CN113191273A - Oil field well site video target detection and identification method and system based on neural network - Google Patents
Oil field well site video target detection and identification method and system based on neural network Download PDFInfo
- Publication number
- CN113191273A CN113191273A CN202110485534.8A CN202110485534A CN113191273A CN 113191273 A CN113191273 A CN 113191273A CN 202110485534 A CN202110485534 A CN 202110485534A CN 113191273 A CN113191273 A CN 113191273A
- Authority
- CN
- China
- Prior art keywords
- target
- neural network
- image
- oil field
- well site
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 58
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 47
- 238000000034 method Methods 0.000 title claims abstract description 25
- 238000007781 pre-processing Methods 0.000 claims abstract description 12
- 238000000605 extraction Methods 0.000 claims abstract description 10
- 238000012549 training Methods 0.000 claims description 30
- 230000006870 function Effects 0.000 claims description 16
- 238000010586 diagram Methods 0.000 claims description 11
- 238000012360 testing method Methods 0.000 claims description 10
- 238000003062 neural network model Methods 0.000 claims description 8
- 241001465754 Metazoa Species 0.000 claims description 7
- 238000005516 engineering process Methods 0.000 claims description 7
- 238000005086 pumping Methods 0.000 claims description 5
- 238000013434 data augmentation Methods 0.000 claims description 3
- 238000005286 illumination Methods 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 230000007935 neutral effect Effects 0.000 claims description 2
- 238000004519 manufacturing process Methods 0.000 abstract description 11
- 238000012544 monitoring process Methods 0.000 description 8
- 238000004590 computer program Methods 0.000 description 7
- 238000012545 processing Methods 0.000 description 7
- 230000000694 effects Effects 0.000 description 3
- 239000000284 extract Substances 0.000 description 3
- 238000012163 sequencing technique Methods 0.000 description 3
- 238000003860 storage Methods 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 239000003086 colorant Substances 0.000 description 2
- 239000003208 petroleum Substances 0.000 description 2
- 239000000779 smoke Substances 0.000 description 2
- 241000282472 Canis lupus familiaris Species 0.000 description 1
- 241001494479 Pecora Species 0.000 description 1
- 230000003321 amplification Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 238000003199 nucleic acid amplification method Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Images
Classifications
-
- 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
- G06V20/46—Extracting features or characteristics from the video content, e.g. video fingerprints, representative shots or key frames
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/07—Target detection
Abstract
The invention discloses a neural network-based oilfield well site video target detection and identification method and system, wherein the method comprises the following steps: preprocessing an oil field scene image formed by frame extraction of an oil field well site video; inputting the preprocessed image into a pre-trained deep neural network oilfield well site target detection and recognition device for recognition, and obtaining target position and target category information appearing in the image; and matching the obtained target position and target category information with the pre-established oil field alarm grade information, and outputting an identification result. Aiming at diversified scenes of an oil field well site field, the invention utilizes a self-defined two-stage depth neural network structure, can improve the identification accuracy, realizes the rapid and effective detection and identification of multiple targets in the oil field scene, and brings safety guarantee for oil field production and transportation.
Description
Technical Field
The invention belongs to the technical field of computer digital image processing and pattern recognition, relates to the field of target detection and recognition of an oil field well site operation area, and particularly relates to a method and a system for detecting and recognizing an oil field well site video target based on a neural network.
Background
Most oil fields are located in areas with inconvenient field traffic and severe natural environment, and meanwhile, the problems of theft and damage are also solved, so that the safety of petroleum production and transportation is seriously threatened. An important component of oilfield exploitation is safety control. Safety work includes safety issues with the complex nature of the environment, oil production and transportation safety issues, and well site operator safety.
The most important technology in security monitoring at present is to adopt moving frame detection, also commonly called motion detection, which is commonly used for unattended monitoring video recording and automatic alarm. The camera collects images according to different frame rates, the obtained images are calculated and compared by the CPU according to a certain algorithm, when the pictures are changed, if people walk, the lens is moved and the like, the digital result obtained by calculation and comparison exceeds a threshold value, and the system is indicated to automatically perform corresponding processing. Aiming at complex scenes in an oil field factory area, the oil field factory area comprises complex targets such as an oil pumping unit, an engineering truck, a worker, an animal and a car.
At present, a general moving frame detection scheme cannot detect a specific target accurately and accurately, and is difficult to monitor accurately and implement purposeful management and control.
Disclosure of Invention
The invention aims to provide a neural network-based oilfield wellsite video target detection and identification method and system, so as to solve one or more technical problems. Aiming at diversified scenes of an oil field well site field, the invention utilizes a self-defined two-stage depth neural network structure, can improve the identification accuracy, realizes the rapid and effective detection and identification of multiple targets in the oil field scene, and brings safety guarantee for oil field production and transportation.
In order to achieve the purpose, the invention adopts the following technical scheme:
the invention discloses a neural network-based oilfield well site video target detection and identification method, which comprises the following steps of:
preprocessing an oil field scene image formed by frame extraction of an oil field well site video to obtain a preprocessed image;
inputting the preprocessed image into a pre-trained deep neural network oilfield well site target detection and recognition device for recognition, and obtaining target position and target category information appearing in the image; in the pre-trained deep neural network oilfield well site target detection and recognition device, the target detection and recognition network based on the deep neural network comprises:
the first-stage network model is used for carrying out target identification on a high-resolution image which is shot by an oil field well site and has the resolution ratio larger than a preset threshold value, and outputting position information and first category information of a target object in the image;
the second-level network model is used for inputting a preset concerned target object image and outputting second category information; the preset concerned target object image is obtained by removing a background target from a target image output by a first-level network model;
and matching the obtained target position and target category information with the pre-established oil field alarm grade information, and outputting an identification result.
The invention is further improved in that the training step of the pre-trained deep neural network oilfield wellsite target detection and recognizer specifically comprises the following steps:
inputting a learning sample set picture marked with a target position and category information label into a target detection and identification network based on a deep neural network for training, and adjusting target detection and identification network parameters based on the deep neural network based on a combined learning loss function of the target position and the category;
optimizing by using a random gradient descending mode to generate an oilfield well site target detection and recognizer, performing model test once every epoch iteration, and selecting a model with the best performance or the performance meeting the preset requirement as a trained deep neural network oilfield well site target detection and recognizer.
The invention is further improved by adopting a Yolo network as a first-stage neural network model.
In a further improvement of the present invention, the step of training the first-stage neural network model includes:
sending the learning sample set picture to a BackBone module of a first-level network model Yolo network to obtain a characteristic diagram of an input image;
and (3) sending the characteristic graph into a neutral network and a prediction network, designing a loss function according to the prediction label information and the marking information for training, and selecting a model parameter combination which enables the error of the test set to be minimum as a parameter of a first-stage network model through continuous training and testing.
A further improvement of the present invention is to employ the SSD model as the second level network model.
In a further improvement of the present invention, the training step of the second-level network model comprises:
screening the primary target detection and classification results output by the first-stage neural network model, removing background images and obtaining preset concerned target images;
and sending the preset concerned target object image into a second-level network model for training and prediction to obtain the trained second-level network model.
A further development of the invention is that the first category information comprises: human, car, animal; the second category information includes: red safety helmet, blue safety helmet, worker wearing working clothes, worker not wearing working clothes, car, tank car, engineering truck and oil pumping unit.
The invention has the further improvement that when the learning sample set picture marked with the target position and category information label is obtained, a data enhancement mode is adopted, and the data set amplification comprises the following steps:
carrying out geometric distortion, illumination distortion and image shielding on the single picture;
the collected images are combined by adopting a multi-image combination Data enhancement technology, and four training images are combined into one image according to a certain proportion by utilizing a Mosaic Data Augmentation technology.
The invention discloses an oilfield well site video target detection and identification system based on a neural network, which comprises:
the preprocessing module is used for preprocessing an oil field scene image formed by frame extraction of an oil field well site video to obtain a preprocessed image;
the recognition and classification module is used for inputting the preprocessed image into a pre-trained deep neural network oilfield well site target detection and recognizer for recognition to obtain the target position and the target category information appearing in the image;
and the alarm matching module is used for matching the obtained target position and target category information with pre-established oil field alarm grade information and outputting an identification result.
Compared with the prior art, the invention has the following beneficial effects:
aiming at well site image information without obvious structured labels, the invention can realize the detection and identification of different types of targets (vehicles, personnel, animals and the like) in an oil field scene by utilizing a self-defined deep neural network (including the deep neural network). According to the invention, based on the self-defined deep neural network (including the first network model and the second network model), the identification accuracy can be improved, the intelligent digital monitoring of the oil field is greatly improved, the personnel management and the rapid alarm of the safety event are facilitated, and the safety event in the production process of the oil field is reduced.
Aiming at diversified scenes of an oil field well site field, the invention performs data enhancement processing on the acquired data set, performs optimization of small target detection and identification by utilizing a self-defined two-stage deep neural network structure and considering the characteristic that an oil field camera extracts image information, can improve the identification accuracy, realizes rapid and effective detection and identification of multiple targets in the oil field scene, and provides safety guarantee for oil field production and transportation.
The invention adopts a deep neural network model to collect a large number of special scene images of the oil field, accurately identifies multiple targets through a training model, identifies complex targets such as a pumping unit, an engineering vehicle, a worker, an animal, a car and the like in real time, provides an alarm system, classifies different events into different safety categories, and brings great guarantee to the safety of petroleum production and transportation.
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 are briefly introduced below; it is obvious that the drawings in the following description are some embodiments of the invention, and that for a person skilled in the art, other drawings can be derived from them without inventive effort.
FIG. 1 is a schematic flow chart of a neural network-based oilfield wellsite video target detection and identification method according to an embodiment of the present invention;
FIG. 2 is a schematic representation of the identification effect of an oilfield wellsite in an embodiment of the invention.
Detailed Description
In order to make the purpose, technical effect and technical solution of the embodiments of the present invention clearer, the following clearly and completely describes the technical solution of the embodiments of the present invention with reference to the drawings in the embodiments of the present invention; it is to be understood that the described embodiments are only some of the embodiments of the present invention. Other embodiments, which can be derived by one of ordinary skill in the art from the disclosed embodiments without inventive faculty, are intended to be within the scope of the invention.
Referring to fig. 1 and 2, a method for intelligent real-time monitoring and identification of an oilfield wellsite video based on a neural network according to an embodiment of the present invention includes the following steps:
preprocessing an oil field scene image formed by frame extraction of an oil field well site video;
inputting the preprocessed image into a pre-trained deep neural network oilfield well site target detection and recognizer for recognition, and outputting target position and target category information appearing in the current image;
and matching the obtained target position and target category information with the pre-established oil field alarm grade information, and outputting an identification result.
In the embodiment of the invention, the training steps of the pre-trained deep neural network oilfield well site target detection and recognizer comprise:
establishing a target detection and identification network based on a deep neural network;
inputting a learning sample set picture marked with a target position and category information label into an established detection and recognition network based on a deep neural network for training, and adjusting detection and recognition network parameters based on the deep neural network based on a combined learning loss function of the position and the category of the target; optimizing by using a random gradient descending mode to generate an oilfield well site target detection and classification recognizer, performing model test every epoch iteration, selecting a model with the best performance and storing the model, wherein the model stored finally is the trained deep neural oilfield well site target detection and recognition device.
In the embodiment of the present invention, training a detection and recognition network based on a deep neural network established by inputting a learning sample set image labeled with a target position and category information label, and adjusting a detection and recognition network parameter based on the deep neural network based on a joint learning loss function of the target position and category specifically includes:
and 3.1, marking the acquired oil field scene image information, wherein the marked label information comprises labels such as safety helmets with different colors, work clothes with different colors, different vehicle types, different types of animals, naked flames, smoke and the like. In consideration of the real situation, the data sets of some types of label images are less or the workload of manual labeling is larger, the data sets are amplified in a data enhancement mode, classical geometric distortion, illumination distortion, image shielding and the like are carried out on a single picture, aiming at the large monitoring range of most cameras in the oil field well site scene and the existence of a large number of small targets, the invention adopts a multi-picture combination data enhancement technology to combine the collected images, and the Mosaic data augmentation technology is utilized to combine four training images into one image according to a certain proportion, so that the training model society can recognize objects in a smaller range.
3.2, aiming at the problems that a well field camera is generally shot in a far field scene, target objects occupy a small proportion in an image, a single-stage network model is adopted, the missing detection rate and the false detection rate are too high, the frame resetting phenomenon is serious and the like, the invention adopts a two-stage network model to detect and identify the targets, a first-stage network carries out target identification on a high-resolution image shot in the well field of the oil field, and position information and large category information of the target objects, such as position and category information of people, vehicles, animals, smoke and fire, existing in the image are output. Background targets are screened out from the target images output by the first-level network model, images of other targets concerned by oil field monitoring are input into a second-level network, and further, finely classified targets, such as red safety caps, blue safety caps, staff wearing work clothes, cars, trucks, engineering vehicles, sheep, dogs and other specific target category information are identified, as shown in fig. 2.
3.3, a large number of cameras exist in an oil field well site, hundreds of paths of video streams and thousands of paths of video streams need to be processed in real time, and a target object needs to be detected and identified.
And 3.4, according to the first-stage neural network model in the step 3.3, sending the image set preprocessed in the step 3.1 to a BackBone module of a Yolo network of the first-stage network model to obtain the characteristic map information of the input image, wherein the step is to further perform characteristic extraction and compression on the image information, extract a key characteristic map from the image with high resolution of a well site, and facilitate efficient calculation and accurate target detection and identification of the network model in the following steps.
And 3.5, sending the characteristic diagram extracted from the image in the step 3.4 into a Neck network and a prediction network, then designing a loss function according to the prediction label information and the artificial marking information for training, and selecting a model parameter combination which enables the error of the test set to be minimum as a parameter of a first-stage network through continuous training and testing.
And 3.5, screening the primary target detection and classification results obtained in the step 3.5, removing background images, sending the concerned target images into a second-level network model, and training and predicting. The second-level network model adopts an SSD model.
And 3.6, the second-level network model further classifies and detects the image information and the category information obtained by the first-level network in the step 3.5, wherein the second-level network model feature extraction selection and extraction module firstly extracts and processes the features of the image to form a second-level network feature map, and then sends the feature map into a classifier for classification, so that more subdivided target types, such as target positions and category information of red safety caps, blue safety caps, workers wearing working clothes, workers not wearing working clothes, cars, tank trucks, engineering vehicles, pumping units and the like, can be obtained.
The invention provides an oilfield well site video target detection and identification system based on a neural network, which comprises:
the preprocessing module is used for preprocessing an oil field scene image formed by frame extraction of an oil field well site video to obtain a preprocessed image;
the recognition and classification module is used for inputting the preprocessed image into a pre-trained deep neural network oilfield well site target detection and recognizer for recognition to obtain the target position and the target category information appearing in the image;
and the alarm matching module is used for matching the obtained target position and target category information with pre-established oil field alarm grade information and outputting an identification result.
In the embodiment of the invention, a first-level network adopts a Yolo network, aiming at complex oil field scenes, including offices, factories, operation areas, near scenes, far scenes, rainy and snowy days, daytime, night and the like, most of target objects have small area occupation in images, accurate sequencing on a large number of candidate detections is very important for excellent-performance target detectors, however, classification scores are used as sequencing bases in previous research work and cannot reliably represent sequencing, the detection performance is damaged, and particularly, a new loss function called Varifocal loss is designed and used for training intensive object detectors to predict IACS.
The network model Loss function, variafocal local, is defined as:
where p is the predicted IACS score and q is the target IoU score. For positive samples in training, q is set to IoU between the generated bbox and gt box, while for negative samples in training, the training target q for all classes is 0.
When a dense object detector is trained to enable continuous IACS regression, the invention solves the class imbalance problem by borrowing a sample weighting thought from focal loss; however, unlike focal loss, which treats positive and negative samples equally, the present invention chooses to treat them asymmetrically.
The experimental results of the inventive examples are shown in table 1.
TABLE 1 test Effect of oilfield scene data sets on different network models
As can be seen from the table 1, the method can realize effective identification of multiple complex targets in an oil field scene by improving a data set preprocessing enhancement strategy and customizing a variable local Loss function, improves the detection precision of an original algorithm, realizes rapid and effective detection and identification of multiple targets in the oil field scene, and brings safety guarantee for oil field production and transportation.
In summary, the embodiment of the present invention discloses a method for intelligent real-time monitoring and identification of an oilfield wellsite video based on a neural network, the method comprising: adopting the existing public image data set and the self-drawing image data set, marking and preprocessing the image data set to establish a database, and forming a learning sample; a joint learning loss function of the secondary deep neural network and a self-defined oilfield operation target object is performed based on the manufactured image data set; training parameters of a deep learning network to enable a loss function to be minimum; deploying the trained network parameters to a production environment for identification; and matching the recognition result with the security level information in the database, and outputting the security level information which is possibly generated and to which the recognized image belongs so as to achieve the purpose of warning. The invention alarms various safety problems in the oil field production process, and can effectively achieve the purpose of monitoring.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art can make modifications and equivalents to the embodiments of the present invention without departing from the spirit and scope of the present invention, which is set forth in the claims of the present application.
Claims (9)
1. An oilfield well site video target detection and identification method based on a neural network is characterized by comprising the following steps:
preprocessing an oil field scene image formed by frame extraction of an oil field well site video to obtain a preprocessed image;
inputting the preprocessed image into a pre-trained deep neural network oilfield well site target detection and recognition device for recognition, and obtaining target position and target category information appearing in the image; in the pre-trained deep neural network oilfield well site target detection and recognition device, the target detection and recognition network based on the deep neural network comprises:
the first-stage network model is used for carrying out target identification on a high-resolution image which is shot by an oil field well site and has the resolution ratio larger than a preset threshold value, and outputting position information and first category information of a target object in the image;
the second-level network model is used for inputting a preset concerned target object image and outputting second category information; the preset concerned target object image is obtained by removing a background target from a target image output by a first-level network model;
and matching the obtained target position and target category information with the pre-established oil field alarm grade information, and outputting an identification result.
2. The method according to claim 1, wherein the training step of the pre-trained deep neural network oilfield wellsite target detection and recognition device specifically comprises:
inputting a learning sample set picture marked with a target position and category information label into a target detection and identification network based on a deep neural network for training, and adjusting target detection and identification network parameters based on the deep neural network based on a combined learning loss function of the target position and the category;
optimizing by using a random gradient descending mode to generate an oilfield well site target detection and recognizer, performing model test once every epoch iteration, and selecting a model with the best performance or the performance meeting the preset requirement as a trained deep neural network oilfield well site target detection and recognizer.
3. The method for detecting and identifying the video target of the oilfield wellsite based on the neural network as claimed in claim 1, wherein a Yolo network is adopted as the first-stage neural network model.
4. The method of claim 3, wherein the training of the first stage neural network model comprises:
sending the learning sample set picture to a BackBone module of a first-level network model Yolo network to obtain a characteristic diagram of an input image;
and (3) sending the characteristic graph into a neutral network and a prediction network, designing a loss function according to the prediction label information and the marking information for training, and selecting a model parameter combination which enables the error of the test set to be minimum as a parameter of a first-stage network model through continuous training and testing.
5. The method of claim 1, wherein an SSD model is used as the second-level network model.
6. The method of claim 5, wherein the training step of the second-stage network model comprises:
screening the primary target detection and classification results output by the first-stage neural network model, removing background images and obtaining preset concerned target images;
and sending the preset concerned target object image into a second-level network model for training and prediction to obtain the trained second-level network model.
7. The method of claim 1, wherein the first category information comprises: human, car, animal; the second category information includes: red safety helmet, blue safety helmet, worker wearing working clothes, worker not wearing working clothes, car, tank car, engineering truck and oil pumping unit.
8. The method for detecting and identifying the video target of the oilfield well site based on the neural network as claimed in claim 2, wherein when acquiring the learning sample set picture labeled with the target position and category information tag, the data enhancement method is adopted to amplify the data set, and comprises:
carrying out geometric distortion, illumination distortion and image shielding on the single picture;
the collected images are combined by adopting a multi-image combination Data enhancement technology, and four training images are combined into one image according to a certain proportion by utilizing a Mosaic Data Augmentation technology.
9. An oilfield wellsite video target detection and identification system based on a neural network, comprising:
the preprocessing module is used for preprocessing an oil field scene image formed by frame extraction of an oil field well site video to obtain a preprocessed image;
the recognition and classification module is used for inputting the preprocessed image into a pre-trained deep neural network oilfield well site target detection and recognizer for recognition to obtain the target position and the target category information appearing in the image;
and the alarm matching module is used for matching the obtained target position and target category information with pre-established oil field alarm grade information and outputting an identification result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110485534.8A CN113191273A (en) | 2021-04-30 | 2021-04-30 | Oil field well site video target detection and identification method and system based on neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110485534.8A CN113191273A (en) | 2021-04-30 | 2021-04-30 | Oil field well site video target detection and identification method and system based on neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113191273A true CN113191273A (en) | 2021-07-30 |
Family
ID=76983592
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110485534.8A Pending CN113191273A (en) | 2021-04-30 | 2021-04-30 | Oil field well site video target detection and identification method and system based on neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113191273A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114581728A (en) * | 2022-02-22 | 2022-06-03 | 中国人民解放军军事科学院国防科技创新研究院 | Training image set generation method, device and equipment |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180181830A1 (en) * | 2015-06-05 | 2018-06-28 | Schlumberger Technology Corporation | Wellsite equipment health monitoring |
| CN109753898A (en) * | 2018-12-21 | 2019-05-14 | 中国三峡建设管理有限公司 | A kind of safety cap recognition methods and device |
| CN109886245A (en) * | 2019-03-02 | 2019-06-14 | 山东大学 | A kind of pedestrian detection recognition methods based on deep learning cascade neural network |
| CN110569843A (en) * | 2019-09-09 | 2019-12-13 | 中国矿业大学(北京) | Intelligent detection and identification method for mine target |
| CN111104903A (en) * | 2019-12-19 | 2020-05-05 | 南京邮电大学 | Depth perception traffic scene multi-target detection method and system |
| CN111178424A (en) * | 2019-12-26 | 2020-05-19 | 中国石油大学(北京) | Petrochemical production site safety compliance real-time detection system and method |
| CN111476083A (en) * | 2020-02-07 | 2020-07-31 | 山东理工大学 | Automatic identification method for wearing of safety helmet of electric power staff |
| CN112149512A (en) * | 2020-08-28 | 2020-12-29 | 成都飞机工业（集团）有限责任公司 | Helmet wearing identification method based on two-stage deep learning |
-
2021
- 2021-04-30 CN CN202110485534.8A patent/CN113191273A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180181830A1 (en) * | 2015-06-05 | 2018-06-28 | Schlumberger Technology Corporation | Wellsite equipment health monitoring |
| CN109753898A (en) * | 2018-12-21 | 2019-05-14 | 中国三峡建设管理有限公司 | A kind of safety cap recognition methods and device |
| CN109886245A (en) * | 2019-03-02 | 2019-06-14 | 山东大学 | A kind of pedestrian detection recognition methods based on deep learning cascade neural network |
| CN110569843A (en) * | 2019-09-09 | 2019-12-13 | 中国矿业大学(北京) | Intelligent detection and identification method for mine target |
| CN111104903A (en) * | 2019-12-19 | 2020-05-05 | 南京邮电大学 | Depth perception traffic scene multi-target detection method and system |
| CN111178424A (en) * | 2019-12-26 | 2020-05-19 | 中国石油大学(北京) | Petrochemical production site safety compliance real-time detection system and method |
| CN111476083A (en) * | 2020-02-07 | 2020-07-31 | 山东理工大学 | Automatic identification method for wearing of safety helmet of electric power staff |
| CN112149512A (en) * | 2020-08-28 | 2020-12-29 | 成都飞机工业（集团）有限责任公司 | Helmet wearing identification method based on two-stage deep learning |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114581728A (en) * | 2022-02-22 | 2022-06-03 | 中国人民解放军军事科学院国防科技创新研究院 | Training image set generation method, device and equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108062349B (en) | Video monitoring method and system based on video structured data and deep learning | |
| CN103069434B (en) | For the method and system of multi-mode video case index | |
| CN108052859B (en) | Abnormal behavior detection method, system and device based on clustering optical flow characteristics | |
| KR101748121B1 (en) | System and method for detecting image in real-time based on object recognition | |
| CN110689054B (en) | Worker violation monitoring method | |
| CN111339883A (en) | Method for identifying and detecting abnormal behaviors in transformer substation based on artificial intelligence in complex scene | |
| KR102122859B1 (en) | Method for tracking multi target in traffic image-monitoring-system | |
| CN110619277A (en) | Multi-community intelligent deployment and control method and system | |
| KR102122850B1 (en) | Solution for analysis road and recognition vehicle license plate employing deep-learning | |
| CN107688830B (en) | Generation method of vision information correlation layer for case serial-parallel | |
| CN112434827B (en) | Safety protection recognition unit in 5T operation and maintenance | |
| CN110852179B (en) | Suspicious personnel invasion detection method based on video monitoring platform | |
| CN110728252A (en) | Face detection method applied to regional personnel motion trail monitoring | |
| CN113743256A (en) | Construction site safety intelligent early warning method and device | |
| CN112163572A (en) | Method and device for identifying object | |
| Tao et al. | Smoky vehicle detection based on range filtering on three orthogonal planes and motion orientation histogram | |
| CN116052082A (en) | Power distribution station room anomaly detection method and device based on deep learning algorithm | |
| CN116846059A (en) | Edge detection system for power grid inspection and monitoring | |
| CN113191273A (en) | Oil field well site video target detection and identification method and system based on neural network | |
| Wang et al. | Vision-based highway traffic accident detection | |
| CN117423157A (en) | Mine abnormal video action understanding method combining migration learning and regional invasion | |
| CN117172984A (en) | Safety risk identification method and system based on equipment maintenance digital twin scene | |
| CN116682162A (en) | Robot detection algorithm based on real-time video stream | |
| CN116419059A (en) | Automatic monitoring method, device, equipment and medium based on behavior label | |
| CN113762115B (en) | Distribution network operator behavior detection method based on key point detection |
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 |