A flame detection method, system and flame detection device for monitoring videos

Abstract

The present invention discloses a method, system and device for flame detection of surveillance videos. The method includes: obtaining video data, and determining a candidate region set after preprocessing; extracting abstract features and high-dimensional histogram of oriented gradients (HOG) features from the candidate region set; finding the best candidate region from the candidate region set according to a logistic regression observation model, and determining the target position in the best candidate region; performing dimensionality reduction processing on the abstract features and high-dimensional HOG features of the target position according to a dimensionality reduction algorithm, and determining the weight of each-dimensional feature vector according to the feature contribution degree; and determining whether the target position is a flame according to the Cayley-Klein metric learning criterion and each-dimensional feature vector. In the embodiments of the present invention, flame detection is performed on video data through a machine learning method and a neural network algorithm, so as to solve the problem that traditional manual detection and identification of fires may lead to a more serious fire situation due to negligence.

Description

Technical Field

[0001] The present invention relates to the field of intelligent fire protection technology, and particularly to a method, a system and a flame detection device for monitoring video flame detection. Background Art

[0002] With the rapid development of computer technology, very large scale integrated circuit technology and communication technology, digital relay protection devices and other automatic devices have been widely used in substations. Since the automation level of substations is increasing day by day and gradually developing towards the unattended operation mode, higher requirements are put forward for the intelligence of monitoring equipment. When a fire breaks out in a substation, the monitoring equipment can issue an alarm in time, which can reduce the harm to personnel and circuits; at present, it mainly relies on manual real-time monitoring of the monitoring video. However, due to the long-term viewing of multiple monitors by staff, fatigue and inattention are inevitable. Once a fire breaks out, the staff cannot issue an alarm in time, missing the first rescue time and causing more losses. Therefore, it is necessary to solve the problem that the traditional manual detection and identification of fires may lead to more serious fire situations due to negligence. Summary of the Invention

[0003] In view of this, embodiments of the present invention provide a method, a system and a flame detection device for monitoring video flame detection, so as to solve the problem that the traditional manual detection and identification of fires may lead to more serious fire situations due to negligence.

[0004] According to a first aspect, an embodiment of the present invention provides a method for monitoring video flame detection, including: acquiring video data, and determining a candidate region set after preprocessing; extracting abstract features and high-dimensional histogram of oriented gradients (HOG) features from the candidate region set; finding the best candidate region from the candidate region set according to the logistic regression observation model, and determining the target position in the best candidate region; performing dimensionality reduction processing on the abstract features and high-dimensional HOG features of the target position according to a dimensionality reduction algorithm, and determining the weight of each-dimensional feature vector according to the feature contribution degree; and detecting the flame at the target position according to the Cayley-Klein metric learning criterion and each-dimensional feature vector.

[0005] Combined with the first aspect, in a first implementation manner of the first aspect, after detecting the flame at the target position according to the Cayley-Klein metric learning criterion and each-dimensional feature vector, it further includes: when it is detected that the target position is a flame, issuing a fire alarm.

[0006] Combined with the first aspect, in a second implementation manner of the first aspect, the acquiring video data, and determining a candidate region set after preprocessing includes: using a particle filter algorithm to perform local search on the video data to determine the candidate region set.

[0007] In combination with the first aspect, in the third implementation manner of the first aspect, extracting the abstract features and the high-dimensional histogram of oriented gradients (HOG) features from the candidate region set includes: extracting the abstract features from the candidate region through a meta-learning algorithm.

[0008] According to the second aspect, an embodiment of the present invention provides a flame detection device, including: a memory and a processor, which are communicatively connected to each other. The memory stores computer instructions, and the processor executes the computer instructions to execute the flame detection method for the surveillance video described in the first aspect or any implementation manner of the first aspect.

[0009] According to the third aspect, an embodiment of the present invention provides a flame detection system for a surveillance video, including: at least one camera device, at least one communication module, a server, and the flame detection device described in the second aspect; the at least one camera device is configured to collect the surveillance video of the surveillance environment, the at least one communication module is respectively connected to the server and the at least one communication module, and is configured to upload the surveillance video to the server, and the server is connected to the flame detection device and transmits the surveillance video to the flame detection device.

[0010] In combination with the third aspect, in the first implementation manner of the third aspect, the system further includes: at least one analog-to-digital conversion module and at least one positioning module; the positioning module is connected to the communication module, and is configured to collect the location information of the camera device and upload the location information to the server through the communication module; the analog-to-digital conversion module is respectively connected to the camera device and the communication module, and is configured to convert the surveillance video into video data and upload it to the server through the communication module.

[0011] In combination with the first implementation manner of the third aspect, in the second implementation manner of the third aspect, the system further includes: an alarm prompt device and a display device; the alarm prompt device is connected to the flame detection device and is configured to issue a fire alarm when the flame detection device detects a flame; the display device is connected to the server and is configured to display the surveillance video collected by the camera device and the corresponding location information of the camera device.

[0012] In combination with the third aspect, in the third implementation manner of the third aspect, the server includes: a storage module; the storage module is configured to store at least one of the video data received by the server.

[0013] Compared with the prior art, the present invention has the following beneficial effects: By using a meta-learning algorithm, a logistic regression observation model, a dimensionality reduction algorithm, and the Cayley-Klein metric learning criterion to monitor and identify flames in surveillance videos, it is possible to avoid the problem that traditional manual fire detection and identification may lead to a more serious fire situation due to negligence; and by implementing the flame detection system for surveillance videos of the present invention, staff can obtain the location of the fire and the size of the fire situation from the display device according to the alarm prompt of the alarm prompt device, and take corresponding measures in a timely manner, thereby being able to reduce the losses caused by the fire. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The features and advantages of the present invention will be more clearly understood by referring to the accompanying drawings. The drawings are schematic and should not be construed as limiting the present invention in any way. In the drawings:

[0015] Figure 1 A flowchart of the flame detection method for surveillance videos in an embodiment of the present invention is shown;

[0016] Figure 2 A structural block diagram of the flame detection device in an embodiment of the present invention is shown;

[0017] Figure 3 A flowchart of the flame detection system for surveillance videos in an embodiment of the present invention is shown. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0019] In the description of the present invention, it should be noted that the orientation or positional relationship indicated by the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. is based on the orientation or positional relationship shown in the accompanying drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as limiting the present invention. In addition, the terms "first", "second", and "third" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance.

[0020] In the description of the present invention, it should be noted that unless otherwise clearly specified and defined, the terms "installation", "connection", and "coupling" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the communication inside two components. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific situations.

[0021] In addition, the technical features involved in different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0022] An embodiment of the present invention provides a method for detecting flames in surveillance videos. This method is applicable to fire detection and recognition from surveillance videos. As Figure 1 shown, this method includes the following steps:

[0023] Step S101: Obtain video data, and determine a candidate region set after preprocessing; in practical applications, obtain the video data collected by a camera, and after preprocessing to reduce the influence of illumination and other adverse factors in the video data, determine the candidate region set of this video data.

[0024] Step S102: Extract abstract features and high-dimensional histogram of oriented gradients (HOG) features from the candidate region set; in practical applications, the abstract features of the candidate region set are extracted from 9 contrast-sensitive directions and 18 contrast-insensitive directions through a meta-learning algorithm. The dimension of this feature vector is 108, and at the same time, the high-dimensional histogram of oriented gradients (abbreviation: FHOG) features of the candidate region set are extracted.

[0025] Step S103: Find the best candidate region from the candidate region set according to the logistic regression observation model, and judge the target position in the best candidate region; in practical applications, after classifying the candidate region set using the logistic regression observation model, find the best candidate region and infer the target position in the best candidate region;

[0026] Step S104: According to the dimensionality reduction algorithm, perform dimensionality reduction processing on the abstract features and high-dimensional histogram of oriented gradients features of the target position, and determine the weight of each dimension feature vector according to the feature contribution degree; in practical applications, the dimensionality reduction algorithm using the principal component analysis method is used to perform dimensionality reduction mapping on the abstract features and FHOG features to 100 dimensions, and according to the contribution degree of the features, determine the weight of each dimension feature vector and perform weighted processing on the feature vectors.

[0027] Step S105: Determine whether the target position is a flame according to the Cayley-Klein metric learning criterion and each-dimensional feature vector. In practical applications, the distance between each-dimensional feature is judged based on the metric criterion to realize the detection and recognition of flames in video data.

[0028] By implementing the flame detection method for surveillance videos in the embodiments of the present invention, the surveillance videos are monitored and recognized for flames through the meta-learning algorithm, the logistic regression observation model, the dimensionality reduction algorithm, and the Cayley-Klein metric learning criterion, so as to avoid the problem that the traditional manual detection and recognition of fires may make the fire situation more serious due to negligence.

[0029] Optionally, in some embodiments of the present invention, after step S105 in the above embodiments, the method further includes: when it is determined that the target position is a flame, an alarm is issued; when a flame appears in the detected video data, an alarm is sent in time to remind the staff to take corresponding measures.

[0030] Optionally, in some embodiments of the present invention, during the execution of step S101, it includes: using the particle filter algorithm to perform local search on the video data to determine the candidate region set.

[0031] Optionally, in some embodiments of the present invention, during the execution of step S102, it includes: extracting abstract features from the candidate regions through the meta-learning algorithm.

[0032] The embodiments of the present invention also provide a flame detection device, as Figure 2 shown. The flame detection device may include a processor 51 and a memory 52, where the processor 51 and the memory 52 may be connected through a bus or other means, Figure 2 taking the connection through the bus as an example.

[0033] The processor 51 may be a central processing unit (CPU). The processor 51 may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. chips, or a combination of the above types of chips.

[0034] The memory 52, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs, non-transitory computer-executable programs, and modules, such as the program instructions / modules corresponding to the flame detection method for monitoring videos in the embodiments of the present invention. The processor 51 executes various functional applications and data processing of the processor by running the non-transitory software programs, instructions, and modules stored in the memory 52, that is, implements the flame detection method for monitoring videos in the above method embodiments.

[0035] The memory 52 may include a program storage area and a data storage area. Among them, the program storage area can store an operating system and application programs required for at least one function; the data storage area can store data created by the processor 51 and the like. In addition, the memory 52 may include a high-speed random access memory, and may also include non-transitory memories, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state storage devices. In some embodiments, the memory 52 may optionally include a memory remotely disposed relative to the processor 51, and these remote memories can be connected to the processor 51 through a network. Examples of the above networks include but are not limited to the Internet, enterprise intranets, local area networks, mobile communication networks, and combinations thereof.

[0036] The one or more modules are stored in the memory 52 and, when executed by the processor 51, execute the Figure 1 flame detection method for monitoring videos in the embodiments shown.

[0037] Specific details of the above flame detection device can be understood by referring to the Figure 1 corresponding related descriptions and effects in the shown embodiments, which will not be elaborated here.

[0038] The embodiments of the present invention further provide a flame detection system for monitoring videos, as shown in Figure 3As shown in the figure, the system includes: at least one camera device 1, at least one communication module 2, a server 3, and the flame detection device 5 in the above embodiments; the at least one camera device 1 is used to collect the surveillance video of the monitoring environment, the at least one communication module 2 is respectively connected to the server 3 and the at least one communication module 2, and is used to upload the surveillance video to the server 3, and the server 3 is connected to the flame detection device 5 and transmits the surveillance video to the flame detection device 5. In practical applications, camera devices 1 and communication modules 2 are set at different fire monitoring points. The camera device 1 can be a high-definition camera. The communication module 2 includes a wireless communication module (wireless communication methods such as WIFI, Bluetooth, GPRS, etc.) and / or a wired communication method (which can be a CAN bus or an RS485 bus). After the camera device 1 collects the surveillance video of the monitoring point, it uploads the surveillance video to the server 3 through the communication module 2, and the server 3 transmits the surveillance video to the flame detection device 5 for flame detection and recognition.

[0039] By implementing the flame detection system for surveillance videos in the embodiments of the present invention, the communication module 2 uploads the surveillance video collected by the camera device 1 to the server 3, and the server 3 transmits the surveillance video to the flame detection device 5 for flame recognition, thus being able to avoid the problem that the traditional manual detection and recognition of fires may lead to a more serious fire situation due to negligence.

[0040] Optionally, in some embodiments of the present invention, the flame detection system for surveillance videos in the above embodiments further includes: at least one analog-to-digital conversion module 4 and at least one positioning module 6; the positioning module 6 is connected to the communication module 2 and is used to collect the location information of the camera device 1 and upload the location information to the server 3 through the communication module 2; the analog-to-digital conversion module 4 is respectively connected to the camera device 1 and the communication module 2 and is used to convert the surveillance video into video data and upload it to the server 3 through the communication module 2. In practical applications, the positioning module 6 can be a GPS positioning chip, which is installed together with each camera device 1 to position each camera device 1.

[0041] Optionally, in some embodiments of the present invention, the flame detection system for the monitoring video in the above embodiments further includes: an alarm prompting device 7 and a display device 8; the alarm prompting device 7 is connected to the flame detection device 5 and is used to issue a fire alarm when the flame detection device 5 detects a flame; the display device 8 is connected to the server 3 and is used to display the monitoring video collected by the camera device 1 and the location information corresponding to the camera device 1. In practical applications, the alarm prompting device 7 includes at least one of a sound prompting device and a light prompting device to prompt relevant staff through sound and light; the display device 8 can be a high-definition display screen. When the flame detection device 5 monitors a flame, the display device 8 displays the location information of the camera device 1 where the fire occurs and the monitoring video collected by the camera device 1. Staff can judge the size of the fire situation based on the monitoring video and take corresponding measures in a timely manner.

[0042] Optionally, in some embodiments of the present invention, the server 3 in the above embodiments includes: a storage module; the storage module is used to store at least one video data received by the server 3.

[0043] Those skilled in the art can understand that to implement all or part of the processes in the above method embodiments, it can be completed by instructing relevant hardware through a computer program. The program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the above method embodiments. Among them, the storage medium can be a magnetic disk, an optical disc, a read-only memory (ROM), a random access memory (RAM), a flash memory, a hard disk drive (HDD), or a solid-state drive (SSD), etc.; the storage medium can also include a combination of the above types of memories.

[0044] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A flame detection method for monitoring videos, characterized in that, Including: Obtain video data, and determine a set of candidate regions after preprocessing; Extract abstract features and high-dimensional histogram of oriented gradients (HOG) features from the set of candidate regions; Find the best candidate region from the set of candidate regions according to a logistic regression observation model, and determine the target position in the best candidate region; According to a dimensionality reduction algorithm, perform dimensionality reduction processing on the abstract features and high-dimensional HOG features of the target position, and determine the weight of each-dimensional feature vector according to the feature contribution degree; Use a dimensionality reduction algorithm in the form of principal component analysis to perform dimensionality reduction mapping of the abstract features and FHOG features to 100 dimensions, determine the weight of each-dimensional feature vector according to the feature contribution degree, and perform weighted processing on the feature vectors; Detect the flame at the target position according to the Cayley-Klein metric learning criterion and each-dimensional feature vector; Extract abstract features and high-dimensional HOG features from the set of candidate regions, including: Extract the abstract features from the candidate regions through a meta-learning algorithm.

2. The flame detection method for surveillance video according to claim 1, characterized in that, When detecting the flame at the target position according to the Cayley-Klein metric learning criterion and each-dimensional feature vector, it further includes: When the flame at the target position is detected, issue a fire alarm.

3. The flame detection method for surveillance videos according to claim 1, characterized in that, The obtaining of the video data and determining the set of candidate regions after preprocessing includes: Use a particle filter algorithm to perform local search on the video data to determine the set of candidate regions.

4. A flame detection device, characterized in that, Including: A memory and a processor, which are communicatively connected to each other. The memory stores computer instructions, and the processor executes the computer instructions to execute the method for detecting flame in a surveillance video according to any one of claims 1-3.

5. A flame detection system for monitoring videos, characterized in that, Including: At least one camera device, at least one communication module, a server, and the flame detection device according to claim 4 above; The at least one camera device is used to collect a surveillance video of a surveillance environment. The at least one communication module is respectively connected to the server and the at least one communication module, and is used to upload the surveillance video to the server. The server is connected to the flame detection device and transmits the surveillance video to the flame detection device.

6. The flame detection system according to claim 5, wherein, It further includes: At least one analog-to-digital conversion module and at least one positioning module; The positioning module is connected to the communication module, and is used to collect the location information of the camera device and upload the location information to the server through the communication module; The analog-to-digital conversion module is respectively connected to the camera device and the communication module, and is used to convert the surveillance video into video data and upload it to the server through the communication module.

7. The flame detection system according to claim 6, wherein It further includes: An alarm prompt device and a display device; The alarm prompt device is connected to the flame detection device, and is used to issue a fire alarm when the flame detection device detects a flame; The display device is connected to the server, and is used to display the surveillance video collected by the camera device and the location information corresponding to the camera device.

8. The flame detection system according to claim 5, wherein The server includes: a storage module; The storage module is used to store at least one piece of the video data received by the server.

Images