Forest fire smoke image detection method based on total bounded variation

Abstract

The invention relates to the technical field of information processing, in particular to a forest fire smoke image detection method based on total bounded variation, which is characterized by comprising the following steps: 1, inputting an image in a video sequence; 2, adopting an image blocking method to score the TBV values of the blocked images, searching suspected smoke blocking areas from a blocking result image through two times of comparison of the TBV values, and finally, extracting the final suspected smoke areas through fusion clustering processing of feature data; 3, performing smoke motion feature analysis on the suspected smoke area by adopting an inter-frame difference method, if the suspected smoke area moves, judging the suspected smoke area as a fire smoke image, and giving a fire alarm; and if the suspected smoke area does not move, returning to video input to carry out the processing flow of the next image. According to the invention, errors caused by tedious calculation are avoided, the output result is accurate and stable, and the detection effect is good.

Description

Technical field [0001] The invention relates to the field of information processing technology, and in particular to a forest fire smoke image detection method based on total bounded variation. Background technique [0002] Forest fires, as the most harmful disaster to forests, have attracted the attention of countries all over the world; traditional detection technology based on smoke sensors has a small monitoring range, and the cost of laying it in large forest areas is high, and such sensors are prone to aging and reduced sensitivity. Currently, video systems are widely used in forest monitoring, and some fire events can be detected through manual monitoring. However, due to the vast forest area and numerous video images, manual monitoring is difficult. In recent years, with the continuous development and progress of research on image processing technology methods, video image processing and image recognition technology methods have become increasingly mature. Video smo...

AI Technical Summary

Problems solved by technology

In terms of smoke texture feature extraction, there are many related research methods. Among them, GLCM, LBP, Wavelet and other methods are the most widely used; Toreyin et al. use texture feature method to detect smoke; Flame and smoke real-time detection system; Tian et al. introduced LBP to extract smoke texture features; Yuan et al. proposed a smoke detection method based on pyramid histogram sequence; There are many false detections, and some methods have no obvious advantages in forest fire smoke detection

And feature research is limited to static low-level features such as color, outline, and motion, which is not enough to distinguish smoke from some suspected smoke objects, such as: clouds, fog, etc. There are few studies on high-level features and spatio-temporal features; forest fire smoke in different environments Presenting a variety of states, smoke visual feature extraction is the difficulty of smoke detection, how to build a stable and efficient feature extraction algorithm, and integrate static and dynamic information in the video has become the key to reducing smoke false detection

Images