CN111553183A - Ship detection model training method, ship detection method and ship detection device - Google Patents
Ship detection model training method, ship detection method and ship detection device Download PDFInfo
- Publication number
- CN111553183A CN111553183A CN201911377878.6A CN201911377878A CN111553183A CN 111553183 A CN111553183 A CN 111553183A CN 201911377878 A CN201911377878 A CN 201911377878A CN 111553183 A CN111553183 A CN 111553183A
- Authority
- CN
- China
- Prior art keywords
- ship
- training sample
- training
- detection model
- ship 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.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/41—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
Abstract
The invention discloses a ship detection model training method, a ship detection method and a ship detection device, wherein the ship detection model training method comprises the following steps: acquiring a ship training sample, wherein the ship training sample comprises a plurality of sea surface environment images; extracting the ship training sample to obtain a first ship training sample; dividing the first ship training sample into m x n grids to obtain a target training sample, wherein m is smaller than n; and training according to the target ship training sample to obtain a ship detection model. According to the invention, the ship training sample is non-uniformly divided, and the longitudinal detection density is increased under the condition of keeping the transverse detection density unchanged, so that the shape of the training sample after cutting is consistent with that of the ship, the ship detection is more accurate, and the influence of sea waves on the ship is reduced.
Description
Technical Field
The invention relates to the technical field of recognition, in particular to a ship detection model training method, a ship detection method and a ship detection device.
Background
The identification of the marine vessel has important strategic significance on marine security protection, and great help is provided for the strategic implementation of the ocean strong country in China. The marine ship detection means that characteristic information of a ship input in an image or video mode is extracted, the extracted characteristic information is matched with ship information in a database, and the matched ship information and an input ship signal are divided into the same type. The marine ship detection is widely applied, for example, the approach of malicious ships, the entering of non-national ships into China's territory, an intelligent monitoring system, marine big data and the like. However, in the sea surface vessel identification, the sea surface vessel identification is greatly influenced by wind waves, fog, light and the like, wherein the vessel is influenced by sea waves, so that the false alarm rate of the vessel detection is high.
Disclosure of Invention
Therefore, the technical problem to be solved by the invention is to overcome the defect of low detection rate of the off-board ship in the sea wave problem in the prior art, so that the ship detection model training method, the ship detection method and the ship detection device are provided.
According to a first aspect, the embodiment of the invention discloses a ship detection model training method, which comprises the following steps: acquiring a ship training sample, wherein the ship training sample comprises a plurality of sea surface environment images; extracting the ship training sample to obtain a first ship training sample; dividing the first ship training sample into m x n grids to obtain a target training sample, wherein m is smaller than n; and training according to the target ship training sample to obtain a ship detection model.
With reference to the first aspect, in a first implementation manner of the first aspect, the extracting the ship training samples to obtain first ship training samples includes: extracting images of the ship training samples with the influence of sea waves; and marking the image with the influence of the sea waves to obtain a first ship training sample.
With reference to the first embodiment of the first aspect, in a second embodiment of the first aspect, the dividing the training samples of the first vessel into m × n grids to obtain target training samples, where m is smaller than n includes: performing convolution operation on the first ship training sample according to a first preset convolution core to obtain a first target training sample; and performing dimensionality reduction processing on the first target training sample according to a second preset convolution kernel to obtain a target training sample.
With reference to the first aspect, in a third implementation manner of the first aspect, after the training of the ship detection model according to the target ship training samples, the method further includes: acquiring ship image test samples, wherein the ship image test samples comprise a positive test sample containing a ship and a negative test sample containing no ship; obtaining a test result according to the ship image test sample and the ship detection model; and when the test result meets a preset condition, determining the ship detection model as an available ship detection model.
According to a second aspect, an embodiment of the present invention further discloses a ship detection method, including the following steps: acquiring a ship detection image, wherein the ship detection image comprises a ship and sea waves; inputting the ship detection image into the ship detection model to obtain a ship detection result; the ship detection model is generated by training through the ship detection model training method according to the first aspect or any embodiment of the first aspect.
With reference to the second aspect, in a first embodiment of the second aspect, after the acquiring the ship inspection image, the method further includes: segmenting the ship detection image into a grid of m x n, m being smaller than n.
According to a third aspect, an embodiment of the present invention further discloses a training device for a ship detection model, including: the system comprises a first acquisition module, a second acquisition module and a third acquisition module, wherein the first acquisition module is used for acquiring a ship training sample, and the ship training sample comprises a plurality of sea surface environment images; the first extraction module is used for extracting the ship training sample to obtain a first ship training sample; the segmentation module is used for segmenting the first ship training sample into m x n grids to obtain a target training sample, wherein m is smaller than n; and the training module is used for training according to the target ship training sample to obtain a ship detection model.
According to a fourth aspect, an embodiment of the present invention further discloses a ship detection apparatus, including: the second acquisition module is used for acquiring a ship detection image; a detection module, configured to input the ship detection image into the ship detection model to obtain a ship detection result, where the ship detection model is generated by training through a ship detection model training method according to any one of the first aspect and the first embodiment.
According to a fifth aspect, an embodiment of the present invention further discloses a computer device, including: at least one processor; and a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the one processor to cause the at least one processor to perform the ship detection model training method of the first aspect or any of the embodiments of the first aspect or the ship detection method of any of the embodiments of the second aspect or the second aspect.
According to a sixth aspect, the present invention further discloses a computer-readable storage medium, on which computer instructions are stored, which when executed by a processor implement the ship detection model training method according to the first aspect or any of the embodiments of the first aspect, or the ship detection method according to any of the embodiments of the second aspect or the second aspect.
The technical scheme of the invention has the following advantages:
1. the ship detection model training method and device provided by the invention are characterized in that a ship training sample is obtained, the ship training sample comprises a plurality of sea surface environment images, the ship training sample is extracted to obtain a first ship training sample, the first ship training sample is divided into m x n grids to obtain a target training sample, m is smaller than n, and a ship detection model is obtained according to training of the target ship training sample. According to the invention, the ship training sample is non-uniformly divided, and the longitudinal detection density is increased under the condition of keeping the transverse detection density unchanged, so that the shape of the training sample after cutting is consistent with that of the ship, the ship detection is more accurate, and the influence of sea waves on the ship is reduced.
2. According to the ship detection method and device, the ship detection image is acquired, the ship detection image comprises the ship and the sea waves, the characteristic extraction is carried out on the ship detection image to obtain the ship characteristic, the ship characteristic is input into the ship detection model to obtain the ship detection result, and the ship detection image is input into the ship detection model, so that the ship detection is more accurate, and the influence of the sea waves on the ship is reduced.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
FIG. 1 is a flowchart of a specific example of a training method for a ship detection model according to embodiment 1 of the present invention;
fig. 2 is a flowchart of a specific example of a ship detection method according to embodiment 2 of the present invention;
FIG. 3 is a schematic block diagram of a specific example of the ship detection model training apparatus according to embodiment 3 of the present invention;
fig. 4 is a schematic block diagram of a specific example of the ship detection apparatus according to embodiment 4 of the present invention;
fig. 5 is a schematic block diagram of a specific example of a computer device in embodiment 5 of the present invention.
Detailed Description
The technical solutions of the present invention will be described clearly and completely with reference to the accompanying drawings, and it should be understood that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The terms "first," "second," and "third" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
In addition, the technical features involved in the different embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.
Example 1
The embodiment provides a ship detection model training method, which is applied to ship identification affected by sea waves, and as shown in fig. 1, the method comprises the following steps:
s11: and acquiring a ship training sample, wherein the ship training sample comprises a plurality of sea surface environment images.
For example, the ship training sample may be obtained by decoding a video captured by a certain camera, or may be obtained by searching from a search engine, the obtaining method of the ship training sample is not limited in the embodiment of the present application, and may be selected according to actual situations, the ship training sample may include a plurality of images including sea environment, such as ocean waves, marine flying animals, and the like, the ship training sample may also be various, such as various types of ships, various sea environments, and the like, the embodiment of the present application does not limit the ship training sample, and a person skilled in the art may select the ship training sample according to actual situations.
S12: and extracting the ship training sample to obtain a first ship training sample.
For example, the first ship training sample can be used for marking images of more sea waves and ship types and positions in the images, the images of more sea waves can be screened by adopting a manual visual observation method, and the images can be marked by adopting a manual marking method. The screening method and the labeling method are not limited in the present application, and those skilled in the art can select the method according to actual situations.
S13: and dividing the first ship training sample into m x n grids to obtain a target training sample, wherein m is smaller than n.
Illustratively, when a ship is detected by a traditional ship detection model, a first ship training sample is divided into uniform a-a grids, each grid provides a plurality of candidate frames, the candidate frames with marked ship center points are used for predicting the ship in an image, but the image is divided into uniform grids, the horizontal direction and the vertical direction of the image have the same detection density, actually, the shapes of a plurality of ships are long, the image has the characteristics of dense horizontal distribution and sparse vertical distribution, when the ship influenced by sea waves is detected, the ship detection result may not be accurate enough.
S14: and training according to the target ship training sample to obtain a ship detection model.
In an exemplary embodiment, a target ship training sample is input into a neural network learning network model for training, the training method may be supervised or unsupervised, the training method is not limited in the embodiment of the present application, a weight of the neural network learning network model can be continuously adjusted according to the target ship training sample, and the neural network learning network model is continuously trained and optimized to obtain a ship detection model.
The ship detection model training method provided by the invention comprises the steps of obtaining a ship training sample, extracting the ship training sample to obtain a first ship training sample, dividing the first ship training sample into m x n grids to obtain a target training sample, wherein m is smaller than n, and training according to the target ship training sample to obtain a ship detection model. According to the invention, the ship training sample is non-uniformly divided, and the longitudinal detection density is increased under the condition of keeping the transverse detection density unchanged, so that the shape of the training sample after cutting is consistent with that of the ship, the ship detection is more accurate, and the influence of sea waves on the ship is reduced.
As an optional embodiment of the present application, extracting a ship training sample to obtain a first ship training sample includes:
first, an image of the presence of the effects of waves in a training sample of a ship is extracted.
For example, in the embodiment of the present application, an image 2/3 of the image area occupied by sea waves may be regarded as a picture with more sea waves, or an image obtained by blocking more than half of a ship by sea waves may be regarded as a picture with more sea waves.
And secondly, marking the image with the influence of the sea waves to obtain a first ship training sample.
For example, in the embodiment of the present application, the extracted image affected by ocean waves may be preprocessed in a manual labeling manner, the type and the position of the ship in the image are labeled, so as to obtain a first ship training sample, and in order to improve the detection accuracy, the first ship training sample may be checked multiple times, so that it is ensured that the first ship training sample may be stored in a format of a PASCAL V0C2007 data set after no problem exists, and may be directly called during training.
As an optional embodiment of the present application, the dividing the first ship training sample into m × n grids to obtain the target training sample, where m is smaller than n, includes:
firstly, carrying out convolution operation on a first ship training sample according to a first preset convolution kernel to obtain a first target training sample.
Illustratively, the first predetermined convolution kernel may be 3 × 3 convolution kernels, and may also be 5 × 5 convolution kernels, and the first predetermined convolution kernel is not limited in the embodiment of the present application, and a person skilled in the art may set the first predetermined convolution kernel according to practical situations, and in the embodiment of the present application, the size of the input image may be defined as 448 × 448, and the convolution operation may be to convolve the feature map of 28 × 512 according to the size and number of convolution kernels of 3 × 512, so as to obtain 14 × 28 first target training samples, and implement the segmentation of the first ship training samples.
And secondly, performing dimensionality reduction processing on the first target training sample according to a second preset convolution kernel to obtain a target training sample.
For example, the second preset convolution kernel may be 1 × 1, in this embodiment of the present application, the 1 × 1 convolution kernel may be disposed between the first preset convolution kernels, so as to effectively reduce the number of channels of the first target training sample, and the pooling layer further samples the first target training sample by using 2 × 2 convolution kernels, so as to reduce information redundancy and reduce the influence of the sea wave on ship detection.
As an optional embodiment of the present application, after the ship detection model is trained according to the target ship training sample, the ship detection model training method further includes:
first, ship image test samples are obtained, which include a positive test sample containing a ship and a negative test sample containing no ship.
For example, the ship image test sample may be obtained by randomly dividing a part of the ship image sample obtained in step S11, such as a part different from the remaining 30% of the ship image training sample as the image test sample. The acquisition mode of the ship image test sample is not limited in the embodiment, and can be determined according to the requirement.
And secondly, obtaining a test result according to the ship image test sample and the ship detection model.
Exemplarily, in the embodiment of the present application, all ship image test samples are input to a ship detection model, an intersection ratio calculation is performed according to a plurality of prediction boundary boxes output by the ship detection model and detection results and marks in the prediction boundary boxes, and an average value of calculation results of all intersection ratios is used as an accuracy rate, so as to obtain an accuracy rate of an output result of a neural network; or carrying out cross comparison calculation according to a plurality of prediction boundary frames output by the ship detection model and detection results and marks in the prediction boundary frames, judging whether each cross comparison calculation result meets a preset threshold value, indicating that the detection result of the ship image test sample is accurate when the preset threshold value is met, and taking the ratio of the ship image test sample with the accurate detection result to all ship image test samples as the accuracy of the output result of the neural network. The embodiment does not limit what kind of method is specifically adopted to obtain the test result, and can determine according to the needs.
And thirdly, when the test result meets the preset condition, determining the ship detection model as an available ship detection model. The preset condition may be that the accuracy of the ship identification is 99%, and the accuracy standard is not limited in the embodiment of the present application and may be determined according to an actual situation.
Example 2
An embodiment of the present invention further provides a ship detection method, which is applied to detecting a ship affected by ocean waves, and as shown in fig. 2, the method includes the following steps:
s21: a ship detection image is acquired, and the ship detection image comprises a ship and sea waves.
Illustratively, the ship detection image may be a monitoring video acquired by a camera device, and the ship detection image may be directly uploaded to a terminal by a user in a wired or wireless manner, or may be recorded in the terminal. The acquisition mode of the ship detection image is not particularly limited in the embodiment of the application, and a person skilled in the art can determine the acquisition mode according to actual use requirements.
S22: inputting the ship characteristics into a ship detection model to obtain a ship detection result; the ship detection model is generated by training through the ship detection model training method as in embodiment 1 above.
According to the ship detection method, the ship detection image is acquired, the ship detection image comprises the ship and the sea waves, the characteristic extraction is carried out on the ship detection image to obtain the ship characteristic, the ship characteristic is input into the ship detection model to obtain the ship detection result, and the ship detection image is input into the ship detection model, so that the ship detection is more accurate, and the influence of the sea waves on the ship is reduced.
As an optional embodiment of the present application, after acquiring the ship detection image, the ship detection method further includes: the ship detection image is segmented into a grid of m x n, m being smaller than n.
Example 3
An embodiment of the present invention further provides a training device for a ship detection model, as shown in fig. 3, including:
the first acquisition module 31 is configured to acquire a ship training sample, where the ship training sample includes multiple sea surface environment images; the specific implementation manner is shown in step S11 in embodiment 1, and details are not described here.
The first extraction module 32 is configured to extract a ship training sample to obtain a first ship training sample; the specific implementation manner is shown in step S12 in embodiment 1, and details are not described here.
A dividing module 33, configured to divide the first ship training sample into m × n grids to obtain a target training sample, where m is smaller than n; the specific implementation manner is shown in step S13 in embodiment 1, and details are not described here.
And the training module 34 is used for training the ship detection model according to the target ship training sample. The specific implementation manner is shown in step S14 in embodiment 1, and details are not described here.
The ship detection model training device provided by the invention is characterized in that a ship training sample is obtained, the ship training sample comprises a plurality of sea surface environment images, the ship training sample is extracted to obtain a first ship training sample, the first ship training sample is divided into m x n grids to obtain a target training sample, m is smaller than n, and a ship detection model is obtained according to training of the target ship training sample. According to the invention, the ship training sample is non-uniformly divided, and the longitudinal detection density is increased under the condition of keeping the transverse detection density unchanged, so that the shape of the training sample after cutting is consistent with that of the ship, the ship detection is more accurate, and the influence of sea waves on the ship is reduced.
As an optional embodiment of the present application, the first extraction module 32 includes:
and the first extraction submodule is used for extracting the image of the influence of the sea waves in the ship training sample. The specific implementation manner is shown in the corresponding steps in embodiment 1, and is not described herein again.
And the marking module is used for marking the image with the influence of the sea waves to obtain a first ship training sample. The specific implementation manner is shown in the corresponding steps in embodiment 1, and is not described herein again.
As an alternative embodiment of the present application, the segmentation module 33 includes:
and the convolution module is used for performing convolution operation on the first ship training sample according to a first preset convolution kernel to obtain a first target training sample. The specific implementation manner is shown in the corresponding steps in embodiment 1, and is not described herein again.
And the dimension reduction module is used for carrying out dimension reduction processing on the first target training sample according to a second preset convolution kernel to obtain a target training sample. The specific implementation manner is shown in the corresponding steps in embodiment 1, and is not described herein again.
As an optional embodiment of the present application, the ship detection model training apparatus further includes:
the third acquisition module is used for acquiring ship image test samples, wherein the ship image test samples comprise a positive test sample containing ships and a negative test sample containing no ships; the specific implementation manner is shown in the corresponding steps in embodiment 1, and is not described herein again.
The test module is used for obtaining a test result according to the ship image test sample and the ship detection model; the specific implementation manner is shown in step S15 in embodiment 1, and details are not described here.
And the determining module is used for determining the ship detection model as an available ship detection model when the test result meets the preset condition. The specific implementation manner is shown in the corresponding steps in embodiment 1, and is not described herein again.
Example 4
An embodiment of the present invention further provides a ship detection apparatus, as shown in fig. 4, including:
a second acquisition module 41 for acquiring a ship detection image; the specific implementation manner is shown in step S21 in embodiment 2, and details are not described here.
The detection module 42 is used for inputting the ship detection image into the ship detection model to obtain a ship detection result; the ship detection model is generated by training through the ship detection model training method as in embodiment 1 above. The specific implementation manner is shown in step S22 in embodiment 2, and details are not described here.
According to the ship detection device, the ship detection image is acquired, the ship detection image comprises the ship and the sea waves, the characteristic extraction is carried out on the ship detection image to obtain the ship characteristic, the ship characteristic is input into the ship detection model to obtain the ship detection result, and the ship detection image is input into the ship detection model, so that the ship detection is more accurate, and the influence of the sea waves on the ship is reduced.
As an optional embodiment of the present application, the ship detection apparatus further includes:
and the image segmentation module is used for segmenting the ship detection image into a grid of m x n, wherein m is smaller than n. The specific implementation manner is shown in the corresponding steps in embodiment 2, and is not described herein again.
Example 5
An embodiment of the present invention further provides a computer device, as shown in fig. 5, the computer device may include a processor 51 and a memory 52, where the processor 51 and the memory 52 may be connected by a bus or in another manner, and fig. 5 takes the example of connection by a bus as an example.
The processor 51 may be a Central Processing Unit (CPU). The Processor 51 may also be other general purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field Programmable Gate Arrays (FPGAs) or other Programmable logic devices, discrete Gate or transistor logic devices, discrete hardware components, or combinations thereof.
The memory 52, as a non-transitory computer readable storage medium, may be used for storing non-transitory software programs, non-transitory computer executable programs, and modules, such as program instructions/modules corresponding to the ship detection model training method or the ship detection method in the embodiment of the present invention (for example, the first obtaining module 31, the first extracting module 32, the segmenting module 33, and the training module 34 shown in fig. 3 or the second obtaining module 41 and the detecting module 42 shown in fig. 4). The processor 51 executes various functional applications and data processing of the processor by running non-transitory software programs, instructions and modules stored in the memory 52, namely, implementing the ship detection model training method or the ship detection method in the above method embodiments.
The memory 52 may include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function; the storage data area may store data created by the processor 51, and the like. Further, the memory 52 may include high speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid state storage device. In some embodiments, the memory 52 may optionally include memory located remotely from the processor 51, and these remote memories may be connected to the processor 51 via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The one or more modules are stored in the memory 52 and, when executed by the processor 51, perform a ship detection model training method as in the embodiment shown in fig. 1 or a ship detection method as in the embodiment shown in fig. 2.
The details of the electronic device may be understood with reference to the corresponding descriptions and effects in the embodiments shown in fig. 1 or fig. 2, and are not described herein again.
Example 6
The embodiment of the invention also provides a computer storage medium, wherein the computer storage medium stores computer executable instructions, and the computer executable instructions can execute the ship detection model training method or the ship detection method in any method embodiment. The storage medium may be a magnetic Disk, an optical Disk, a Read-only Memory (ROM), a Random Access Memory (RAM), a flash Memory (FlashMemory), a Hard Disk (Hard Disk Drive, abbreviated as HDD) or a Solid State Drive (SSD), etc.; the storage medium may also comprise a combination of memories of the kind described above.
It should be understood that the above examples are only for clarity of illustration and are not intended to limit the embodiments. Other variations and modifications will be apparent to persons skilled in the art in light of the above description. And are neither required nor exhaustive of all embodiments. And obvious variations or modifications therefrom are within the scope of the invention.
Claims (10)
1. A ship detection model training method is characterized by comprising the following steps:
acquiring a ship training sample, wherein the ship training sample comprises a plurality of sea surface environment images;
extracting the ship training sample to obtain a first ship training sample;
dividing the first ship training sample into m x n grids to obtain a target training sample, wherein m is smaller than n;
and training according to the target ship training sample to obtain a ship detection model.
2. The method of claim 1, wherein said extracting the vessel training samples to obtain a first vessel training sample comprises:
extracting images of the ship training samples with the influence of sea waves;
and marking the image with the influence of the sea waves to obtain a first ship training sample.
3. The method of claim 2, wherein said dividing said first vessel training sample into m x n grids, resulting in a target training sample, m being less than n, comprises:
performing convolution operation on the first ship training sample according to a first preset convolution core to obtain a first target training sample;
and performing dimensionality reduction processing on the first target training sample according to a second preset convolution kernel to obtain a target training sample.
4. The method of claim 1, wherein after the training of the vessel detection model from the target vessel training samples, the method further comprises:
acquiring ship image test samples, wherein the ship image test samples comprise a positive test sample containing a ship and a negative test sample containing no ship;
obtaining a test result according to the ship image test sample and the ship detection model;
and when the test result meets a preset condition, determining the ship detection model as an available ship detection model.
5. A method of ship inspection, comprising the steps of:
acquiring a ship detection image, wherein the ship detection image comprises a ship and sea waves;
inputting the ship detection image into the ship detection model to obtain a ship detection result; the ship detection model is generated by training through the ship detection model training method according to any one of claims 1 to 4.
6. The method of claim 5, wherein after said acquiring a ship inspection image, the method further comprises:
segmenting the ship detection image into a grid of m x n, m being smaller than n.
7. A ship detection model training device is characterized by comprising:
the system comprises a first acquisition module, a second acquisition module and a third acquisition module, wherein the first acquisition module is used for acquiring a ship training sample, and the ship training sample comprises a plurality of sea surface environment images;
the first extraction module is used for extracting the ship training sample to obtain a first ship training sample;
the segmentation module is used for segmenting the first ship training sample into m x n grids to obtain a target training sample, wherein m is smaller than n;
and the training module is used for training according to the target ship training sample to obtain a ship detection model.
8. A watercraft detection device, comprising:
the second acquisition module is used for acquiring a ship detection image;
the detection module is used for inputting the ship detection image into the ship detection model to obtain a ship detection result; the ship detection model is generated by training through the ship detection model training method according to any one of claims 1 to 4.
9. A computer device, comprising: at least one processor; and a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the at least one processor to cause the at least one processor to perform the vessel inspection model training method of any one of claims 1 to 4 or the vessel inspection method of any one of claims 5 to 6.
10. A computer readable storage medium having stored thereon computer instructions, which when executed by a processor, implement the ship inspection model training method of any one of claims 1 to 4 or the ship inspection method of any one of claims 5 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911377878.6A CN111553183A (en) | 2019-12-27 | 2019-12-27 | Ship detection model training method, ship detection method and ship detection device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911377878.6A CN111553183A (en) | 2019-12-27 | 2019-12-27 | Ship detection model training method, ship detection method and ship detection device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111553183A true CN111553183A (en) | 2020-08-18 |
Family
ID=72005609
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911377878.6A Pending CN111553183A (en) | 2019-12-27 | 2019-12-27 | Ship detection model training method, ship detection method and ship detection device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111553183A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112183463A (en) * | 2020-10-23 | 2021-01-05 | 珠海大横琴科技发展有限公司 | Ship identification model verification method and device based on radar image |
| CN112329613A (en) * | 2020-11-03 | 2021-02-05 | 珠海大横琴科技发展有限公司 | Sea wave influence launching detection method and device, electronic equipment and storage medium |
| CN116524405A (en) * | 2023-05-04 | 2023-08-01 | 广东海洋大学 | Ocean storm wave height identification method and system |
-
2019
- 2019-12-27 CN CN201911377878.6A patent/CN111553183A/en active Pending
Non-Patent Citations (2)
| Title |
|---|
| 吴科君: "基于深度学习的海面船舶目标检测", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
| 王喜宾 等著: "《基于优化支持向量机的个性化推荐研究》", 30 June 2017, 重庆大学出版社 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112183463A (en) * | 2020-10-23 | 2021-01-05 | 珠海大横琴科技发展有限公司 | Ship identification model verification method and device based on radar image |
| CN112183463B (en) * | 2020-10-23 | 2021-10-15 | 珠海大横琴科技发展有限公司 | Ship identification model verification method and device based on radar image |
| CN112329613A (en) * | 2020-11-03 | 2021-02-05 | 珠海大横琴科技发展有限公司 | Sea wave influence launching detection method and device, electronic equipment and storage medium |
| CN116524405A (en) * | 2023-05-04 | 2023-08-01 | 广东海洋大学 | Ocean storm wave height identification method and system |
| CN116524405B (en) * | 2023-05-04 | 2024-02-23 | 广东海洋大学 | Ocean storm wave height identification method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110310264B (en) | DCNN-based large-scale target detection method and device | |
| CN110569899A (en) | Dam face defect classification model training method and device | |
| CN111553183A (en) | Ship detection model training method, ship detection method and ship detection device | |
| CN110796048A (en) | Ship target real-time detection method based on deep neural network | |
| CN111435446A (en) | License plate identification method and device based on L eNet | |
| CN112329881A (en) | License plate recognition model training method, license plate recognition method and device | |
| CN110969213A (en) | Ship detection method and device based on fast RCNN and electronic equipment | |
| CN111144425B (en) | Method and device for detecting shot screen picture, electronic equipment and storage medium | |
| CN111444911B (en) | Training method and device of license plate recognition model and license plate recognition method and device | |
| CN111695397A (en) | Ship identification method based on YOLO and electronic equipment | |
| CN111553184A (en) | Small target detection method and device based on electronic purse net and electronic equipment | |
| CN111435445A (en) | Training method and device of character recognition model and character recognition method and device | |
| CN111695572A (en) | Ship retrieval method and device based on convolutional layer feature extraction | |
| CN116596875B (en) | Wafer defect detection method and device, electronic equipment and storage medium | |
| CN115953746B (en) | Ship monitoring method and device | |
| CN116485779B (en) | Adaptive wafer defect detection method and device, electronic equipment and storage medium | |
| CN112785595B (en) | Target attribute detection, neural network training and intelligent driving method and device | |
| CN110555344B (en) | Lane line recognition method, lane line recognition device, electronic device, and storage medium | |
| CN111667450A (en) | Ship quantity counting method and device and electronic equipment | |
| CN111122582B (en) | Poultry egg detection method, image processing method and device | |
| CN111950527A (en) | Target detection method and device based on YOLO V2 neural network | |
| KR101723028B1 (en) | Image processing system for integrated management of image information changing in real time | |
| CN113139077B (en) | Method, device, terminal and storage medium for identifying ship identity | |
| CN117197048B (en) | Ship water gauge reading detection method, device and equipment | |
| CN111625672B (en) | Image processing method, image processing device, computer equipment and storage medium |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200818 |