CN107704818A - A kind of fire detection system based on video image - Google Patents
A kind of fire detection system based on video image Download PDFInfo
- Publication number
- CN107704818A CN107704818A CN201710900675.5A CN201710900675A CN107704818A CN 107704818 A CN107704818 A CN 107704818A CN 201710900675 A CN201710900675 A CN 201710900675A CN 107704818 A CN107704818 A CN 107704818A
- Authority
- CN
- China
- Prior art keywords
- fire
- background
- pixels
- foreground
- module
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/41—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/194—Segmentation; Edge detection involving foreground-background segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30181—Earth observation
Abstract
A kind of fire detection system based on video image, including image capture module, moving object detection module, fire disaster analyzing module and fire alarm module, described image acquisition module is used for the video image for obtaining monitored area, the moving object detection module carries out the renewal of background modeling and background model using clustering technique to the video image, and the background pixel in image is removed according to the background model, so as to obtain the foreground pixel in image, the fire disaster analyzing module is used to analyze the moving target detected and extracts the characteristic parameter of flame, and fire is determined whether according to the characteristic parameter, the fire alarm module is used to be alarmed when judging and having fire generation.Beneficial effects of the present invention are:A kind of fire detection system based on video image is provided, by gathering the video image of monitored area, and by being handled the video and being extracted the characteristic parameter of flame, so as to determine whether the generation of fire.
Description
Technical Field
The invention relates to the technical field of image processing, in particular to a fire detection system based on video images.
Background
In recent years, fire accidents frequently occur, and great loss is caused to the property of people, so that the prevention of the fire accidents is very important, the traditional fire monitoring method is to monitor the fire by using a fire detector, and the method cannot effectively early warn the fire in an open outdoor environment and large-area indoor places in time. With the development of scientific technology and the development of image processing technology becoming mature, people gradually apply the image processing technology to fire detection, and the method can quickly and effectively extract fire parameters in the images to judge the fire, so that the effective monitoring of the fire is realized.
In a fire detection system of technical vision, a main task of moving target detection research is to automatically detect a moving target in a video image, and the moving target image is deformed due to the existence of interferences such as complex background, illumination intensity change, moving target shielding and the like in the acquisition process of the video image, so that the moving target detection technology needs to be continuously researched in practical application.
Disclosure of Invention
In view of the above problems, the present invention is directed to a fire detection system based on video images.
The purpose of the invention is realized by the following technical scheme:
a fire detection system based on video images comprises an image acquisition module, a moving target detection module, a fire analysis module and a fire alarm module, wherein the image acquisition module is used for acquiring video images of a monitoring area, the moving target detection module adopts a clustering technology to perform background modeling and updating on the video images and removes background pixels in the video images according to a background model so as to obtain foreground pixels in the video images, the fire analysis module is used for analyzing the detected moving targets, extracting characteristic parameters of flames and judging whether fire occurs or not according to the characteristic parameters, and the fire alarm module is used for giving an alarm when the fire occurs.
The beneficial effects created by the invention are as follows: the fire detection system based on the video images is provided, and whether a fire occurs or not is judged by acquiring the video images of a monitoring area, processing the video and extracting the characteristic parameters of flame. In the moving target detection technology, the system improves the corresponding processing speed and the accuracy of a background model by carrying out background modeling and updating on a video image, and adopts the setting of a self-adaptive threshold value in a foreground detection module so as to obtain a higher foreground detection rate, so that the system has a higher target identification rate and can meet the requirement of actual processing.
Drawings
The invention is further described with the aid of the accompanying drawings, in which, however, the embodiments do not constitute any limitation to the invention, and for a person skilled in the art, without inventive effort, further drawings may be derived from the following figures.
FIG. 1 is a schematic structural view of the present invention;
fig. 2 is a schematic structural diagram of a moving object detection module according to the present invention.
Reference numerals:
an image acquisition module 1; a moving object detection module 2; a fire analysis module 3; a fire alarm module 4; a background modeling unit 21; an adaptive update unit 22; a foreground detection unit 23.
Detailed Description
The invention is further described in connection with the following examples.
Referring to fig. 1 and 2, the fire detection system based on a video image in this embodiment includes an image acquisition module 1, a moving target detection module 2, a fire analysis module 3, and a fire alarm module 4, where the image acquisition module 1 is configured to acquire a video image of a monitored area, the moving target detection module 2 performs background modeling and updating on the video image by using a clustering technique, and removes background pixels in the video image according to the background model, so as to obtain foreground pixels in the video image, the fire analysis module 3 is configured to analyze a detected moving target and extract characteristic parameters of flames, and determine whether a fire occurs according to the characteristic parameters, and the fire alarm module 4 is configured to alarm when a fire occurs.
The preferred embodiment provides a fire detection system based on video images, which judges whether a fire occurs by acquiring the video images of a monitoring area, processing the video and extracting the characteristic parameters of flame. In the moving target detection technology, the system improves the corresponding processing speed and the accuracy of a background model by carrying out background modeling and updating on a video image, and adopts the setting of a self-adaptive threshold value in a foreground detection module so as to obtain a higher foreground detection rate, so that the system has a higher target identification rate and can meet the requirement of actual processing.
Preferably, the moving object detecting module 2 includes a background modeling unit 21, an adaptive updating unit 22 and a foreground detecting unit 23, where the background modeling unit 21 performs background modeling on the video image by using an FCM clustering technique, the adaptive updating unit 22 is configured to perform adaptive updating on a background model after the modeling is completed, and the foreground detecting unit 23 is configured to remove background pixels in the video image, so as to obtain foreground pixels in the video image.
Preferably, the background modeling unit 21 performs background modeling on the video image by using FCM clustering technique, which uses an improved validity indicator to determine the optimal clustering number, and defines a validity indicator V new Then V is new The calculation formula of (c) is:
in the formula u ij Representing a sample x j Degree of membership belonging to the i-th class and having u ij ∈[0,1], v i Denotes the ith cluster center, v h Representing the h-th class of clustering centers, n representing the number of samples in the data set, and k representing the number of classes;
when V is new And when the minimum value is smaller, the optimal clustering result is corresponded.
The preferred embodiment provides a new effectiveness index, follows the clustering characteristics of tight intra-class and separation between classes, can correctly process the situation that a plurality of isolated data points exist or the classes are overlapped, and can still ensure the correctness of the clustering result when the data set is repeated.
Preferably, the adaptive updating unit 22 is configured to perform adaptive updating on the background model after the background modeling is completed, and define I t (x, y) is the gray level of the pixel (x, y) in the t-th frame, k is the number of clusters of the pixel, v i (x, y) represents the center value of the ith cluster, N i (x, y) represents the number of elements included in the ith cluster, and the pixel I is calculated t (x, y) and a certain cluster center value v i Distance D (I) of (x, y) t ,v i ):
D(I t ,v i )=d min (I t (x,y),v i (x,y))
a. Defining a clustering threshold p when D (I) t ,v i ) When rho is less than or equal to the pixel value, the pixel is classified into a corresponding class v i (x, y), and update parameters:
N i (x,y)=N i (x,y)+1
in the formula,. DELTA.t i Update time difference, ω, for ith cluster center i The weight of the ith cluster is taken as k is the number of clusters;
b. when D (I) t ,v i )&When gt, rho, a new classification v is created again k+1 (x, y), and setting relevant parameters as:
v k+1 (x,y)=I t (x,y)
N k+1 (x,y)=1
k=k+1
in the formula, v k+1 (x, y) is the central value of the (k + 1) th cluster, N k+1 (x, y) represents the number of elements in the (k + 1) th cluster.
The preferred embodiment provides an adaptive updating method for a background model based on an improved FCM cluster, which can dynamically modify, delete or create a new cluster to perform adaptive updating of the background model, and has higher calculation efficiency.
Preferably, the foreground detecting unit 23 is configured to remove background pixels from the video image to obtain foreground pixels from the video image, and an improved threshold selecting method is adopted to define a gray scale range of each pixel in the video image as [1,L [ ]]N is the number of pixels having a gray value of i i The probability of occurrence of a pixel having a gray value i is p i The pixels are divided into two categories by the gray value t, i.e. the background portion C 0 = {1, …, t } and foreground portion C 1 = { t +1, …, L }, the optimal segmentation threshold t is then * The calculation formula of (2) is as follows:
in the formula, t * For optimum threshold value, ω 0 (t) and ω 1 (t) probability of occurrence of pixels in the background part and the foreground part, respectively, mu 0 (t) and μ 1 (t) are the mean values of the pixels of the background part and the foreground part respectively,andthe variance of the pixels in the background and foreground parts, respectively, and μ (t) is the mean of all pixels in the video image, σ 2 (t) is the variance of the pixels in the video image.
The optimal embodiment improves an OTSU-based adaptive threshold selection method, replaces the mean value in the traditional algorithm with the variance, introduces the intra-class variance into the calculation process of the threshold selection function, effectively embodies the discrete degree of the gray scale, is suitable for the conditions of fuzzy moving targets and small difference between the foreground gray scale and the background gray scale in images, and improves the precision of the detection of the moving targets.
Finally, it should be noted that the above embodiments are only used for illustrating the technical solutions of the present invention, and not for limiting the protection scope of the present invention, although the present invention is described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions can be made on the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.
Claims (5)
1. A fire detection system based on video images is characterized by comprising an image acquisition module, a moving target detection module, a fire analysis module and a fire alarm module, wherein the image acquisition module is used for acquiring video images of a monitored area, the moving target detection module adopts a clustering technology to perform background modeling and updating on the video images and removes background pixels in the video images according to a background model so as to obtain foreground pixels in the video images, the fire analysis module is used for analyzing the detected moving targets, extracting characteristic parameters of flames and judging whether a fire occurs or not according to the characteristic parameters, and the fire alarm module is used for giving an alarm when the fire occurs.
2. The fire detection system based on video images as claimed in claim 1, wherein the moving object detection module comprises a background modeling unit, an adaptive updating unit and a foreground detection unit, the background modeling unit employs FCM clustering technique to perform background modeling of the video images, the adaptive updating unit is configured to perform adaptive updating on a background model after the modeling is completed, the foreground detection unit is configured to remove background pixels in the video images, thereby obtaining foreground pixels in the video images.
3. A method as claimed in claim 2The fire detection system based on the video images is characterized in that the background modeling unit carries out background modeling on the video images by adopting an FCM clustering technology, an improved validity index is adopted to determine the optimal clustering number, and the validity index is defined as V new Then V is new The calculation formula of (c) is:
in the formula u ij Represents a sample x j Degree of membership belonging to the i-th class and having u ij ∈[0,1], m represents a fuzzy weight index, v i Denotes the ith cluster center, v h Representing the h-th class of clustering centers, n representing the number of samples in the data set, and k representing the number of classes;
when V is new And when the minimum value is smaller, the optimal clustering result is corresponded.
4. A fire detection system according to claim 2, wherein the adaptive update unit is adapted to adaptively update the background model after background modeling is completed, defining I t (x, y) is the gray value of the pixel (x, y) at the t-th frame, v i (x, y) represents the center value of the ith cluster, N i (x, y) represents the number of elements included in the ith cluster, and the pixel value I is calculated t (x, y) and a certain cluster center value v i Distance D (I) of (x, y) t ,v i ):
D(I t ,v i )=d min (I t (x,y),v i (x,y))
a. Defining a clustering threshold p when D (I) t ,v i ) When the value is less than or equal to rho, the pixel value is classified into a corresponding class v i (x, y), and updating the parametersComprises the following steps:
N i (x,y)=N i (x,y)+1
in the formula,. DELTA.t i Update time difference, ω, for ith cluster center i The weight of the ith cluster is taken as k is the number of clusters;
b. when D (I) t ,v i )&When gt, rho, a new classification v is created again k+1 (x, y), and setting relevant parameters as:
v k+1 (x,y)=I t (x,y)
N k+1 (x,y)=1
k=k+1
in the formula, v k+1 (x, y) is the central value of the (k + 1) th cluster, N k+1 (x, y) represents the number of elements in the (k + 1) th cluster.
5. The fire detection system according to claim 2, wherein the foreground detection unit is configured to remove background pixels from the video image to obtain foreground pixels from the video image, and an improved threshold selection method is used to define a gray scale range of each pixel in the video image as [1,L ]]N is the number of pixels having a gray value of i i The probability of occurrence of a pixel having a gray value i is p i The pixels are divided into two categories by the gray value t, i.e. the background portion C 0 = {1, …, t } and foreground portion C 1 = { t +1, …, L }, then the optimal segmentation threshold t * The calculation formula of (2) is as follows:
in the formula, t * For optimum threshold value, ω 0 (t) and ω 1 (t) probability of occurrence of pixels in the background part and the foreground part, respectively, mu 0 (t) and μ 1 (t) are the mean values of the pixels of the background part and the foreground part respectively,andthe variance of the pixels in the background and foreground parts, respectively, and μ (t) is the mean of all pixels in the video image, σ 2 (t) is the variance of the pixels in the video image.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710900675.5A CN107704818A (en) | 2017-09-28 | 2017-09-28 | A kind of fire detection system based on video image |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710900675.5A CN107704818A (en) | 2017-09-28 | 2017-09-28 | A kind of fire detection system based on video image |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107704818A true CN107704818A (en) | 2018-02-16 |
Family
ID=61174618
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710900675.5A Pending CN107704818A (en) | 2017-09-28 | 2017-09-28 | A kind of fire detection system based on video image |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107704818A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110263654A (en) * | 2019-05-23 | 2019-09-20 | 深圳市中电数通智慧安全科技股份有限公司 | A kind of flame detecting method, device and embedded device |
CN112699801A (en) * | 2020-12-30 | 2021-04-23 | 上海船舶电子设备研究所(中国船舶重工集团公司第七二六研究所) | Fire identification method and system based on video image |
CN113538499A (en) * | 2021-08-20 | 2021-10-22 | 深圳市爱协生科技有限公司 | Image threshold segmentation method and device, computer equipment and storage medium |
CN114093116A (en) * | 2020-08-25 | 2022-02-25 | 中国电信股份有限公司 | Method, device and system for fire detection |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101834981A (en) * | 2010-05-04 | 2010-09-15 | 崔志明 | Video background extracting method based on online cluster |
CN102201146A (en) * | 2011-05-18 | 2011-09-28 | 中国科学技术大学 | Active infrared video based fire smoke detection method in zero-illumination environment |
CN104992447A (en) * | 2015-07-24 | 2015-10-21 | 安徽工业大学 | Automatic image detection method for moving microorganisms in sewage |
CN105788142A (en) * | 2016-05-11 | 2016-07-20 | 中国计量大学 | Video image processing-based fire detection system and detection method |
-
2017
- 2017-09-28 CN CN201710900675.5A patent/CN107704818A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101834981A (en) * | 2010-05-04 | 2010-09-15 | 崔志明 | Video background extracting method based on online cluster |
CN102201146A (en) * | 2011-05-18 | 2011-09-28 | 中国科学技术大学 | Active infrared video based fire smoke detection method in zero-illumination environment |
CN104992447A (en) * | 2015-07-24 | 2015-10-21 | 安徽工业大学 | Automatic image detection method for moving microorganisms in sewage |
CN105788142A (en) * | 2016-05-11 | 2016-07-20 | 中国计量大学 | Video image processing-based fire detection system and detection method |
Non-Patent Citations (1)
Title |
---|
张谢华: ""煤矿智能视频监控系统关键技术的研究"", 《中国博士学位论文全文数据库 工程科技Ⅰ辑》 * |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110263654A (en) * | 2019-05-23 | 2019-09-20 | 深圳市中电数通智慧安全科技股份有限公司 | A kind of flame detecting method, device and embedded device |
CN114093116A (en) * | 2020-08-25 | 2022-02-25 | 中国电信股份有限公司 | Method, device and system for fire detection |
CN112699801A (en) * | 2020-12-30 | 2021-04-23 | 上海船舶电子设备研究所(中国船舶重工集团公司第七二六研究所) | Fire identification method and system based on video image |
CN113538499A (en) * | 2021-08-20 | 2021-10-22 | 深圳市爱协生科技有限公司 | Image threshold segmentation method and device, computer equipment and storage medium |
CN113538499B (en) * | 2021-08-20 | 2024-01-19 | 深圳市爱协生科技股份有限公司 | Image threshold segmentation method, device, computer equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110516609B (en) | Fire disaster video detection and early warning method based on image multi-feature fusion | |
CN112069975A (en) | Comprehensive flame detection method based on ultraviolet, infrared and vision | |
CN101751744B (en) | Detection and early warning method of smoke | |
CN107704818A (en) | A kind of fire detection system based on video image | |
CN106128022B (en) | A kind of wisdom gold eyeball identification violent action alarm method | |
CN105701467A (en) | Many-people abnormal behavior identification method based on human body shape characteristic | |
CN111222478A (en) | Construction site safety protection detection method and system | |
CN106228709B (en) | A kind of wisdom gold eyeball identifies that one adds paper money alarm method and device | |
CN111860143B (en) | Real-time flame detection method for inspection robot | |
CN106127814A (en) | A kind of wisdom gold eyeball identification gathering of people is fought alarm method and device | |
CN107688793A (en) | A kind of outside transformer substation fire automatic monitoring method for early warning | |
CN108830161B (en) | Smog identification method based on video stream data | |
CN111476964A (en) | Remote forest fire prevention monitoring system and method | |
KR101196678B1 (en) | Real-time fire detection device and method | |
CN110942450A (en) | Multi-production-line real-time defect detection method based on deep learning | |
CN113963301A (en) | Space-time feature fused video fire and smoke detection method and system | |
CN112800975A (en) | Behavior identification method in security check channel based on image processing | |
CN115311623A (en) | Equipment oil leakage detection method and system based on infrared thermal imaging | |
WO2019088333A1 (en) | Method for recognizing human body activity on basis of depth map information and apparatus therefor | |
CN112926522B (en) | Behavior recognition method based on skeleton gesture and space-time diagram convolution network | |
CN107729811B (en) | Night flame detection method based on scene modeling | |
CN113192038A (en) | Method for identifying and monitoring abnormal smoke and fire in existing flame environment based on deep learning | |
CN113298077A (en) | Transformer substation foreign matter identification and positioning method and device based on deep learning | |
CN114283367B (en) | Artificial intelligent open fire detection method and system for garden fire early warning | |
CN113299034B (en) | Flame identification early warning method suitable for multiple scenes |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180216 |
|
RJ01 | Rejection of invention patent application after publication |