CN110263622B - Train fire monitoring method, device, terminal and storage medium - Google Patents
Train fire monitoring method, device, terminal and storage medium Download PDFInfo
- Publication number
- CN110263622B CN110263622B CN201910377069.9A CN201910377069A CN110263622B CN 110263622 B CN110263622 B CN 110263622B CN 201910377069 A CN201910377069 A CN 201910377069A CN 110263622 B CN110263622 B CN 110263622B
- Authority
- CN
- China
- Prior art keywords
- fire
- train
- images
- positive
- candidate contour
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
- G06Q50/265—Personal security, identity or safety
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/40—Business processes related to the transportation industry
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Tourism & Hospitality (AREA)
- Multimedia (AREA)
- Strategic Management (AREA)
- Primary Health Care (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- General Health & Medical Sciences (AREA)
- Economics (AREA)
- Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Educational Administration (AREA)
- Development Economics (AREA)
- Computer Security & Cryptography (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Train Traffic Observation, Control, And Security (AREA)
- Image Analysis (AREA)
Abstract
A train fire monitoring method comprising: continuously acquiring a plurality of images in the running process of the train; detecting a plurality of candidate contour areas in each image by using a YOLO target detection algorithm; inputting the plurality of candidate contour areas into a pre-trained SVM model, and outputting a result of fire on the train when at least one of the candidate contour areas is identified to have the fire through the SVM model, wherein the positions of the candidate contour areas with the fire in each image are inconsistent. The invention also provides a train fire monitoring device, a terminal and a storage medium. According to the invention, a plurality of images of the train can be continuously acquired in a non-station road section through a video monitoring technology, and the images are analyzed to determine whether fire occurs on the train, so that an effective auxiliary technical means is provided for safety monitoring of the train, and the running safety of the train is improved.
Description
Technical Field
The invention relates to the technical field of video monitoring, in particular to a train fire monitoring method, a train fire monitoring device, a train fire monitoring terminal and a train fire monitoring storage medium.
Background
The technology of high-speed railways and high-speed trains in China has been developed for over 20 years, and the favorite proud achievement is obtained. However, aiming at the characteristics of large scale of the railway network, wide coverage area, long transportation line and the like in China, the running safety of the train becomes an important problem to be solved by the national railway development, and once the train safety accident occurs, immeasurable loss and disaster can be brought to the country and people. The train fire safety precaution is an important point of train safety precaution, and has become an important subject of national and research institutions.
At present, the train operation safety and fire safety monitoring adopts a purely manual mode, related data information is obtained by means of manual on-site inspection, a great amount of time and labor are consumed for manual detection, the detection efficiency is low, and errors are easy to occur; in addition, the detection cannot be performed manually on a road section other than the station.
Therefore, it is necessary to provide a technical solution for monitoring train fires without relying on manual work and on non-station road sections.
Disclosure of Invention
In view of the above, it is necessary to provide a train fire monitoring method, device, terminal and storage medium, which can continuously acquire multiple images of a train on a non-station road section through a video monitoring technology, analyze the multiple images to determine whether a fire occurs on the train, provide an effective auxiliary technical means for safety monitoring of the train, and improve the safety of train operation.
A first aspect of the present invention provides a train fire monitoring method, the method comprising:
continuously acquiring a plurality of images in the running process of the train;
detecting a plurality of candidate contour areas in each image by using a YOLO target detection algorithm;
inputting the candidate contour areas into a pre-trained SVM model, and identifying whether a fire occurs in at least one candidate contour area through the SVM model;
When the SVM model recognizes that the fire occurs in at least one candidate contour area, judging whether the positions of the candidate contour areas with the fire are consistent in the plurality of images;
and outputting a result of the fire on the train when judging that the positions of the candidate outline areas with the fire are inconsistent in each image.
Preferably, the determining whether the positions of the candidate contour areas in which the fire occurs in the plurality of images are identical includes:
taking the candidate contour area with the fire as a first target contour area;
acquiring images with continuous preset numbers from the plurality of images;
judging whether the displacement of the first target contour area in the continuous preset number of images is changed or not;
when the displacement of the first target contour area in the continuous preset number of images is judged to be changed, the method further comprises the following steps: determining that the positions of the candidate outline areas with the fire conditions in each image are inconsistent, and determining that the fire conditions are the fire conditions on the train;
When the displacement of the first target contour area in the continuous preset number of images is not changed, the method further comprises: and determining that the positions of the candidate outline areas where the fire occurs are consistent in each image, and determining that the fire is not the fire on the train.
Preferably, the training process of the SVM model includes:
Acquiring a plurality of images of fire as a positive sample set and a plurality of images of red objects as a negative sample set;
Generating a positive and negative sample training set and a positive and negative sample testing set according to the obtained positive and negative sample set;
inputting the positive and negative sample training sets into an SVM for training to obtain an SVM model;
Inputting the positive and negative sample test set into the SVM model for testing to obtain a test passing rate;
If the test passing rate is greater than or equal to a preset passing rate threshold value, determining that the training of the SVM model is finished;
And if the test passing rate is smaller than the preset passing rate threshold, increasing the number of the positive and negative sample training sets, and carrying out training on the SVM model again.
Preferably, the generating the positive and negative sample training set and the positive and negative sample testing set according to the obtained positive and negative sample set includes:
and randomly selecting a first preset number of images from the positive and negative sample sets to serve as the positive and negative sample training set, and randomly selecting a second preset number of images to serve as the positive and negative sample testing set by adopting a random number generation algorithm.
Preferably, when no fire is identified by the SVM model as occurring in the candidate contour region; or when it is identified by the SVM model that a fire has occurred in at least one of the candidate contour regions, but the positions of the candidate contour regions where the fire has occurred are consistent in each image, the method further includes:
And outputting a result of no fire condition on the train.
Preferably, after said outputting the result of the train having a fire, the method further comprises:
sending alarm information to a train driver of the train;
and simultaneously, sending alarm information containing the locomotive number of the train to a dispatching room of a front station.
Preferably, after the continuously acquiring the plurality of images during the running of the train, the method further comprises:
performing illumination or contrast normalization processing on the plurality of images;
and carrying out noise reduction treatment on the plurality of images subjected to the normalization treatment by adopting a bilateral filtering algorithm.
A second aspect of the present invention provides a train fire monitoring apparatus, the apparatus comprising:
The acquisition module is used for continuously acquiring a plurality of images in the running process of the train;
the detection module is used for detecting a plurality of candidate contour areas in each image by using a YOLO target detection algorithm; the identification module is used for inputting the candidate contour areas into a pre-trained SVM model, and identifying whether a fire occurs in at least one candidate contour area or not through the SVM model;
The judging module is used for judging whether the positions of the candidate contour areas with the fire conditions in each image are consistent or not when the identification module identifies that the fire conditions occur in at least one candidate contour area through the SVM model; and the output module is used for outputting a result of the fire condition on the train when the judging module determines that the positions of the candidate outline areas with the fire condition in each image are inconsistent.
A third aspect of the invention provides a terminal comprising a processor for implementing the train fire monitoring method when executing a computer program stored in a memory.
A fourth aspect of the present invention provides a computer readable storage medium having a computer program stored thereon, which when executed by a processor, implements the train fire monitoring method.
In summary, according to the train fire monitoring method, device, terminal and storage medium of the invention, a plurality of images are continuously acquired in the running process of the train, then a YOLO target detection algorithm is adopted to detect a plurality of candidate contour areas in each image, and finally the candidate contour areas are input into a pre-trained SVM model for recognition, so that whether fire occurs in the candidate contour areas can be obtained, when the condition that the candidate contour areas occur is determined, whether the position of the candidate contour areas where the fire occurs is further determined to be fixed, and when the position is determined to be fixed, the result that the fire occurs on the train is output. The method and the device provide effective auxiliary technical means for safety detection of trains and pedestrians, break through the mechanism of completely relying on pure manual vision, hearing and touch to passively judge whether fire occurs on the trains, avoid the risk of easy mistakes of manual operation, greatly lighten the workload of on-site staff, improve the safety of train operation, improve the working efficiency of the staff, realize effective control on transportation safety production and provide effective high-definition image basis for a train dispatching room. And secondly, the combination position further ensures whether the fire is the fire on the train, so that the occurrence of stopping and delay events caused by false alarm or misinformation is avoided, and the railway transportation efficiency is improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are required to be used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only embodiments of the present invention, and that other drawings can be obtained according to the provided drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flowchart of a train fire monitoring method according to an embodiment of the present invention.
Fig. 2 is a block diagram of a train fire monitoring apparatus according to a second embodiment of the present invention.
Fig. 3 is a schematic structural diagram of a terminal according to a third embodiment of the present invention.
The invention will be further described in the following detailed description in conjunction with the above-described figures.
Detailed Description
In order that the above-recited objects, features and advantages of the present invention will be more clearly understood, a more particular description of the invention will be rendered by reference to specific embodiments thereof which are illustrated in the appended drawings. It should be noted that, without conflict, the embodiments of the present invention and features in the embodiments may be combined with each other.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, and the described embodiments are merely some, rather than all, embodiments of the present invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.
Example 1
Fig. 1 is a flowchart of a train fire monitoring method according to an embodiment of the present invention.
In this embodiment, the train fire monitoring method may be applied to a terminal, and for a terminal that needs to monitor train fire, the function of train fire monitoring provided by the method of the present invention may be directly integrated on the terminal, or may be run in the terminal in the form of a software development kit (Software Development Kit, SKD).
As shown in fig. 1, the train fire monitoring method is applied to the specific scene in the train running process, and specifically includes the following steps, the sequence of the steps in the flowchart may be changed according to different requirements, and some may be omitted.
S11: multiple images are continuously acquired during the running of the train.
The train in this embodiment is composed of tens or even tens of container cars, each container car being about 35-40 meters long. The train generally includes a head portion at the front end of the first car and a tail portion at the end of the last car.
In this embodiment, a plurality of images of a train during traveling may be acquired by a high-definition digital image acquisition device. Because the running speed of the train is high, the running speed can reach hundreds of meters per second, and whether fire occurs on the train can not be accurately determined through a single image, the high-speed continuous shooting digital photographing technology can be adopted, and when the train passes through, the high-definition digital image acquisition equipment can rapidly carry out continuous shooting on the train so as to acquire a plurality of high-definition digital images including the train. Or the high-definition digital image acquisition equipment is used for acquiring a video stream, and a plurality of images are obtained by extracting frames in the video stream so as to comprehensively identify whether a fire condition occurs on the train through a plurality of continuous images.
In this embodiment, the high-definition digital image capturing device may include a plurality of high-definition cameras, and the high-definition cameras are mounted on at least one supporting frame disposed along the train travel route on both sides of the monitored train track. Specifically, a plurality of high-definition cameras can be respectively arranged on a top beam and two upright posts of a rail portal frame, and video images of the top and two sides of a train are obtained in real time and sent to a streaming media storage device. Of course, in other embodiments, the high-definition camera may have other mounting positions, so long as the high-definition camera can clearly obtain clear images of the top and two sides of the train.
The high-definition digital image acquisition equipment can send the acquired high-definition images to the indoor video analysis servers of the dispatching rooms of all stations through technologies such as a special video optical transceiver and a wired network, so that a train management department can master the running condition of a train at any time through the indoor video analysis servers.
Preferably, after the continuously acquiring the plurality of images during the running of the train, the method further comprises:
preprocessing the plurality of images.
In this embodiment, preprocessing the plurality of images includes: performing illumination or contrast normalization processing on the plurality of images; and carrying out noise reduction treatment on the plurality of images subjected to the normalization treatment by adopting a bilateral filtering algorithm.
Because the train running states of all weather (different time periods, different light intensities and different climatic characteristics) are monitored, after the pretreatment is carried out on a plurality of high-definition images, the normalization of the illumination or contrast of the images under different illumination conditions in different time periods can be solved, so that the characteristics of the parts needing to be identified in the images are optimized and highlighted, and the other characteristics of the parts not needing to be identified are weakened, so that the accuracy and the identification speed of the image identification are improved.
The bilateral filtering algorithm can effectively remove noise, such as salt and pepper noise generated by high-definition digital image acquisition equipment, and meanwhile has good edge detail retaining capability. The processing procedure of the bilateral filtering algorithm is the prior art, and the present invention is not described in detail here.
In this embodiment, the image gray scale normalization processing is not required for the plurality of images, that is, the plurality of images or the plurality of preprocessed images are color images.
S12: a plurality of candidate contour regions in each image are detected using a YOLO target detection algorithm.
In this embodiment, after obtaining a plurality of images including a train, it is necessary to detect a plurality of targets in each image according to a YOLO target detection algorithm set in advance. The YOLO (You Only Look Once: beer, fast, stronger) object detection algorithm is a fast multi-object detection algorithm, capable of detecting a plurality of candidate objects simultaneously, and the contour area of each candidate object is selected by a form frame of a rectangular frame, and the contour area of the candidate object is called a candidate contour area.
In this embodiment, the red area is targeted, for example, a red area corresponding to a fire occurring on a train, a red area corresponding to a red bag hooked on the outside of a train car, or a red area corresponding to a fire occurring in other areas along the railway.
S13: and inputting the candidate contour areas into a pre-trained SVM model, and identifying whether a fire occurs in at least one candidate contour area through the SVM model.
In this embodiment, after detecting a plurality of candidate contour regions, it is necessary to screen out a target contour region representing a possible occurrence of a fire from among the plurality of candidate contour regions, and further determine whether the fire actually occurs in the target contour region.
The support vector machine model (Support Vector Machine, SVM) is trained in advance, and the support vector machine (Support Vector Machine, SVM) is used as a classification model, has a good classification effect and is usually used for pattern recognition, classification and regression analysis. The main idea of SVM: firstly, the linear separable condition is analyzed, and for the linear inseparable condition, the nonlinear mapping algorithm is used for converting the sample which is linearly inseparable in the low-dimensional input space into the high-dimensional characteristic space so as to make the high-dimensional characteristic space linearly separable, so that the linear analysis of the nonlinear characteristics of the sample by adopting the linear algorithm is possible.
And identifying whether a fire occurs in at least one candidate contour area through a support vector machine model. When a fire condition occurs in a certain candidate contour area output through the SVM model, the probability of occurrence of the fire condition on the train is high. And when no fire occurs in all the candidate contour areas through the SVM model, namely when no fire occurs in the candidate contour areas through the SVM model, indicating that no fire occurs on the train.
Preferably, the training process of the SVM model includes:
131 Acquiring a plurality of images of fire as a positive sample set and a plurality of images of red objects as a negative sample set;
Since most of the areas in the images obtained when a fire occurs are red, and most of the areas in the images of the red object are also red, the images of the red object are very easily misjudged as the images when a fire occurs, so a large number of images when a fire occurs can be obtained in advance as a positive sample set, and a large number of images when a fire does not occur but are red objects as a negative sample set.
132 Generating a positive and negative sample training set and a positive and negative sample testing set according to the obtained positive and negative sample set;
In the preferred embodiment, the idea of Cross Validation (Cross Validation) may be adopted when training the SVM model, and the obtained positive and negative sample sets are divided into a positive and negative sample training set and a positive and negative sample testing set according to a suitable proportion, for example, 6:4.
Specifically, the generating the positive and negative sample training set and the positive and negative sample testing set according to the obtained positive and negative sample set includes:
and randomly selecting a first preset number of images from the positive and negative sample sets to serve as the positive and negative sample training set, and randomly selecting a second preset number of images to serve as the positive and negative sample testing set by adopting a random number generation algorithm.
The positive and negative sample training sets are used for training the SVM model, the positive and negative sample testing sets are used for testing the performance of the trained SVM model, and if the testing accuracy is higher, the performance of the trained SVM model is better; if the accuracy of the test is low, the performance of the trained SVM model is poor.
If the total number of the divided positive and negative sample training sets is still larger, that is, all the positive and negative sample training sets are used for participating in training of the SVM model, the cost of searching the optimal parameters corresponding to the SVM model is larger, so that samples with a first preset proportion can be randomly selected in the generated positive and negative sample training sets to participate in training, and the number of the samples participating in training is reduced.
In the preferred embodiment, in order to increase the randomness of the positive and negative sample training sets involved in training, a random number generation algorithm may be used for random selection.
In the preferred embodiment, the first preset number may be a preset fixed value, for example, 40, that is, 40 samples are randomly selected from the generated positive and negative sample training sets to participate in training of the SVM model. The first preset number may also be a preset scale value, for example, 1/10, i.e., 1/10 of the samples in the positive and negative sample training sets are randomly selected to participate in the training of the SVM model.
133 Inputting the positive and negative sample training sets into an SVM to train so as to obtain an SVM model;
134 Inputting the positive and negative sample test set into the SVM model for testing to obtain a test passing rate;
135 If the test passing rate is greater than or equal to a preset passing rate threshold value, determining that the training of the SVM model is finished; and if the test passing rate is smaller than the preset passing rate threshold, increasing the number of the positive and negative sample training sets, and carrying out training on the SVM model again.
In this embodiment, the positive sample training set may be marked by using a first identifier, the negative sample training set may be marked by using a second identifier, and then, a part of the positive sample training set carrying the first identifier and a part of the negative sample training set carrying the second identifier are trained to obtain an SVM model. When the test is performed, whether fire occurs in the candidate contour area or not can be determined through the output identification of the SVM model, for example, if the first identification is output, the fire occurs in the candidate contour area can be determined, and if the second identification is output, the fire does not occur in the candidate contour area can be determined.
If the SVM model recognizes that a fire occurs in at least one candidate contour area, executing S14; and if no fire condition occurs in the candidate contour area through the SVM model, executing S16.
S14: and judging whether the positions of the candidate outline areas where the fire occurs in each image are consistent.
In this embodiment, even if it is recognized that a fire occurs in at least one of the candidate contour areas according to the pre-trained SVM model, it is not ensured that the fire is on the train, and possibly that other devices on the railway along the line have a fire, and the fire is just shot when the train arrives, it is necessary to further determine whether the fire recognized by the SVM model is a fire on the train by determining whether the positions of the candidate contour areas where the fire occurs in each image are consistent, so as to avoid false alarm or misreport.
Preferably, the determining whether the positions of the candidate contour areas in which the fire occurs in the plurality of images are identical includes:
taking the candidate contour area with the fire as a first target contour area;
acquiring images with continuous preset numbers from the plurality of images;
And judging whether the displacement of the first target contour area in the continuous preset number of images is changed or not.
And determining whether the positions of the candidate outline areas with the fire are consistent in the images by judging whether the displacement of the candidate outline areas with the fire in the images with the continuous preset number is changed or not. When the displacement of the first target contour area in the continuous preset number of images is judged to be changed, determining that the positions of the candidate contour areas with the fire are inconsistent in each image, and considering the fire as the fire on the train; and when judging that the displacement of the first target contour area in the continuous preset number of images is not changed, determining that the positions of the candidate contour areas with the fire are consistent in each image, and considering that the fire is not the fire on the train.
For example, assuming that 4 images are obtained, a LOYO target detection algorithm is used to detect 3 candidate contour regions in each image, wherein some candidate contour regions may have fire, some candidate contour regions may have red objects, and 2 candidate contour regions can be clearly identified from the pre-trained SVM image, and another 1 candidate contour region has no fire. At this time, it is not possible to accurately determine whether the fire occurring in the 2 candidate outlines is a fire on a train or a fire on other equipment along the railway, and thus it is necessary to further determine whether the positions of the target outline areas in the 4 images are inconsistent in the 2 candidate outline areas. A fire is considered on-train when at least 1 of the 2 candidate contour regions are not consistently located in the 4 images. When the positions of the 2 candidate contour areas in the 4 images are consistent, the fire is not on the train, and is considered to be on other equipment on the railway along the line.
It should be understood that the positions described in this embodiment refer to relative positions, i.e. the outline area on each image is relative to that image, and if the fire is on a train, the fire should move with the train, and the relative positions in the images will change; if the fire is not on a train, the fire should be fixed and the relative position in the multiple images fixed, consistent.
When it is judged that the positions of the candidate contour regions in each image where the fire has occurred are inconsistent, S15 is performed; when it is judged that the positions of the candidate contour regions in each image where the fire has occurred are identical, S16 is performed.
S15: and outputting the result of the fire condition on the train.
In this embodiment, when it is identified by the SVM model that a fire occurs in at least one of the candidate contour regions, and the positions of the candidate contour regions with fire in each image are inconsistent, a result that there is no fire on the train is output.
S16: and outputting a result of no fire condition on the train.
In this embodiment, when it is recognized by the SVM model that no fire occurs in the candidate contour region; or when the SVM model recognizes that fire occurs in at least one candidate contour area, and the positions of the candidate contour areas with the fire are consistent in each image, outputting a result that no fire occurs on the train.
Further, after the outputting the result of the fire on the train, the method further includes:
sending alarm information to a train driver of the train;
and simultaneously, sending alarm information containing the locomotive number of the train to a dispatching room of a front station.
In this embodiment, the locomotive number is a number for permanently marking a train, including: model and number, locomotive number is unique in the jurisdiction of the whole road. And when a fire is found, sending alarm information to a dispatching room of a train driver and a front station at the same time. The warning information is sent to the train driver, so that the driver can know the running train condition in real time, the driver is reminded to stop the train in time to process the fire, and the greater loss caused by further spreading of the fire is avoided; and the warning information containing the locomotive number of the train is sent to the front station, so that the staff at the front station can conveniently and timely drive to the place where the train with the fire condition is located to rescue. Namely, the effect of double guarantee on timely rescue of the train with fire is achieved.
In summary, in the train fire monitoring method, during the running process of the train, a plurality of images are continuously acquired, then a YOLO target detection algorithm is adopted to detect a plurality of candidate contour areas in each image, finally the candidate contour areas are input into a pre-trained SVM model for recognition, whether fire occurs in the candidate contour areas can be obtained, when the condition that the fire occurs in the candidate contour areas is determined, whether the position of the candidate contour areas with the fire is fixed is further judged, and when the position is fixed, the result that the fire occurs in the train is output. The method and the device provide effective auxiliary technical means for safety detection of trains and pedestrians, break through the mechanism of completely relying on pure manual vision, hearing and touch to passively judge whether fire occurs on the trains, avoid the risk of easy mistakes of manual operation, greatly lighten the workload of on-site staff, improve the safety of train operation, improve the working efficiency of the staff, realize effective control on transportation safety production and provide effective high-definition image basis for a train dispatching room. Secondly, the combination position further ensures whether the fire is the fire on the train, so that the occurrence of stopping and delay events caused by false alarm or misinformation is avoided, and the railway transportation efficiency is improved; furthermore, the train with fire can be warned in real time, the fire can be controlled in time, and further expansion of the fire is avoided.
Example two
Fig. 2 is a block diagram of a train fire monitoring apparatus according to a second embodiment of the present invention.
In some embodiments, the train fire monitoring apparatus 20 may include a plurality of functional modules comprised of program code segments. Program code for each program segment in the train fire monitoring apparatus 20 may be stored in a memory of the terminal and executed by the at least one processor to perform (see fig. 1 for details) monitoring for the presence of a train fire.
In this embodiment, the train fire monitoring device 20, which operates in a specific scenario during the running of the train, may be divided into a plurality of functional modules according to the functions performed by the train fire monitoring device. The functional module may include: the device comprises an acquisition module 201, a preprocessing module 202, a detection module 203, an identification module 204, an output module 205, a judgment module 206, a determination module 207 and a sending module 208. The module referred to in the present invention refers to a series of computer program segments capable of being executed by at least one processor and of performing a fixed function, stored in a memory. In the present embodiment, the functions of the respective modules will be described in detail in the following embodiments.
An acquisition module 201, configured to continuously acquire a plurality of images during the running process of the train.
The train in this embodiment is composed of tens or even tens of container cars, each container car being about 35-40 meters long. The train generally includes a head portion at the front end of the first car and a tail portion at the end of the last car.
In this embodiment, a plurality of images of a train during traveling may be acquired by a high-definition digital image acquisition device. Because the running speed of the train is high, the running speed can reach hundreds of meters per second, and whether fire occurs on the train can not be accurately determined through a single image, the high-speed continuous shooting digital photographing technology can be adopted, and when the train passes through, the high-definition digital image acquisition equipment can rapidly carry out continuous shooting on the train so as to acquire a plurality of high-definition digital images including the train. Or the high-definition digital image acquisition equipment is used for acquiring a video stream, and a plurality of images are obtained by extracting frames in the video stream so as to comprehensively identify whether a fire condition occurs on the train through a plurality of continuous images.
In this embodiment, the high-definition digital image capturing device may include a plurality of high-definition cameras, and the high-definition cameras are mounted on at least one supporting frame disposed along the train travel route on both sides of the monitored train track. Specifically, a plurality of high-definition cameras can be respectively arranged on a top beam and two upright posts of a rail portal frame, and video images of the top and two sides of a train are obtained in real time and sent to a streaming media storage device. Of course, in other embodiments, the high-definition camera may have other mounting positions, so long as the high-definition camera can clearly obtain clear images of the top and two sides of the train.
The high-definition digital image acquisition equipment can send the acquired high-definition images to the indoor video analysis servers of the dispatching rooms of all stations through technologies such as a special video optical transceiver and a wired network, so that a train management department can master the running condition of a train at any time through the indoor video analysis servers.
Preferably, after the acquisition module 201 continuously acquires a plurality of images during the running of the train, the train fire monitoring apparatus 20 further includes:
A preprocessing module 202, configured to preprocess the plurality of images.
In this embodiment, preprocessing the plurality of images includes: performing illumination or contrast normalization processing on the plurality of images; and carrying out noise reduction treatment on the plurality of images subjected to the normalization treatment by adopting a bilateral filtering algorithm.
Because the train running states of all weather (different time periods, different light intensities and different climatic characteristics) are monitored, after the pretreatment is carried out on a plurality of high-definition images, the normalization of the illumination or contrast of the images under different illumination conditions in different time periods can be solved, so that the characteristics of the parts needing to be identified in the images are optimized and highlighted, and the other characteristics of the parts not needing to be identified are weakened, so that the accuracy and the identification speed of the image identification are improved.
The bilateral filtering algorithm can effectively remove noise, such as salt and pepper noise generated by high-definition digital image acquisition equipment, and meanwhile has good edge detail retaining capability. The processing procedure of the bilateral filtering algorithm is the prior art, and the present invention is not described in detail here.
In this embodiment, the image gray scale normalization processing is not required for the plurality of images, that is, the plurality of images or the plurality of preprocessed images are color images.
The detection module 203 is configured to detect a plurality of candidate contour regions in each image by using a YOLO target detection algorithm.
In this embodiment, after obtaining a plurality of images including a train, it is necessary to detect a plurality of targets in each image according to a YOLO target detection algorithm set in advance. The YOLO (You Only Look Once: beer, fast, stronger) object detection algorithm is a fast multi-object detection algorithm, capable of detecting a plurality of candidate objects simultaneously, and the contour area of each candidate object is selected by a form frame of a rectangular frame, and the contour area of the candidate object is called a candidate contour area.
In this embodiment, the red area is targeted, for example, a red area corresponding to a fire occurring on a train, a red area corresponding to a red bag hooked on the outside of a train car, or a red area corresponding to a fire occurring in other areas along the railway.
The identifying module 204 is configured to input the plurality of candidate contour regions into a pre-trained SVM model, and identify, by using the SVM model, whether a fire occurs in at least one of the candidate contour regions.
In this embodiment, after detecting a plurality of candidate contour regions, it is necessary to screen out a target contour region representing a possible occurrence of a fire from among the plurality of candidate contour regions, and further determine whether the fire actually occurs in the target contour region.
The support vector machine model (Support Vector Machine, SVM) is trained in advance, and the support vector machine (Support Vector Machine, SVM) is used as a classification model, has a good classification effect and is usually used for pattern recognition, classification and regression analysis. The main idea of SVM: firstly, the linear separable condition is analyzed, and for the linear inseparable condition, the nonlinear mapping algorithm is used for converting the sample which is linearly inseparable in the low-dimensional input space into the high-dimensional characteristic space so as to make the high-dimensional characteristic space linearly separable, so that the linear analysis of the nonlinear characteristics of the sample by adopting the linear algorithm is possible.
And identifying whether a fire occurs in at least one candidate contour area through a support vector machine model. When a fire condition occurs in a certain candidate contour area output through the SVM model, the probability of occurrence of the fire condition on the train is high. And when no fire occurs in all the candidate contour areas through the SVM model, namely when no fire occurs in the candidate contour areas through the SVM model, indicating that no fire occurs on the train.
Preferably, the training process of the SVM model includes:
131 Acquiring a plurality of images of fire as a positive sample set and a plurality of images of red objects as a negative sample set;
Since most of the areas in the images obtained when a fire occurs are red, and most of the areas in the images of the red object are also red, the images of the red object are very easily misjudged as the images when a fire occurs, so a large number of images when a fire occurs can be obtained in advance as a positive sample set, and a large number of images when a fire does not occur but are red objects as a negative sample set.
132 Generating a positive and negative sample training set and a positive and negative sample testing set according to the obtained positive and negative sample set;
In the preferred embodiment, the idea of Cross Validation (Cross Validation) may be adopted when training the SVM model, and the obtained positive and negative sample sets are divided into a positive and negative sample training set and a positive and negative sample testing set according to a suitable proportion, for example, 6:4.
Specifically, the generating the positive and negative sample training set and the positive and negative sample testing set according to the obtained positive and negative sample set includes:
and randomly selecting a first preset number of images from the positive and negative sample sets to serve as the positive and negative sample training set, and randomly selecting a second preset number of images to serve as the positive and negative sample testing set by adopting a random number generation algorithm.
The positive and negative sample training sets are used for training the SVM model, the positive and negative sample testing sets are used for testing the performance of the trained SVM model, and if the testing accuracy is higher, the performance of the trained SVM model is better; if the accuracy of the test is low, the performance of the trained SVM model is poor.
If the total number of the divided positive and negative sample training sets is still larger, that is, all the positive and negative sample training sets are used for participating in training of the SVM model, the cost of searching the optimal parameters corresponding to the SVM model is larger, so that samples with a first preset proportion can be randomly selected in the generated positive and negative sample training sets to participate in training, and the number of the samples participating in training is reduced.
In the preferred embodiment, in order to increase the randomness of the positive and negative sample training sets involved in training, a random number generation algorithm may be used for random selection.
In the preferred embodiment, the first preset number may be a preset fixed value, for example, 40, that is, 40 samples are randomly selected from the generated positive and negative sample training sets to participate in training of the SVM model. The first preset number may also be a preset scale value, for example, 1/10, i.e., 1/10 of the samples in the positive and negative sample training sets are randomly selected to participate in the training of the SVM model.
133 Inputting the positive and negative sample training sets into an SVM to train so as to obtain an SVM model;
134 Inputting the positive and negative sample test set into the SVM model for testing to obtain a test passing rate;
135 If the test passing rate is greater than or equal to a preset passing rate threshold value, determining that the training of the SVM model is finished; and if the test passing rate is smaller than the preset passing rate threshold, increasing the number of the positive and negative sample training sets, and carrying out training on the SVM model again.
In this embodiment, the positive sample training set may be marked by using a first identifier, the negative sample training set may be marked by using a second identifier, and then, a part of the positive sample training set carrying the first identifier and a part of the negative sample training set carrying the second identifier are trained to obtain an SVM model. When the test is performed, whether fire occurs in the candidate contour area or not can be determined through the output identification of the SVM model, for example, if the first identification is output, the fire occurs in the candidate contour area can be determined, and if the second identification is output, the fire does not occur in the candidate contour area can be determined.
An output module 205 for identifying, by the identification module 204, that no fire has occurred in the candidate contour region by the SVM model; and outputting a result of no fire condition on the train.
A judging module 206, configured to judge whether the positions of the candidate contour areas with fire are consistent in each image when the identifying module 204 identifies that at least one of the candidate contour areas has fire through the SVM model.
In this embodiment, even if it is recognized that a fire occurs in at least one of the candidate contour areas according to the pre-trained SVM model, it is not ensured that the fire is on the train, and possibly that other devices on the railway along the line have a fire, and the fire is just shot when the train arrives, it is necessary to further determine whether the fire recognized by the SVM model is a fire on the train by determining whether the positions of the candidate contour areas where the fire occurs in each image are consistent, so as to avoid false alarm or misreport.
Preferably, the determining module 206 determines whether the positions of the candidate contour areas in which the fire occurs are identical in the plurality of images includes:
taking the candidate contour area with the fire as a first target contour area;
acquiring images with continuous preset numbers from the plurality of images;
And judging whether the displacement of the first target contour area in the continuous preset number of images is changed or not.
And determining whether the positions of the candidate outline areas with the fire are consistent in the images by judging whether the displacement of the candidate outline areas with the fire in the images with the continuous preset number is changed or not. When it is determined that the displacement of the first target contour area in the continuous preset number of images changes, a determining module 207 is configured to determine that the positions of the candidate contour areas with fire in each image are inconsistent, and consider the fire to be a fire on the train; when it is determined that the displacement of the first target contour area in the continuous preset number of images does not change, the determining module 207 is further configured to determine that the positions of the candidate contour areas with fire in each image are consistent, and consider that the fire is not a fire on the train.
For example, assuming that 4 images are obtained, a LOYO target detection algorithm is used to detect 3 candidate contour regions in each image, wherein some candidate contour regions may have fire, some candidate contour regions may have red objects, and 2 candidate contour regions can be clearly identified from the pre-trained SVM image, and another 1 candidate contour region has no fire. At this time, it is not possible to accurately determine whether the fire occurring in the 2 candidate outlines is a fire on a train or a fire on other equipment along the railway, and thus it is necessary to further determine whether the positions of the target outline areas in the 4 images are inconsistent in the 2 candidate outline areas. A fire is considered on-train when at least 1 of the 2 candidate contour regions are not consistently located in the 4 images. When the positions of the 2 candidate contour areas in the 4 images are consistent, the fire is not on the train, and is considered to be on other equipment on the railway along the line.
It should be understood that the positions described in this embodiment refer to relative positions, i.e. the outline area on each image is relative to that image, and if the fire is on a train, the fire should move with the train, and the relative positions in the images will change; if the fire is not on a train, the fire should be fixed and the relative position in the multiple images fixed, consistent.
The output module 205 is further configured to output a result that there is no fire on the train when the identifying module 204 identifies that there is a fire in at least one of the candidate contour areas through the SVM model, but the judging module 206 judges that the positions of the candidate contour areas with fire in each image are consistent.
The output module 205 is further configured to output a result of a fire on the train when the judging module 206 judges that the positions of the candidate contour areas where the fire occurs are inconsistent in each image.
Further, after the outputting of the result of the fire on the train, the train fire monitoring apparatus 20 further includes:
a sending module 208, configured to send an alarm message to a train driver of the train;
The sending module 208 is further configured to send an alert message including a locomotive number of the train to a dispatch room of a front station.
In this embodiment, the locomotive number is a number for permanently marking a train, including: model and number, locomotive number is unique in the jurisdiction of the whole road. And when a fire is found, sending alarm information to a dispatching room of a train driver and a front station at the same time. The warning information is sent to the train driver, so that the driver can know the running train condition in real time, the driver is reminded to stop the train in time to process the fire, and the greater loss caused by further spreading of the fire is avoided; and the warning information containing the locomotive number of the train is sent to the front station, so that the staff at the front station can conveniently and timely drive to the place where the train with the fire condition is located to rescue. Namely, the effect of double guarantee on timely rescue of the train with fire is achieved.
In summary, in the train fire monitoring device, a plurality of images are continuously acquired in the running process of the train, then a YOLO target detection algorithm is adopted to detect a plurality of candidate contour areas in each image, finally the candidate contour areas are input into a pre-trained SVM model for recognition, whether fire occurs in the candidate contour areas or not can be obtained, when the condition that the fire occurs in the candidate contour areas is determined, whether the position of the candidate contour areas with the fire is fixed is further judged, and when the position is fixed, the result that the fire occurs in the train is output. The method and the device provide effective auxiliary technical means for safety detection of trains and pedestrians, break through the mechanism of completely relying on pure manual vision, hearing and touch to passively judge whether fire occurs on the trains, avoid the risk of easy mistakes of manual operation, greatly lighten the workload of on-site staff, improve the safety of train operation, improve the working efficiency of the staff, realize effective control on transportation safety production and provide effective high-definition image basis for a train dispatching room. Secondly, the combination position further ensures whether the fire is the fire on the train, so that the occurrence of stopping and delay events caused by false alarm or misinformation is avoided, and the railway transportation efficiency is improved; furthermore, the train with fire can be warned in real time, the fire can be controlled in time, and further expansion of the fire is avoided.
Example III
Fig. 3 is a schematic structural diagram of a terminal according to a third embodiment of the present invention. In the preferred embodiment of the invention, the terminal 3 comprises a memory 31, at least one processor 32, at least one communication bus 33 and a transceiver 34.
It will be appreciated by those skilled in the art that the configuration of the terminal shown in fig. 3 is not limiting of the embodiments of the present invention, and that it may be a bus type configuration, a star type configuration, or a combination of hardware and software, or a different arrangement of components, as the terminal 3 may include more or less hardware or software than is shown.
In some embodiments, the terminal 3 includes a terminal capable of automatically performing numerical calculation and/or information processing according to a preset or stored instruction, and its hardware includes, but is not limited to, a microprocessor, an application specific integrated circuit, a programmable gate array, a digital processor, an embedded device, and the like. The terminal 3 may further comprise a client device, which includes, but is not limited to, any electronic product capable of performing man-machine interaction with a client through a keyboard, a mouse, a remote controller, a touch pad, a voice control device, etc., for example, a personal computer, a tablet computer, a smart phone, a digital camera, etc.
It should be noted that the terminal 3 is only used as an example, and other electronic products that may be present in the present invention or may be present in the future are also included in the scope of the present invention by way of reference.
In some embodiments, the memory 31 is used to store program codes and various data, such as train fire monitoring apparatus 20 installed in the terminal 3, and to enable high speed, automatic access to programs or data during operation of the terminal 3. The Memory 31 includes Read-Only Memory (ROM), random-access Memory (Random Access Memory, RAM), programmable Read-Only Memory (Programmable Read-Only Memory, PROM), erasable programmable Read-Only Memory (Erasable Programmable Read-Only Memory, EPROM), one-time programmable Read-Only Memory (OTPROM), electrically erasable rewritable Read-Only Memory (EEPROM), compact disc Read-Only Memory (Compact Disc Read-Only Memory, CD-ROM) or other optical disc Memory, magnetic tape Memory, or any other medium that can be used for carrying or storing data, which is readable by a computer.
In some embodiments, the at least one processor 32 may be comprised of an integrated circuit, such as a single packaged integrated circuit, or may be comprised of multiple integrated circuits packaged with the same or different functionality, including one or more central processing units (Central Processing unit, CPU), microprocessors, digital processing chips, graphics processors, combinations of various control chips, and the like. The at least one processor 32 is a Control Unit (Control Unit) of the terminal 3, connects the respective components of the whole terminal 3 using various interfaces and lines, and performs various functions of the terminal 3 and processes data, such as a train fire monitoring function, by running or executing programs or modules stored in the memory 31, and calling data stored in the memory 31.
In some embodiments, the at least one communication bus 33 is arranged to enable connected communication between the memory 31 and the at least one processor 32 or the like.
Although not shown, the terminal 3 may further include a power source (such as a battery) for supplying power to the respective components, and preferably, the power source may be logically connected to the at least one processor 32 through a power management device, so as to perform functions of managing charging, discharging, power consumption management, etc. through the power management device. The power supply may also include one or more of any of a direct current or alternating current power supply, recharging device, power failure detection circuit, power converter or inverter, power status indicator, etc. The terminal 3 may further include various sensors, bluetooth modules, wi-Fi modules, etc., which will not be described herein.
It should be understood that the embodiments described are for illustrative purposes only and are not limited to this configuration in the scope of the patent application.
The integrated units implemented in the form of software functional modules described above may be stored in a computer readable storage medium. The software functional modules described above are stored in a storage medium and include instructions for causing a computer device (which may be a personal computer, a terminal, or a network device, etc.) or a processor (processor) to perform portions of the methods described in the various embodiments of the invention.
In a further embodiment, in connection with fig. 2, the at least one processor 32 may execute the operating means of the terminal 3 as well as various installed applications (such as the train fire monitoring apparatus 20), program code, etc., such as the various modules described above.
The memory 31 has program code stored therein, and the at least one processor 32 can invoke the program code stored in the memory 31 to perform related functions. For example, each of the modules depicted in fig. 2 is program code stored in the memory 31 and executed by the at least one processor 32 to perform the functions of the modules for train fire monitoring purposes.
In one embodiment of the invention, the memory 31 stores a plurality of instructions that are executed by the at least one processor 32 to perform train fire monitoring functions.
Specifically, the specific implementation method of the above instruction by the at least one processor 32 may refer to the description of the relevant steps in the corresponding embodiment of fig. 1, which is not repeated herein.
In the several embodiments provided in the present invention, it should be understood that the disclosed apparatus, device and method may be implemented in other manners. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the modules is merely a logical function division, and there may be other manners of division when actually implemented.
The modules described as separate components may or may not be physically separate, and components shown as modules may or may not be physical units, may be located in one place, or may be distributed over multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional module in the embodiments of the present invention may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units can be realized in a form of hardware or a form of hardware and a form of software functional modules.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim concerned. Furthermore, it will be obvious that the term "comprising" does not exclude other elements or that the singular does not exclude a plurality. A plurality of units or means recited in the apparatus claims can also be implemented by means of one unit or means in software or hardware. The terms first, second, etc. are used to denote a name, but not any particular order.
Finally, it should be noted that the above-mentioned embodiments are merely for illustrating the technical solution of the present invention and not for limiting the same, and although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications and equivalents may be made to the technical solution of the present invention without departing from the spirit and scope of the technical solution of the present invention.
Claims (9)
1. A train fire monitoring method, the method comprising:
continuously acquiring a plurality of images in the running process of the train;
detecting a plurality of candidate contour areas in each image by using a YOLO target detection algorithm;
inputting the candidate contour areas into a pre-trained SVM model, and identifying whether a fire occurs in at least one candidate contour area through the SVM model;
When the SVM model recognizes that the fire occurs in at least one candidate contour area, judging whether the positions of the candidate contour areas with the fire are consistent in the plurality of images;
The judging whether the positions of the candidate outline areas with the fire are consistent in the plurality of images comprises the following steps: taking the candidate contour area with fire as a first target contour area, acquiring images with continuous preset numbers from the images, wherein the images with continuous preset numbers comprise the first target contour area, when the displacement of the first target contour area in the images with continuous preset numbers changes, determining that the relative position of the candidate contour area with fire in each image is inconsistent, and determining that the fire is on the train, or when the displacement of the first target contour area in the images with continuous preset numbers does not change, determining that the relative position of the candidate contour area with fire in each image is consistent, and determining that the fire is not on the train, wherein the images are acquired through an image acquisition device positioned on a support frame on the track side of the train;
and outputting a result of the fire on the train when judging that the positions of the candidate outline areas with the fire are inconsistent in each image.
2. The method of claim 1, wherein the training process of the SVM model comprises:
Acquiring a plurality of images of fire as a positive sample set and a plurality of images of red objects as a negative sample set;
Generating a positive and negative sample training set and a positive and negative sample testing set according to the obtained positive and negative sample set;
inputting the positive and negative sample training sets into an SVM for training to obtain an SVM model;
Inputting the positive and negative sample test set into the SVM model for testing to obtain a test passing rate;
If the test passing rate is greater than or equal to a preset passing rate threshold value, determining that the training of the SVM model is finished;
And if the test passing rate is smaller than the preset passing rate threshold, increasing the number of the positive and negative sample training sets, and carrying out training on the SVM model again.
3. The method of claim 2, wherein generating a positive and negative sample training set and a positive and negative sample testing set from the obtained positive and negative sample sets comprises:
and randomly selecting a first preset number of images from the positive and negative sample sets to serve as the positive and negative sample training set, and randomly selecting a second preset number of images to serve as the positive and negative sample testing set by adopting a random number generation algorithm.
4. The method of claim 1, wherein when no fire is identified by the SVM model as occurring in the candidate contour region; or when it is identified by the SVM model that a fire has occurred in at least one of the candidate contour regions, but the positions of the candidate contour regions where the fire has occurred are consistent in each image, the method further includes:
And outputting a result of no fire condition on the train.
5. The method of claim 1, wherein after said outputting the result of the train having a fire, the method further comprises:
sending alarm information to a train driver of the train;
and simultaneously, sending alarm information containing the locomotive number of the train to a dispatching room of a front station.
6. The method of any one of claims 1 to 5, wherein after the continuously acquiring a plurality of images during the train traveling, the method further comprises:
performing illumination or contrast normalization processing on the plurality of images;
and carrying out noise reduction treatment on the plurality of images subjected to the normalization treatment by adopting a bilateral filtering algorithm.
7. A train fire monitoring apparatus for implementing a train fire monitoring method as claimed in any one of claims 1 to 6, the apparatus comprising:
The acquisition module is used for continuously acquiring a plurality of images in the running process of the train;
the detection module is used for detecting a plurality of candidate contour areas in each image by using a YOLO target detection algorithm; the identification module is used for inputting the candidate contour areas into a pre-trained SVM model, and identifying whether a fire occurs in at least one candidate contour area or not through the SVM model;
The judging module is used for judging whether the positions of the candidate contour areas with the fire conditions in each image are consistent or not when the identification module identifies that the fire conditions occur in at least one candidate contour area through the SVM model; and the output module is used for outputting a result of the fire condition on the train when the judging module determines that the positions of the candidate outline areas with the fire condition in each image are inconsistent.
8. A terminal comprising a processor for implementing a train fire monitoring method according to any one of claims 1 to 6 when executing a computer program stored in a memory.
9. A computer readable storage medium having a computer program stored thereon, wherein the computer program when executed by a processor implements the train fire monitoring method according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910377069.9A CN110263622B (en) | 2019-05-07 | 2019-05-07 | Train fire monitoring method, device, terminal and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910377069.9A CN110263622B (en) | 2019-05-07 | 2019-05-07 | Train fire monitoring method, device, terminal and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110263622A CN110263622A (en) | 2019-09-20 |
| CN110263622B true CN110263622B (en) | 2024-09-06 |
Family
ID=67914349
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910377069.9A Active CN110263622B (en) | 2019-05-07 | 2019-05-07 | Train fire monitoring method, device, terminal and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110263622B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111613037B (en) * | 2020-04-30 | 2022-09-13 | 杭州拓深科技有限公司 | Method for reducing composite smoke sense false alarm based on intelligent algorithm |
| CN111986436B (en) * | 2020-09-02 | 2022-12-13 | 成都视道信息技术有限公司 | Comprehensive flame detection method based on ultraviolet and deep neural networks |
| CN112712076A (en) * | 2020-12-29 | 2021-04-27 | 中信重工开诚智能装备有限公司 | Visual positioning device and method based on label-free positioning |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101833838A (en) * | 2010-05-27 | 2010-09-15 | 王巍 | Large-range fire disaster analyzing and early warning system |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101334924B (en) * | 2007-06-29 | 2011-06-01 | 丁国锋 | Fire detection system and fire detection method thereof |
| KR101581162B1 (en) * | 2015-06-24 | 2015-12-31 | 전정희 | Automatic detection method, apparatus and system of flame, smoke and object movement based on real time images |
| CN108664970A (en) * | 2018-05-17 | 2018-10-16 | 淘然视界(杭州)科技有限公司 | A kind of fast target detection method, electronic equipment, storage medium and system |
| CN108875600A (en) * | 2018-05-31 | 2018-11-23 | 银江股份有限公司 | A kind of information of vehicles detection and tracking method, apparatus and computer storage medium based on YOLO |
-
2019
- 2019-05-07 CN CN201910377069.9A patent/CN110263622B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101833838A (en) * | 2010-05-27 | 2010-09-15 | 王巍 | Large-range fire disaster analyzing and early warning system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110263622A (en) | 2019-09-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110263623B (en) | Train climbing monitoring method, device, terminal and storage medium | |
| CN110390262A (en) | Video analysis method, apparatus, server and storage medium | |
| CN202793389U (en) | On-line routing inspection device for high-speed rail overhead line system | |
| CN110263622B (en) | Train fire monitoring method, device, terminal and storage medium | |
| CN108369645A (en) | Taxi operation monitoring method, device, storage medium and system | |
| CN108280435A (en) | Passenger abnormal behavior recognition method based on human body posture estimation | |
| CN113298059A (en) | Pantograph foreign matter intrusion detection method, device, computer equipment, system and storage medium | |
| CN105115605A (en) | Track train infrared detection system and detection method | |
| CN111814775A (en) | Target object abnormal behavior identification method, device, terminal and storage medium | |
| CN110264470B (en) | Freight train tarpaulin monitoring method, device, terminal and storage medium | |
| CN108337321A (en) | CBTC signal equipment inspection system and method based on video intelligent recognition | |
| CN117975371A (en) | Pedestrian safety early warning processing method, system, equipment and medium for high-speed rail platform | |
| CN110263718A (en) | Vehicle enters and leaves section detection method and detection system | |
| CN110524553A (en) | A kind of railway Cargo Inspection intelligent robot image analysis system | |
| CN111126771B (en) | System and method for predicting quality supervision and assurance of security inspectors' map recognition based on regional attention | |
| CN114493196A (en) | Data analysis method, system, electronic equipment and storage medium | |
| CN111055890B (en) | Intelligent detection method and detection system for railway vehicle anti-slip | |
| CN114118470A (en) | Intelligent management and control method and system for production and operation of full-automatic driving vehicle base | |
| CN109131453B (en) | Microcomputer interlocking console operation track tracking system | |
| CN113192019A (en) | Train pantograph detection method and system based on ultraviolet camera | |
| CN110276742B (en) | Train tail lamp monitoring method, device, terminal and storage medium | |
| CN103278603B (en) | Utilize number plate recognition technology monitoring motor vehicle to the system and method for environmental impact | |
| CN110287769B (en) | Train tail device monitoring method, system, terminal and storage medium | |
| CN213424011U (en) | Monitoring system for passenger identification of tour inspection channel | |
| CN117351682A (en) | Monitoring system for dynamically detecting characters by unmanned aerial vehicle |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |