CN108121968A - A kind of fish monitoring method - Google Patents
A kind of fish monitoring method Download PDFInfo
- Publication number
- CN108121968A CN108121968A CN201711400429.XA CN201711400429A CN108121968A CN 108121968 A CN108121968 A CN 108121968A CN 201711400429 A CN201711400429 A CN 201711400429A CN 108121968 A CN108121968 A CN 108121968A
- Authority
- CN
- China
- Prior art keywords
- fish
- pictures
- sub
- abnormal
- 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.)
- Granted
Links
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
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/51—Indexing; Data structures therefor; Storage structures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Multimedia (AREA)
- Software Systems (AREA)
- Databases & Information Systems (AREA)
- Image Analysis (AREA)
- Farming Of Fish And Shellfish (AREA)
Abstract
The invention discloses a kind of fish monitoring method, this method includes：（1）Obtain the picture of abnormal fish, classification rower of going forward side by side draws；（2）It obtains training set according to the abnormal image of position classification；（3）The neutral net of deep learning is obtained according to abnormal image training set；（4）The underwater shoal of fish of video identification；（5）The shoal of fish is salvaged according to recognition result.The present invention provides a kind of fish monitoring methods, the judgement of fish unusual condition is realized by big data and video acquisition, and the information of specific abnormal fish appearance is obtained according to the GPS, hydraulic pressure, time of acquisition, so that poultry feeders can salvage abnormal fish to come in time, the pollution to waters is avoided, improves yield.
Description
Technical field
The present invention relates to a kind of method breeded fish, especially a kind of fish monitoring method.
Background technology
Fish is the food that people like eating.In order to meet it is increasingly increased to fish the needs of, it is necessary to largely cultivate fish.And
Turning out the fish that can be sold just needs to pay attention to the physical condition of fish constantly.
The method of existing observation fish physical condition is manually to go to salvage a part of sample in the set time.However, so
It can not timely problem present in electric fish.Moreover, even if the sample salvaged has no problem, it cannot guarantee that and be not present
Abnormal fish.No matter ignore if letting alone abnormal fish for a long time in waters, it will so that abnormal fish polluted water region, its own is deposited
The problem of be transmitted to other healthy fishes, so as to cause the substantial amounts of unnatural death of fish, greatly reduce the survival rate of fish
And growth, cause huge economic loss.
The content of the invention
Therefore, in view of the above-mentioned problems, the present invention provides a kind of fish monitoring methods, big data and video acquisition are passed through
It realizes the judgement of fish unusual condition, and the information of specific abnormal fish appearance is obtained according to the GPS, hydraulic pressure, time of acquisition, and then make
Abnormal fish can be salvaged to come in time by obtaining poultry feeders, avoided the pollution to waters, improved yield.
In order to achieve the above object, the present invention proposes a kind of fish monitoring method, which is characterized in that this method includes：
(1) picture of abnormal fish is obtained, classification rower of going forward side by side draws；
Search and the relevant plurality of pictures of abnormal fish, will be not belonging to the part of abnormal fish in itself and buckle from internet in picture
Except the sub-pictures for obtaining only representing abnormal fish image itself, the exception fish is that fish surface has fish that is red and swollen, festering, change colour,
And every sub-pictures are marked into row position according to the abnormal position of the fish that every sub-pictures are shown, i.e., the sub-pictures show this
The abnormal position of abnormal fish is any in head, body, afterbody；
(2) obtain training set according to the abnormal image of position classification；
Search is labeled as all sub-pictures on head, forms head abnormal image training set；Search is labeled as body
All sub-pictures form body abnormality training set of images and close；Search is labeled as all sub-pictures of afterbody, forms afterbody Abnormal Map
As training set；
(3) neutral net of deep learning is obtained according to abnormal image training set；
Each sub-pictures in head abnormal image training set are resolved into according to certain tactic multiple portions
Divide the subgraph overlapped, each subgraph is input in a nervelet network, the output of the nervelet network is stored in
In one small ordered series of numbers, all small ordered series of numbers centainly are ranked sequentially to obtain a big ordered series of numbers according to described, which is inputted
Into one big neutral net, and then obtain to judge that fish whether there is the first nerves network of head exception according to image；
Each sub-pictures during body abnormality training set of images is closed are resolved into according to certain tactic multiple portions
Divide the subgraph overlapped, each subgraph is input in a nervelet network, the output of the nervelet network is stored in
In one small ordered series of numbers, all small ordered series of numbers centainly are ranked sequentially to obtain a big ordered series of numbers according to described, which is inputted
Into one big neutral net, and then obtain to judge that fish whether there is the nervus opticus network of body abnormality according to image；
Each sub-pictures in afterbody abnormal image training set are resolved into according to certain tactic multiple portions
Divide the subgraph overlapped, each subgraph is input in a nervelet network, the output of the nervelet network is stored in
In one small ordered series of numbers, all small ordered series of numbers centainly are ranked sequentially to obtain a big ordered series of numbers according to described, which is inputted
Into one big neutral net, and then obtain to judge that fish whether there is the third nerve network of afterbody exception according to image；
(4) the underwater shoal of fish of video identification；
Underwater environment is scanned using the infrared thermography that can voluntarily move under water, in infrared thermography is imaged extremely
Less there are during the shape of fish, shot using the direction of camera alignment infrared thermography scanning, until infrared heat
When no longer there is the shape of fish in imager imaging, camera stops shooting, obtains one section of video, and is shot in camera
During mark is carried out to video using GPS, hydraulic pressure, time shaft always, i.e., each frame video is corresponding with the frame video capture
When time, the GPS location of two-dimensional environment residing for camera when showing to shoot the frame video shows to image when shooting the frame video
The hydraulic pressure of depth of water situation, land server is transmitted to by the video residing for head；Land server is by video in the way of every frame
Multiple pictures marked with GPS, hydraulic pressure, time are split into, and several partly overlapping subgraphs will be divided into per pictures
All sub-pictures are input to first nerves network by piece, if first nerves network judges that a certain picture includes head exception fish,
Then export this with GPS, hydraulic pressure, the time mark picture and mark head exception, also by all sub-pictures be input to second god
Through network, if nervus opticus network judges that a certain picture includes body abnormality fish, output should be with GPS, hydraulic pressure, time mark
The picture of note simultaneously marks body abnormality, and all sub-pictures also are input to third nerve network, if third nerve network judges
A certain picture include afterbody exception fish, then export this with GPS, hydraulic pressure, the time mark picture and mark afterbody exception；
(5) shoal of fish is salvaged according to recognition result；
Obtained mark head exception, body abnormality, the picture of afterbody exception are compared, if there is three identical pictures
Then judge to need manual intervention, according to the GPS, hydraulic pressure, time of three identical picture indicias, the shoal of fish is salvaged.
Specific embodiment
Embodiment one.
A kind of fish monitoring method, which is characterized in that this method includes：
(1) picture of abnormal fish is obtained, classification rower of going forward side by side draws；
Search and the relevant plurality of pictures of abnormal fish, will be not belonging to the part of abnormal fish in itself and buckle from internet in picture
Except the sub-pictures for obtaining only representing abnormal fish image itself, the exception fish is that fish surface has fish that is red and swollen, festering, change colour,
And every sub-pictures are marked into row position according to the abnormal position of the fish that every sub-pictures are shown, i.e., the sub-pictures show this
The abnormal position of abnormal fish is any in head, body, afterbody；
(2) obtain training set according to the abnormal image of position classification；
Search is labeled as all sub-pictures on head, forms head abnormal image training set；Search is labeled as body
All sub-pictures form body abnormality training set of images and close；Search is labeled as all sub-pictures of afterbody, forms afterbody Abnormal Map
As training set；
(3) neutral net of deep learning is obtained according to abnormal image training set；
Each sub-pictures in head abnormal image training set are resolved into according to certain tactic multiple portions
Divide the subgraph overlapped, each subgraph is input in a nervelet network, the output of the nervelet network is stored in
In one small ordered series of numbers, all small ordered series of numbers centainly are ranked sequentially to obtain a big ordered series of numbers according to described, which is inputted
Into one big neutral net, and then obtain to judge that fish whether there is the first nerves network of head exception according to image；
Each sub-pictures during body abnormality training set of images is closed are resolved into according to certain tactic multiple portions
Divide the subgraph overlapped, each subgraph is input in a nervelet network, the output of the nervelet network is stored in
In one small ordered series of numbers, all small ordered series of numbers centainly are ranked sequentially to obtain a big ordered series of numbers according to described, which is inputted
Into one big neutral net, and then obtain to judge that fish whether there is the nervus opticus network of body abnormality according to image；
Each sub-pictures in afterbody abnormal image training set are resolved into according to certain tactic multiple portions
Divide the subgraph overlapped, each subgraph is input in a nervelet network, the output of the nervelet network is stored in
In one small ordered series of numbers, all small ordered series of numbers centainly are ranked sequentially to obtain a big ordered series of numbers according to described, which is inputted
Into one big neutral net, and then obtain to judge that fish whether there is the third nerve network of afterbody exception according to image；
(4) the underwater shoal of fish of video identification；
Underwater environment is scanned using the infrared thermography that can voluntarily move under water, in infrared thermography is imaged extremely
Less there are during the shape of fish, shot using the direction of camera alignment infrared thermography scanning, until infrared heat
When no longer there is the shape of fish in imager imaging, camera stops shooting, obtains one section of video, and is shot in camera
During mark is carried out to video using GPS, hydraulic pressure, time shaft always, i.e., each frame video is corresponding with the frame video capture
When time, the GPS location of two-dimensional environment residing for camera when showing to shoot the frame video shows to image when shooting the frame video
The hydraulic pressure of depth of water situation, land server is transmitted to by the video residing for head；Land server is by video in the way of every frame
Multiple pictures marked with GPS, hydraulic pressure, time are split into, and several partly overlapping subgraphs will be divided into per pictures
All sub-pictures are input to first nerves network by piece, if first nerves network judges that a certain picture includes head exception fish,
Then export this with GPS, hydraulic pressure, the time mark picture and mark head exception, also by all sub-pictures be input to second god
Through network, if nervus opticus network judges that a certain picture includes body abnormality fish, output should be with GPS, hydraulic pressure, time mark
The picture of note simultaneously marks body abnormality, and all sub-pictures also are input to third nerve network, if third nerve network judges
A certain picture include afterbody exception fish, then export this with GPS, hydraulic pressure, the time mark picture and mark afterbody exception；
(5) shoal of fish is salvaged according to recognition result；
Obtained mark head exception, body abnormality, the picture of afterbody exception are compared, if there is three identical pictures
Then judge to need manual intervention, according to the GPS, hydraulic pressure, time of three identical picture indicias, the shoal of fish is salvaged.
Camera is high definition or super clear camera.
Embodiment two.
A kind of fish monitoring method, which is characterized in that this method includes：
(1) picture of abnormal fish is obtained, classification rower of going forward side by side draws；
Search and the relevant plurality of pictures of abnormal fish, will be not belonging to the part of abnormal fish in itself and buckle from internet in picture
Except the sub-pictures for obtaining only representing abnormal fish image itself, the exception fish is that fish surface has redness, festers, changes colour, spot
Deng fish, and every sub-pictures are marked into row position according to the abnormal position of the fish that every sub-pictures are shown, i.e. the subgraph
Piece shows that the abnormal position of the exception fish is any in head, body, afterbody；
If there are problems in itself for fish, it will usually be embodied on its body surface, therefore, by being to the judgement on fish surface
It is known that fish is with the presence or absence of abnormal, and usually fish there are the problem of it is larger when, just have more spot on surface, burst
It is rotten etc., therefore, situations such as whether existing simultaneously discoloration by the multiple positions for judging fish surface, it is possible to know the abnormal shape of fish
Whether emergency is processed state；
(2) obtain training set according to the abnormal image of position classification；
Search is labeled as all sub-pictures on head, forms head abnormal image training set；Search is labeled as body
All sub-pictures form body abnormality training set of images and close；Search is labeled as all sub-pictures of afterbody, forms afterbody Abnormal Map
As training set；
(3) neutral net of deep learning is obtained according to abnormal image training set；
Each sub-pictures in head abnormal image training set are resolved into according to certain tactic multiple portions
Divide the subgraph overlapped, each subgraph is input in a nervelet network, the output of the nervelet network is stored in
In one small ordered series of numbers, all small ordered series of numbers centainly are ranked sequentially to obtain a big ordered series of numbers according to described, which is inputted
Into one big neutral net, and then obtain to judge that fish whether there is the first nerves network of head exception according to image；
Each sub-pictures during body abnormality training set of images is closed are resolved into according to certain tactic multiple portions
Divide the subgraph overlapped, each subgraph is input in a nervelet network, the output of the nervelet network is stored in
In one small ordered series of numbers, all small ordered series of numbers centainly are ranked sequentially to obtain a big ordered series of numbers according to described, which is inputted
Into one big neutral net, and then obtain to judge that fish whether there is the nervus opticus network of body abnormality according to image；
Each sub-pictures in afterbody abnormal image training set are resolved into according to certain tactic multiple portions
Divide the subgraph overlapped, each subgraph is input in a nervelet network, the output of the nervelet network is stored in
In one small ordered series of numbers, all small ordered series of numbers centainly are ranked sequentially to obtain a big ordered series of numbers according to described, which is inputted
Into one big neutral net, and then obtain to judge that fish whether there is the third nerve network of afterbody exception according to image；
Using deep learning neutral net, the efficiency and quality of study can be improved, so improve identification accuracy and
Speed；
(4) the underwater shoal of fish of video identification；
Underwater environment is scanned using the infrared thermography that can voluntarily move under water, in infrared thermography is imaged extremely
Less there are during the shape of fish, shot using the direction of camera alignment infrared thermography scanning, until infrared heat
When no longer there is the shape of fish in imager imaging, camera stops shooting, obtains one section of video, and is shot in camera
During mark is carried out to video using GPS, hydraulic pressure, time shaft always, i.e., each frame video is corresponding with the frame video capture
When time, the GPS location of two-dimensional environment residing for camera when showing to shoot the frame video shows to image when shooting the frame video
The hydraulic pressure of depth of water situation, land server is transmitted to by the video residing for head；Land server is by video in the way of every frame
Multiple pictures marked with GPS, hydraulic pressure, time are split into, and several partly overlapping subgraphs will be divided into per pictures
All sub-pictures are input to first nerves network by piece, if first nerves network judges that a certain picture includes head exception fish,
Then export this with GPS, hydraulic pressure, the time mark picture and mark head exception, also by all sub-pictures be input to second god
Through network, if nervus opticus network judges that a certain picture includes body abnormality fish, output should be with GPS, hydraulic pressure, time mark
The picture of note simultaneously marks body abnormality, and all sub-pictures also are input to third nerve network, if third nerve network judges
A certain picture include afterbody exception fish, then export this with GPS, hydraulic pressure, the time mark picture and mark afterbody exception；
In view of infrared imaging instrument and the working environment of camera, the two is required to be waterproof pressure-resistant equipment, and uses
Accumulator provides electric power for the two；Infrared imaging instrument and camera can be arranged on a movable termination, moved
Terminal realizes acting on one's own in water using slurry；Since underwater equipment needs storage battery power supply, and the electricity of accumulator has
Limit, it is therefore desirable to reduction unnecessary electricity waste as far as possible, so only there are the shapes of fish in the imaging of infrared imaging instrument
Just so that camera realizes shooting, and then the image of capture fish during shape；Due to the activity of fish be it is regular, together
When also record GPS, hydraulic pressure, time data be for convenience manually at the same time, the place fish that can will have abnormality
Capture；
(5) shoal of fish is salvaged according to recognition result；
Obtained mark head exception, body abnormality, the picture of afterbody exception are compared, if there is three identical pictures
Then judge to need manual intervention, according to the GPS, hydraulic pressure, time of three identical picture indicias, the shoal of fish is salvaged；
If head, body, afterbody have exception, illustrate that the abnormal conditions of the exception fish are very serious, it is necessary to urgent
It is handled, is otherwise likely to polluted water region and other fishes, so artificial right place is needed to be caught up.
It is it should be noted that made for the present invention further specifically the above content is specific embodiment is combined
It is bright, it is impossible to assert that the specific embodiment of the present invention is only limitted to this, under the above-mentioned guidance of the present invention, those skilled in the art can
To carry out various improvement and deformation on the basis of above-described embodiment, and these are improved or deformation falls in protection model of the invention
In enclosing.
Claims (2)
- A kind of 1. fish monitoring method, which is characterized in that this method includes：（1）Obtain the picture of abnormal fish, classification rower of going forward side by side draws；Search and the relevant plurality of pictures of abnormal fish, will be not belonging to the part of abnormal fish in itself and deduct from internet in picture To the sub-pictures for only representing abnormal fish image itself, the exception fish is that fish surface has fish that is red and swollen, festering, change colour, and root The abnormal position of the fish that is shown according to every sub-pictures marks every sub-pictures into row position, i.e., the sub-pictures show the exception The abnormal position of fish is any in head, body, afterbody；（2）It obtains training set according to the abnormal image of position classification；Search is labeled as all sub-pictures on head, forms head abnormal image training set；Search is labeled as all of body Sub-pictures form body abnormality training set of images and close；Search is labeled as all sub-pictures of afterbody, forms afterbody abnormal image instruction Practice set；（3）The neutral net of deep learning is obtained according to abnormal image training set；Each sub-pictures in head abnormal image training set are resolved into according to certain tactic multiple portions weight The subgraph of conjunction, each subgraph is input in a nervelet network, and the output of the nervelet network is stored in one In small ordered series of numbers, all small ordered series of numbers centainly are ranked sequentially to obtain a big ordered series of numbers according to described, which is input to one In a big neutral net, and then obtain to judge that fish whether there is the first nerves network of head exception according to image；Each sub-pictures during body abnormality training set of images is closed are resolved into according to certain tactic multiple portions weight The subgraph of conjunction, each subgraph is input in a nervelet network, and the output of the nervelet network is stored in one In small ordered series of numbers, all small ordered series of numbers centainly are ranked sequentially to obtain a big ordered series of numbers according to described, which is input to one In a big neutral net, and then obtain to judge that fish whether there is the nervus opticus network of body abnormality according to image；Each sub-pictures in afterbody abnormal image training set are resolved into according to certain tactic multiple portions weight The subgraph of conjunction, each subgraph is input in a nervelet network, and the output of the nervelet network is stored in one In small ordered series of numbers, all small ordered series of numbers centainly are ranked sequentially to obtain a big ordered series of numbers according to described, which is input to one In a big neutral net, and then obtain to judge that fish whether there is the third nerve network of afterbody exception according to image；（4）The underwater shoal of fish of video identification；Underwater environment is scanned using the infrared thermography that can voluntarily move under water, is at least deposited when in infrared thermography imaging In the shape of fish, shot using the direction of camera alignment infrared thermography scanning, until infrared thermal imaging When no longer there is the shape of fish in instrument imaging, camera stops shooting, obtains one section of video, and the phase shot in camera Between always mark carries out video using GPS, hydraulic pressure, time shaft, i.e., when each frame video is corresponding with the frame video capture Time, the GPS location of two-dimensional environment residing for camera when showing to shoot the frame video, camera institute when showing to shoot the frame video Locate the hydraulic pressure of depth of water situation, which is transmitted to land server；Land server splits video in the way of every frame Into multiple pictures marked with GPS, hydraulic pressure, time, and several partly overlapping sub-pictures will be divided into per pictures, it will All sub-pictures are input to first nerves network, defeated if first nerves network judges that a certain picture includes head exception fish Go out this with GPS, hydraulic pressure, the time mark picture and mark head exception, all sub-pictures are also input to nervus opticus net Network if nervus opticus network judges that a certain picture includes body abnormality fish, exports this and is marked with GPS, hydraulic pressure, time Picture simultaneously marks body abnormality, and all sub-pictures also are input to third nerve network, if third nerve network judge it is a certain Picture include afterbody exception fish, then export this with GPS, hydraulic pressure, the time mark picture and mark afterbody exception；（5）The shoal of fish is salvaged according to recognition result；Obtained mark head exception, body abnormality, the picture of afterbody exception are compared, is then sentenced if there is three identical pictures Surely manual intervention is needed, according to the GPS, hydraulic pressure, time of three identical picture indicias, the shoal of fish is salvaged.
- 2. fish monitoring method according to claim 2, which is characterized in that camera is high-definition camera.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711400429.XA CN108121968B (en) | 2017-12-20 | 2017-12-20 | Fish shoal monitoring method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711400429.XA CN108121968B (en) | 2017-12-20 | 2017-12-20 | Fish shoal monitoring method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108121968A true CN108121968A (en) | 2018-06-05 |
| CN108121968B CN108121968B (en) | 2021-11-12 |
Family
ID=62231305
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711400429.XA Active CN108121968B (en) | 2017-12-20 | 2017-12-20 | Fish shoal monitoring method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108121968B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109856138A (en) * | 2018-12-18 | 2019-06-07 | 杭州电子科技大学 | Deep sea net cage healthy fish identifying system and method based on deep learning |
| CN112446872A (en) * | 2020-12-07 | 2021-03-05 | 吉首大学 | Giant salamander abnormal behavior identification method based on image identification |
| CN112465778A (en) * | 2020-11-26 | 2021-03-09 | 江苏国和智能科技有限公司 | Underwater fish shoal observation device and method |
| CN112634202A (en) * | 2020-12-04 | 2021-04-09 | 浙江省农业科学院 | Method, device and system for detecting behavior of polyculture fish shoal based on YOLOv3-Lite |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101614819A (en) * | 2009-08-04 | 2009-12-30 | 北京师范大学 | Submerged plants in shallow water lake automatic identification technology and device |
| CN103678910A (en) * | 2013-12-12 | 2014-03-26 | 河海大学 | Cloud system structure pre-warning system and method for riverway type reservoir tributary bay water bloom |
| CN106022459A (en) * | 2016-05-23 | 2016-10-12 | 三峡大学 | Automatic counting system for fish passing amount of fish passage based on underwater videos |
| CN106550223A (en) * | 2017-01-13 | 2017-03-29 | 湖南理工学院 | A kind of dead fish supervising device and monitoring method for aquaculture |
| CN106908038A (en) * | 2017-01-04 | 2017-06-30 | 成都通甲优博科技有限责任公司 | A kind of monitoring device and monitoring system based on fish eye lens video camera |
| CN106960175A (en) * | 2017-02-21 | 2017-07-18 | 华南理工大学 | The first visual angle dynamic gesture detection method based on depth convolutional neural networks |
| CN107451512A (en) * | 2016-06-01 | 2017-12-08 | 海南热带海洋学院 | A kind of algae pollution detection method of integrated wireless communications, video image processing technology |
-
2017
- 2017-12-20 CN CN201711400429.XA patent/CN108121968B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101614819A (en) * | 2009-08-04 | 2009-12-30 | 北京师范大学 | Submerged plants in shallow water lake automatic identification technology and device |
| CN103678910A (en) * | 2013-12-12 | 2014-03-26 | 河海大学 | Cloud system structure pre-warning system and method for riverway type reservoir tributary bay water bloom |
| CN106022459A (en) * | 2016-05-23 | 2016-10-12 | 三峡大学 | Automatic counting system for fish passing amount of fish passage based on underwater videos |
| CN107451512A (en) * | 2016-06-01 | 2017-12-08 | 海南热带海洋学院 | A kind of algae pollution detection method of integrated wireless communications, video image processing technology |
| CN106908038A (en) * | 2017-01-04 | 2017-06-30 | 成都通甲优博科技有限责任公司 | A kind of monitoring device and monitoring system based on fish eye lens video camera |
| CN106550223A (en) * | 2017-01-13 | 2017-03-29 | 湖南理工学院 | A kind of dead fish supervising device and monitoring method for aquaculture |
| CN106960175A (en) * | 2017-02-21 | 2017-07-18 | 华南理工大学 | The first visual angle dynamic gesture detection method based on depth convolutional neural networks |
Non-Patent Citations (2)
| Title |
|---|
| R. PRADOS等: ""Real-time fish detection in trawl nets"", 《OCEANS 2017 - ABERDEEN》 * |
| 傅小康等: ""基于鱼体回声图像的鱼类识别算法"", 《测试技术学报》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109856138A (en) * | 2018-12-18 | 2019-06-07 | 杭州电子科技大学 | Deep sea net cage healthy fish identifying system and method based on deep learning |
| CN112465778A (en) * | 2020-11-26 | 2021-03-09 | 江苏国和智能科技有限公司 | Underwater fish shoal observation device and method |
| CN112634202A (en) * | 2020-12-04 | 2021-04-09 | 浙江省农业科学院 | Method, device and system for detecting behavior of polyculture fish shoal based on YOLOv3-Lite |
| CN112446872A (en) * | 2020-12-07 | 2021-03-05 | 吉首大学 | Giant salamander abnormal behavior identification method based on image identification |
| CN112446872B (en) * | 2020-12-07 | 2022-09-27 | 吉首大学 | Giant salamander abnormal behavior identification method based on image identification |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108121968B (en) | 2021-11-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108121968A (en) | A kind of fish monitoring method | |
| CN109591982B (en) | Unmanned monitoring underwater aquaculture robot | |
| CN110200598A (en) | A kind of large-scale plant that raises sign exception birds detection system and detection method | |
| CN110414396A (en) | A kind of unmanned boat perception blending algorithm based on deep learning | |
| CN112506120A (en) | Wisdom fishery management system based on thing networking | |
| CN104867158B (en) | Indoor above water craft Precise Position System and method based on monocular vision | |
| CN109613559A (en) | The flood boundaries floating material discriminating gear and method of view-based access control model and laser radar | |
| CN111487245B (en) | Coral reef-like water area biological quality evolution evaluation system | |
| CN110120073B (en) | Method for guiding recovery of unmanned ship based on lamp beacon visual signal | |
| CN106096617A (en) | A kind of based on the panorama sea horizon extracting method improving active contour model | |
| CN111833331A (en) | Livestock body temperature image processing method based on artificial intelligence and active gun-ball linkage | |
| CN105844264B (en) | It is a kind of based on the recognition methods of tree peony fruit image of the oil of stress | |
| CN208187400U (en) | A kind of long measuring device of Fish based on machine vision | |
| CN111428677B (en) | Method for automatically monitoring water surface circular swimming state of fish cultivated in offshore net cage by unmanned aerial vehicle | |
| KR102579082B1 (en) | METHOD FOR DETECTING A SARGASSUM USING MULTISPECTRAL IMAGES LOADED ON MANNED and UNMANNED AERIAL VEHICLES AND SYSTEM THEREOF | |
| Burns | Population characteristics and migratory movements of humpback whales (Megaptera novaeangliae) identified on their southern migration past Ballina, eastern Australia | |
| Sherlock et al. | Near-field zooplankton, ice-face biota and proximal hydrography of free-drifting Antarctic icebergs | |
| CN207423108U (en) | Autonomous type sea cucumber finishing device visual detection positioning device | |
| CN109625218B (en) | Unmanned monitoring underwater aquaculture robot system based on solar charging | |
| CN115437394A (en) | Unmanned aerial vehicle system for river channel monitoring and autonomous line patrol method thereof | |
| Rezo et al. | Collecting information for biomass estimation in mariculture with a heterogeneous robotic system | |
| Jones et al. | Enhancing UAV images to improve the observation of submerged whales using a water column correction method | |
| Tijani et al. | Fishing forecasting system in Adriatic sea—A model approach based on a normalized scalar product of the SST gradient and CHL gradient vectors | |
| CN105230581B (en) | A kind of fishing feelings detection device based on seabird image | |
| Mitchell et al. | Queensland SharkSmart Drone Trial Final Report |
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 | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20211026 Address after: 221000 Room 101, building a, Xinyi Internet Industrial Park, Xuzhou City, Jiangsu Province Applicant after: Jiangsu Wharton Digital Technology Co.,Ltd. Address before: 528000 room 2903, block 1, No.32 Zhaoxiang Road, Chancheng District, Foshan City, Guangdong Province Applicant before: FOSHAN LUOKEWEITE TECHNOLOGY Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |