Adaptive brightness segmentation-based fire video image analysis algorithm

Abstract

The invention provides an adaptive brightness segmentation-based fire video image analysis algorithm. A video shot by a monitoring device around a fire scene is copied to a computer for performing image processing; the computer preliminarily performs frame-by-frame analysis retrieval on video images to automatically find a region with a brightness change in the images and records the region; a first frame of the video captured from the monitoring device is taken as an initial background, continuous adjacent frames are subjected to difference operation to obtain an absolute value of a brightness difference of two frames of images, and a frame difference is subjected to binarization processing to obtain a binary image; a binary image of a video monitoring region part is subjected to frame-by-frame analysis; the processed binary image is subjected to adaptive threshold segmentation by utilizing an OTSU algorithm; a boundary in the binary image is subjected to linear fitting to obtain an accurate brightness segmentation line; and the brightness segmentation line displayed in real time and automatically extracted from the video is analyzed to determine a fire breaking point and a spreaddirection of a fire, so that fire investigators can perform quick analysis conveniently and the working efficiency is greatly improved.

Description

technical field [0001] The invention provides a fire video image analysis algorithm based on adaptive brightness segmentation, which belongs to the field of computer vision and is essentially an image segmentation problem. [0002] Specifically, the investigators of the fire cause of the fire brigade use the video recordings taken by the monitoring equipment around the fire scene to perform image processing on the computer, automatically extract the shadow and brightness dividing line that is conducive to the analysis of the fire site, and give the trend of fire spread. Therefore, it is helpful to improve the efficiency of fire site inspection, and it is convenient to trace the source of the fire. Background technique [0003] In today's society, fire has always been one of the major disasters faced by human beings, causing a large number of human casualties and property losses. With the development of science and technology, fire early warning technology and fire protectio...

AI Technical Summary

Problems solved by technology

However, due to the long video time and the large amount of data, it needs to consume a lot of manpower and time

In addition, since many monitoring equipments are installed in locations where only the reflection of fire light around the fire site can be observed, in the early stage of a fire, the fire light will irradiate the surrounding environment to a certain extent, but due to the occlusion of objects, there will be a dividing line between shadow and light , these dividing lines are valuable for inferring the location of the fire and the trend of fire spread

However, in the early stage of combustion, the brightness change is not obvious, and it is difficult for the human eye to detect and identify it. Fire investigators often miss these effective information, which affects the effect and efficiency of fire investigation.

This kind of manual manual frame-by-frame observation of whether there is a brightness dividing line is time-consuming and cannot be recorded in real time, and cannot guide the fire investigation work.

Images