diagram of a deformable convolutional neural network method of the image processing method of FIG. 1;
FIG. 4 is a schematic diagram of a fusion of a standard feature map and a deformed feature map into a result feature map;
fig. 5 is a block diagram of the structure of an image processing system according to an embodiment of the present invention.
Detailed Description
In the following description, numerous specific details are set forth in order to provide a more thorough understanding of the present invention. It will be apparent, however, to one skilled in the art, that the present invention may be practiced without one or more of these specific details. In other instances, well-known features have not been described in order to avoid obscuring the invention.
It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of exemplary embodiments according to the invention. As used herein, the singular is intended to include the plural unless the context clearly dictates otherwise. It will be further understood that the terms "comprises" and/or "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
Exemplary embodiments according to the present invention will now be described in more detail with reference to the accompanying drawings. These exemplary embodiments may, however, be embodied in many different forms and should not be construed as limited to only the embodiments set forth herein. It is to be understood that these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the concept of these exemplary embodiments to those skilled in the art. In the drawings, the thicknesses of layers and regions are exaggerated for clarity, and the same elements are denoted by the same reference numerals, and thus the description thereof will be omitted.
One embodiment of the present invention provides an image processing method. The image processing method can be used for ships such as unmanned ships. When the unmanned ship sails, the images acquired by the unmanned ship can be processed in real time through the image processing method. And further, the running of the unmanned ship can be controlled according to the processing result.
As shown in fig. 1, the image processing method includes:
and processing the image to be processed by a convolutional neural network method to determine a standard feature map.
In the navigation process of the naval vessel, the naval vessel can acquire images around the naval vessel in real time through image acquisition equipment (a camera or a camera). The matrix representing the image it acquired is then processed by a convolutional neural network method on the COCO dataset to determine a matrix representing the standard feature map, as shown in figure 2.
The method of the convolutional neural network may be an existing method of the convolutional neural network. The convolutional neural network method processes a matrix representing an image acquired by the convolutional neural network method, and a matrix representing a standard characteristic diagram is obtained through one or more convolutional layers.
Preferably, the convolutional neural network method may be a fast RCNN method. The fast RCNN method is described in the documents S.Ren, K.He, R.Girshick, J.Sun, fast r-cnn: Towards real-time object detection with region pro-social networks, in: NeurIPS, 2015.
The network structure of the convolutional neural network method may be ResNet 101. Thereby, the network structure is stabilized.
The image processing method further includes:
the standard feature map is processed by a deformable convolutional neural network method to determine a deformed feature map.
Because the water surface target is influenced by sea waves, weather and the like, the water surface target has certain geometric deformation in the image to be processed. Therefore, processing a matrix representing an image to be processed by a convolutional neural network method may make the matrix representing the feature standard map less robust and accurate due to the small number of samples. For this, as shown in fig. 3, the matrix representing the standard feature map may be processed by a deformable convolution neural network method to determine a matrix representing the deformed feature map, the dimension of the matrix representing the deformed feature map being the same as the dimension of the matrix representing the standard feature map. The deformable convolutional neural network method adds a 2D offset in the convolutional neural network method, which is learned from a matrix representing a standard eigenmap through an additional volume base layer, thereby increasing the robustness of the matrix representing the deformed eigenmap.
Preferably, the Deformable convolutional neural network method is described in documents J.Dai, H.Qi, Y.Xiong, Y.Li, G.Zhang, H.Hu, Y.Wei, Deformable convolution networks, in: ICCV,2017.
The network structure of the deformable convolutional neural network method may be ResNet 101. Thereby, the network structure is stabilized.
The image processing method further includes:
and determining a result characteristic diagram according to the standard characteristic diagram and the deformation characteristic diagram.
As shown in fig. 4, the matrix representing the standard feature map and the matrix representing the deformed feature map may be processed by a fusion method to determine a matrix representing the resultant feature map. The dimension of the matrix representing the result feature map is the same as the dimension of the matrix representing the standard feature map.
In one embodiment, a matrix representing the standard profile and a matrix representing the deformed profile are added to determine a matrix representing the resultant profile. In this case, the dimension of the matrix representing the standard feature map is the same as the dimension of the matrix representing the deformed feature map. The dimension of the matrix representing the standard profile, the dimension of the matrix representing the deformed profile, and the dimension of the matrix representing the resultant profile are the same.
In another embodiment, a matrix representing the standard profile and a matrix representing the deformed profile are concatenated to determine a matrix representing the intermediate profile. The dimension of the matrix representing the standard feature map is different from the dimension of the matrix representing the deformed feature map. For example,
the matrix representing the standard feature map is
The matrix representing the deformation profile is
The matrix representing the intermediate feature map is
Or for example, the matrix representing the standard feature map and the matrix representing the deformed feature map are both a × B, and the matrix representing the intermediate feature map may be determined to be a × 2B after the matrix representing the standard feature map and the matrix representing the deformed feature map are connected. Wherein A and B can be any positive integer or 0. In other words, the matrix representing the standard characteristic diagram and the matrix representing the deformed characteristic diagram can be understood as two books. The matrix representing the standard characteristic diagram and the matrix representing the deformation characteristic diagram are connected by overlapping the two books along the thickness direction of the two books. The matrix representing the intermediate characteristic diagram determined by performing the connecting operation on the matrix representing the standard characteristic diagram and the matrix representing the deformed characteristic diagram are two books which are overlapped, and the thickness of the matrix representing the intermediate characteristic diagram is increased.
The matrix representing the intermediate feature map is then processed by a convolutional neural network to determine a matrix representing the resultant feature map such that the dimension of the matrix representing the resultant feature map is the same as the dimension representing the standard feature map. The convolution neural network processes the matrix representing the intermediate feature map in substantially the same manner as the conventional convolution neural network processes the matrix, and thus, the details thereof are not repeated.
Thus, the matrix representing the standard feature map and the matrix representing the deformed feature map are merged into a matrix representing the result feature map. The matrix representing the result characteristic diagram has high robustness and accuracy, and the influence of factors such as sea waves, weather and the like is reduced.
The image processing method of the embodiment processes the matrix representing the image to be processed by the convolutional neural network method to determine the matrix representing the standard characteristic diagram, processes the matrix representing the standard characteristic diagram by the deformable convolutional neural network method to determine the matrix representing the deformed characteristic diagram, and determines the matrix representing the result characteristic diagram according to the matrix representing the standard characteristic diagram and the matrix representing the deformed characteristic diagram.
After determining the matrix representing the result feature map, the present embodiment may replace the aforementioned matrix representing the standard feature map with the matrix representing the result feature map, and then identify the matrix representing the result feature map by a convolutional neural network method to detect the target.
The invention also provides an image processing system. As shown in fig. 5, the image processing system includes first determining means, second determining means, and third determining means; the first determining device is used for processing the image to be processed by a convolutional neural network method to determine a standard feature map. The second determining means is for processing the standard feature map by a deformable convolutional neural network method to determine a resultant feature map. And the third determining device is used for determining a result characteristic diagram according to the standard characteristic diagram and the deformation characteristic diagram.
The image processing system of the embodiment processes the matrix representing the image to be processed by the convolutional neural network method to determine the matrix representing the standard characteristic diagram, processes the matrix representing the standard characteristic diagram by the deformable convolutional neural network method to determine the matrix representing the deformed characteristic diagram, and determines the matrix representing the result characteristic diagram according to the matrix representing the standard characteristic diagram and the matrix representing the deformed characteristic diagram.
The invention also provides a computer device comprising a memory, a processor and a computer program stored on the memory and operable on the processor, wherein the processor executes the program to realize the following steps:
processing the image to be processed by a convolutional neural network method to determine a standard feature map;
processing the standard feature map by a deformable convolutional neural network method to determine a result feature map;
and determining a result characteristic diagram according to the standard characteristic diagram and the deformation characteristic diagram.
The computer device of the embodiment processes the matrix representing the image to be processed by the convolutional neural network method to determine the matrix representing the standard characteristic diagram, processes the matrix representing the standard characteristic diagram by the deformable convolutional neural network method to determine the matrix representing the deformed characteristic diagram, and determines the matrix representing the result characteristic diagram according to the matrix representing the standard characteristic diagram and the matrix representing the deformed characteristic diagram.
The invention also provides a computer-readable storage medium, on which a computer program is stored which, when executed by a processor, carries out the steps of:
processing the image to be processed by a convolutional neural network method to determine a standard feature map;
processing the standard feature map by a deformable convolutional neural network method to determine a result feature map;
and determining a result characteristic diagram according to the standard characteristic diagram and the deformation characteristic diagram.
The computer-readable storage medium of the embodiment processes the matrix representing the image to be processed by the convolutional neural network method to determine the matrix representing the standard characteristic diagram, processes the matrix representing the standard characteristic diagram by the deformable convolutional neural network method to determine the matrix representing the deformed characteristic diagram, and determines the matrix representing the result characteristic diagram according to the matrix representing the standard characteristic diagram and the matrix representing the deformed characteristic diagram.
The invention also provides a ship, and the ship is controlled by the image processing method.
According to the ship of the embodiment, the matrix representing the image to be processed is processed through the convolutional neural network method to determine the matrix representing the standard characteristic diagram, the matrix representing the standard characteristic diagram is processed through the deformable convolutional neural network method to determine the matrix representing the deformed characteristic diagram, the matrix representing the result characteristic diagram is determined according to the matrix representing the standard characteristic diagram and the matrix representing the deformed characteristic diagram, and the matrix representing the result characteristic diagram is high in robustness and accuracy, so that the matrix representing the result characteristic diagram can be used for replacing the matrix representing the standard characteristic diagram to complete detection of the target, the detection accuracy is high, and influences of factors such as sea waves and weather are reduced.
Preferably, the vessel is an unmanned ship.
The object to be measured is detected using the image processing method of the present invention, and the accuracy of the embodiment of the matrix representing the resultant feature map is determined to be 35.3% by adding the matrix representing the standard feature map and the matrix representing the deformed feature map as described above. The join operation was performed as described above to determine the accuracy of an embodiment of the matrix representing the intermediate feature map of 36.5%.
The present invention has been illustrated by the above embodiments, but it should be understood that the above embodiments are for illustrative and descriptive purposes only and are not intended to limit the invention to the scope of the described embodiments. Furthermore, it will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, and that many variations and modifications may be made in accordance with the teachings of the present invention, which variations and modifications are within the scope of the present invention as claimed. The scope of the invention is defined by the appended claims and equivalents thereof.
The flows described in all the preferred embodiments described above are only examples. Unless an adverse effect occurs, various processing operations may be performed in a different order from the order of the above-described flow. The above-mentioned steps of the flow can be added, combined or deleted according to the actual requirement.
Further, the commands, command numbers, and data items described in all the preferred embodiments described above are only examples, and thus the commands, command numbers, and data items may be set in any manner as long as the same functions are achieved. The units of the terminal of the preferred embodiments may also be integrated, further divided or subtracted according to actual needs.
Claims (10)
1. An image processing method, comprising:
processing the image to be processed by a convolutional neural network method to determine a standard feature map;
processing the standard feature map by a deformable convolutional neural network method to determine a deformed feature map;
and determining a result characteristic diagram according to the standard characteristic diagram and the deformation characteristic diagram.
2. The image processing method of claim 1, wherein the step of determining a resultant feature map from the standard feature map and the deformed feature map comprises:
adding the matrix of the standard feature map and the matrix of the deformed feature map to determine a matrix of the resultant feature map.
3. The image processing method of claim 1, wherein the step of determining a resultant feature map from the standard feature map and the deformed feature map comprises:
connecting the matrix of the standard characteristic diagram and the matrix of the deformed characteristic diagram to determine a matrix of an intermediate characteristic diagram;
processing the matrix of intermediate feature maps by a convolutional neural network to determine a resultant feature map, the dimension of the resultant feature map being the same as the dimension of the standard feature map.
4. The image processing method according to claim 1, wherein the convolutional neural network method is a fast RCNN method.
5. The image processing method of claim 1, wherein the network structure of the deformable convolutional neural network method and/or the network structure of the convolutional neural network method is ResNet 101.
6. An image processing system, comprising:
the first determining device is used for processing the image to be processed by a convolutional neural network method to determine a standard feature map;
second determining means for processing the standard feature map by a deformable convolutional neural network method to determine a deformed feature map;
and the third determining device is used for determining a result characteristic diagram according to the standard characteristic diagram and the deformation characteristic diagram.
7. A computer device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor executes the program to perform the steps of:
processing the image to be processed by a convolutional neural network method to determine a standard feature map;
processing the standard feature map by a deformable convolutional neural network method to determine a deformed feature map;
and determining a result characteristic diagram according to the standard characteristic diagram and the deformation characteristic diagram.
8. A computer-readable storage medium, on which a computer program is stored which, when executed by a processor, carries out the steps of:
processing the image to be processed by a convolutional neural network method to determine a standard feature map;
processing the standard feature map by a deformable convolutional neural network method to determine a deformed feature map;
and determining a result characteristic diagram according to the standard characteristic diagram and the deformation characteristic diagram.
9. A ship, characterized in that the ship is controlled by the image processing method of any one of claims 1 to 5.
10. The vessel of claim 9, wherein the vessel is an unmanned vessel.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911310620.4A CN112989898A (en) | 2019-12-18 | 2019-12-18 | Image processing method, system, computer device, readable storage medium and ship |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911310620.4A CN112989898A (en) | 2019-12-18 | 2019-12-18 | Image processing method, system, computer device, readable storage medium and ship |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112989898A true CN112989898A (en) | 2021-06-18 |
Family
ID=76343924
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911310620.4A Pending CN112989898A (en) | 2019-12-18 | 2019-12-18 | Image processing method, system, computer device, readable storage medium and ship |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112989898A (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109035184A (en) * | 2018-06-08 | 2018-12-18 | 西北工业大学 | A kind of intensive connection method based on the deformable convolution of unit |
| CN109766873A (en) * | 2019-02-01 | 2019-05-17 | 中国人民解放军陆军工程大学 | A kind of pedestrian mixing deformable convolution recognition methods again |
| US20210056351A1 (en) * | 2018-06-04 | 2021-02-25 | Jiangnan University | Multi-scale aware pedestrian detection method based on improved full convolutional network |
-
2019
- 2019-12-18 CN CN201911310620.4A patent/CN112989898A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20210056351A1 (en) * | 2018-06-04 | 2021-02-25 | Jiangnan University | Multi-scale aware pedestrian detection method based on improved full convolutional network |
| CN109035184A (en) * | 2018-06-08 | 2018-12-18 | 西北工业大学 | A kind of intensive connection method based on the deformable convolution of unit |
| CN109766873A (en) * | 2019-02-01 | 2019-05-17 | 中国人民解放军陆军工程大学 | A kind of pedestrian mixing deformable convolution recognition methods again |
Non-Patent Citations (2)
| Title |
|---|
| 高鑫等: "基于可变形卷积神经网络的遥感影像密集区域车辆检测方法", 《电子与信息学报》 * |
| 黄龙: "基于全卷积神经网络的室内场景分割算法研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP3252715A1 (en) | Two-camera relative position calculation system, device and apparatus | |
| CN109753838A (en) | Two-dimensional code identification method, device, computer equipment and storage medium | |
| JP4985166B2 (en) | Self-position estimation device | |
| CN106447601B (en) | Unmanned aerial vehicle remote sensing image splicing method based on projection-similarity transformation | |
| US20140219569A1 (en) | Image recognition system and method for identifying similarities in different images | |
| CN110969145B (en) | Remote sensing image matching optimization method and device, electronic equipment and storage medium | |
| CN111860489A (en) | Certificate image correction method, device, equipment and storage medium | |
| CN112613506A (en) | Method and device for recognizing text in image, computer equipment and storage medium | |
| US20200005078A1 (en) | Content aware forensic detection of image manipulations | |
| CN112163995A (en) | Splicing generation method and device for oversized aerial photographing strip images | |
| CN113297986A (en) | Handwritten character recognition method, device, medium and electronic equipment | |
| CN113221897B (en) | Image correction method, image text recognition method, identity verification method and device | |
| CN112989898A (en) | Image processing method, system, computer device, readable storage medium and ship | |
| CN111582013A (en) | Ship retrieval method and device based on gray level co-occurrence matrix characteristics | |
| CN114463430B (en) | Ocean search and rescue system and method based on image processing | |
| CN110660091A (en) | Image registration processing method and device and photographing correction operation system | |
| CN115131433A (en) | Non-cooperative target pose processing method and device and electronic equipment | |
| CN106651950B (en) | Single-camera pose estimation method based on quadratic curve perspective projection invariance | |
| CN111931786B (en) | Image processing method and device and computer readable storage medium | |
| JP2008032478A (en) | Self-location estimator | |
| CN106709942A (en) | Panoramic image mistaken matching elimination method based on characteristic azimuth | |
| CN112862676A (en) | Image splicing method, device and storage medium | |
| CN112907662A (en) | Feature extraction method and device, electronic equipment and storage medium | |
| CN111079523B (en) | Object detection method, device, computer equipment and storage medium | |
| CN115131401B (en) | Remote sensing video target tracking method based on multi-scale multi-direction kernel correlation filtering |
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 |