Smoke detection method and system, computer readable storage medium and equipment

Abstract

The invention provides a smoke detection method which comprises the following steps: reading a new frame of near-infrared image, converting the near-infrared image into a gray level image, and zoomingthe gray level image to obtain a new image; filtering is carried out on each frame of the new image; caching the filtered new image; extracting a motion foreground area in the new image; performing morphological opening operation on the motion foreground area; detecting a connected region contour in the moving foreground region image; clustering the contours of the connected regions to obtain a plurality of key contour regions; judging the brightness of each key contour area, and filtering out an interference area with insufficient brightness according to the brightness; extracting LBP textural features of the key contour area meeting the brightness requirement; inputting the LBP texture feature of each key contour area into a trained SVM classifier to obtain a smoke area; outputting a minimum rectangular envelope according to the smoke area; and if the rectangular envelope of the smoke area is detected, giving out a prompt alarm.

Description

technical field [0001] The invention relates to the field of smoke detection, in particular to a smoke detection method and system based on near-infrared video images. Background technique [0002] Fire has the characteristics of growth, floating, sporadic, blindness, etc. It starts suddenly and spreads quickly without external interference, threatening the safety of people's lives and property all the time. For key areas such as factory equipment with poor heat dissipation and forest vegetation in hot climates, fires are extremely prone to occur, and once a fire occurs, it usually causes serious losses. In order to avoid or minimize the loss caused by fire, it is particularly important to prevent fire in advance. [0003] In the early stages of a fire, the appearance of smoke usually precedes open flames. Therefore, if the smog in the environment can be monitored in real time and given a timely warning through video surveillance, the losses caused by fire will be greatly ...

AI Technical Summary

Problems solved by technology

[0004] In view of the shortcomings of the prior art described above, the purpose of the present invention is to provide a smoke detection method and system based on near-infrared video images, which is used to solve the problems in the prior art due to the near-grayscale characteristics of images captured by near-infrared cameras. The problem that leads to low detection algorithm accuracy

Images