CN-121982823-A - Fire monitoring method, device and related products

CN121982823ACN 121982823 ACN121982823 ACN 121982823ACN-121982823-A

Abstract

The embodiment of the invention provides a fire monitoring method, a device and related products, wherein the method comprises the steps of obtaining a plurality of images acquired in a region to be monitored within a preset time range, wherein the preset time range comprises a time range between a current time and a historical time before the current time, obtaining motion characteristics of pixel points in each image, determining images containing the motion region in each image based on the motion characteristics of the pixel points in each image, taking the images containing the motion region as characteristic images, determining whether the region to be monitored has fire or not based on the characteristic images, and determining fire regions in the region to be monitored and fire grades corresponding to the fire regions based on the characteristic images under the condition that the region to be monitored has fire. According to the embodiment of the invention, the accuracy of identifying the fire area can be improved, and the accuracy of monitoring the fire can be improved.

Inventors

XIE QIN
GUO XUANJIANG
ZHENG HAIPENG
LI WENSI
LI HAICHUAN

Assignees

中国移动(浙江)创新研究院有限公司
中国移动通信集团浙江有限公司
中国移动通信集团有限公司
浙江移动数智科技有限公司
中国移动通信集团江苏有限公司
中国移动通信集团四川有限公司

Dates

Publication Date: 20260505
Application Date: 20251231

Claims (10)

1. A method of monitoring a fire, comprising: acquiring a plurality of images acquired from a region to be monitored in a preset time range, wherein the preset time range comprises a time range between a current time and a historical time before the current time; Acquiring motion characteristics of pixel points in each image, determining an image containing a motion area in each image based on the motion characteristics of the pixel points in each image, and taking the image containing the motion area as a characteristic image; And determining whether a fire exists in the area to be monitored or not based on each characteristic image, and determining the fire area in the area to be monitored and the fire grade corresponding to the fire area based on each characteristic image under the condition that the fire exists in the area to be monitored.
2. The method according to claim 1, wherein the determining an image including a motion region in each image based on the motion characteristics of the pixel points in the each image, and taking the image including the motion region as a feature image, comprises: determining at least one first target pixel point in each image as a seed pixel point based on the motion characteristics of the pixel points in each image, wherein the motion speed of the first target pixel point is not 0; Performing a growing operation on the seed pixel points, the growing operation comprising: Calculating the absolute value of the difference of the motion directions between the seed pixel points and the corresponding pixel points to be grown, judging whether the absolute value of the difference of the motion directions corresponding to the pixel points to be grown does not exceed a preset absolute value threshold value for each pixel point to be grown, if so, taking the pixel points to be grown as the pixel points to be grown, and merging the pixel points to be grown into the area where the seed pixel points are positioned to obtain a motion sub-area; repeating the growing operation for each image until the growing pixel point does not exist in the image, and taking a set of a plurality of motion subareas obtained after repeating the growing operation as a motion area in the image; the image containing the motion region is taken as a characteristic image.
3. The method according to claim 1, wherein the determining whether the area to be monitored has a fire based on the feature images, and determining the fire area in the area to be monitored based on the feature images if the area to be monitored has a fire, comprises: Judging whether the motion area contained in the feature image surrounds a closed area or not according to each feature image, and if so, determining the number of the closed areas in the feature image; when only a closed region surrounded by the motion region exists in the characteristic image, determining the closed region as an initial flame region; when other closed areas except the enclosed closed areas exist in the characteristic image, determining flame characteristic degrees corresponding to the closed areas, determining the closed area corresponding to the flame characteristic degree with the largest value in the flame characteristic degrees as an initial flame area, and determining the closed area adjacent to the initial flame area as an initial smoke area; And determining a fire area in the area to be monitored based on the initial flame area and the initial smoke area corresponding to each characteristic image.
4. The method of claim 3, wherein said determining a flame signature for each of said closed regions comprises: Determining a second target pixel point meeting a preset flame point screening condition in the closed region as a flame point according to each closed region, wherein the flame point screening condition comprises that the gray value is larger than the gray value of other pixels in the closed region, and the minimum distance between the gray value and the edge pixel point of the closed region is larger than the minimum distance between other pixels in the closed region and the edge pixel point of the closed region; Determining a target edge pixel point in each edge pixel point of the closed region according to each closed region, and taking a connecting line between the flame center point and the target edge pixel point as a flame characteristic line, wherein the distance between the target edge pixel point and the flame center point is larger than the distance between each edge pixel point except the target edge pixel point in the closed region and the flame center point; and determining flame characteristic degrees corresponding to the closed areas based on the gray values of the third target pixel points and the number of the third target pixel points on the flame characteristic lines for each closed area.
5. A method according to claim 3, wherein said determining a fire zone in said zone to be monitored based on said initial flame zone and said initial smoke zone corresponding to each of said feature images comprises: determining a characteristic region corresponding to each characteristic image based on the initial flame region and the initial smoke region corresponding to each characteristic image; Determining flame edge areas corresponding to the characteristic images based on the characteristic areas; and determining a fire area in the area to be monitored based on the flame edge area corresponding to each characteristic image and the initial flame area corresponding to each characteristic image.
6. The method of claim 5, wherein the determining a flame edge region corresponding to each of the feature images based on each of the feature regions comprises: Calculating the matching degree between the initial flame areas corresponding to every two adjacent characteristic images to obtain an initial flame area matching degree sequence; Identifying mutation points in the initial flame region matching degree sequence by using a mutation point detection algorithm, and determining a characteristic image corresponding to the mutation points as a mutation image; And aiming at each characteristic image, acquiring a mutation image with acquisition time adjacent to the acquisition time of the characteristic image as a target mutation image, determining the confidence coefficient of each pixel point in a characteristic area corresponding to the characteristic image as a flame edge area pixel point based on the target mutation image, and determining the flame edge area corresponding to the characteristic image based on the confidence coefficient.
7. The method of claim 6, wherein said calculating a degree of matching between the initial flame regions corresponding to each two adjacent feature images comprises: determining the initial flame region and flame center points corresponding to the initial flame region in each of the two adjacent characteristic images; Calculating maximum distance and minimum distance between the flame center point and the edge pixel point of the initial flame region as shape characteristic parameters for each characteristic image, and calculating gray level change degree between every two adjacent pixel points on a flame characteristic line corresponding to the characteristic image; Determining a third target pixel point and a fourth target pixel point in each pixel point on the flame characteristic line, dividing the flame characteristic line into a plurality of line segments based on the third target pixel point and the fourth target pixel point, and respectively calculating a pixel point gray average value corresponding to each line segment; And calculating the matching degree between the initial flame areas corresponding to every two adjacent feature images based on the shape feature parameters corresponding to the feature images and the pixel point gray average value corresponding to each line segment.
8. The method of claim 1, wherein said determining a fire rating of said fire zone comprises: determining a fire grade of the fire zone based on the flame flicker frequency, the degree of matching between the target flame zone in the first target feature image and the target flame zone in the second target feature image, and the area of the target flame zone in the second target feature image; the flame flicker frequency is determined based on a mutation image obtained by detecting mutation points of each feature image, the first target feature image is the first feature image in the plurality of images, the second target feature image is the feature image obtained at the current moment, and the target flame area is determined based on an initial flame area and a flame edge area in the corresponding feature images.
9. A fire monitoring device, the device comprising: The first acquisition module is used for acquiring a plurality of images acquired from the area to be monitored within a preset time range, wherein the preset time range comprises a time range between the current time and a historical time before the current time; The second acquisition module is used for acquiring the motion characteristics of the pixel points in each image, determining an image containing a motion area in each image based on the motion characteristics of the pixel points in each image, and taking the image containing the motion area as a characteristic image; The fire determining module is used for determining whether the fire exists in the to-be-monitored area or not based on the characteristic images, and determining the fire area in the to-be-monitored area and the fire grade corresponding to the fire area based on the characteristic images under the condition that the fire exists in the to-be-monitored area.
10. An electronic device, comprising a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory are in communication with each other through the bus, the memory is used for storing a computer program, and the processor is used for executing the program stored on the memory to realize the steps of the fire monitoring method according to any one of claims 1-8.

Description

Fire monitoring method, device and related products Technical Field The invention relates to the field of image processing, in particular to a fire monitoring method, a fire monitoring device and related products. Background In an actual fire monitoring scene, the prior art mostly depends on a traditional machine learning classification model to realize fire identification, however, the problem of low accuracy of fire area identification exists in the mode. Therefore, a fire monitoring method is needed to greatly improve the accuracy of fire area identification and the accuracy of fire monitoring. Disclosure of Invention The embodiment of the invention aims to provide a fire monitoring method, a fire monitoring device and related products, so as to solve the problem of low fire monitoring accuracy in the related technology. In order to solve the technical problems, the embodiment of the invention is realized as follows: in a first aspect, an embodiment of the present invention provides a fire monitoring method, including: acquiring a plurality of images acquired from a region to be monitored in a preset time range, wherein the preset time range comprises a time range between a current time and a historical time before the current time; Acquiring motion characteristics of pixel points in each image, determining an image containing a motion area in each image based on the motion characteristics of the pixel points in each image, and taking the image containing the motion area as a characteristic image; And determining whether a fire exists in the area to be monitored or not based on each characteristic image, and determining the fire area in the area to be monitored and the fire grade corresponding to the fire area based on each characteristic image under the condition that the fire exists in the area to be monitored. In a second aspect, an embodiment of the present invention provides a fire monitoring apparatus, including: The first acquisition module is used for acquiring a plurality of images acquired from the area to be monitored within a preset time range, wherein the preset time range comprises a time range between the current time and a historical time before the current time; The second acquisition module is used for acquiring the motion characteristics of the pixel points in each image, determining an image containing a motion area in each image based on the motion characteristics of the pixel points in each image, and taking the image containing the motion area as a characteristic image; The fire determining module is used for determining whether the fire exists in the to-be-monitored area or not based on the characteristic images, and determining the fire area in the to-be-monitored area and the fire grade corresponding to the fire area based on the characteristic images under the condition that the fire exists in the to-be-monitored area. In a third aspect, an embodiment of the present invention provides an electronic device, including a processor, a communication interface, a memory, and a communication bus, where the processor, the communication interface, and the memory complete communication with each other through the bus, where the memory is configured to store a computer program, and where the processor is configured to execute the program stored in the memory, to implement the steps of the fire monitoring method according to the first aspect. In a fourth aspect, embodiments of the present invention provide a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the fire monitoring method according to the first aspect. In a fifth aspect, embodiments of the present invention provide a computer program product comprising a computer program which, when executed by a processor, implements the steps of the fire monitoring method as described in the first aspect. As can be seen from the technical solution provided in the foregoing embodiment of the present invention, in the embodiment of the present invention, first, a plurality of images acquired in a to-be-monitored area within a preset time range are acquired, where the preset time range includes a time range between a current time and a history time before the current time, then, motion features of pixels in each image are acquired, an image including a motion area is determined in each image based on the motion features of the pixels in each image, and the image including the motion area is used as a feature image, then, based on each feature image, whether a fire exists in the to-be-monitored area is determined, and in case that a fire exists in the to-be-monitored area is determined, a fire area in the to-be-monitored area and a fire level corresponding to the fire area are determined based on each feature image. Therefore, according to the embodiment of the invention, the characteristic images comprising the motion areas can be determined in each image accord