Cross-media analysis and reasoning fire detection algorithm

Abstract

The invention provides a cross-media analysis and reasoning fire detection algorithm, and the algorithm comprises fire scene identification, feature extraction, region segmentation, feature fitting, movement estimation and a plurality of acquisition devices; the feature fitting is used for identifying a movement path of a fire; the movement pre-estimation is based on the data of the feature fitting, and the movement path of the fire area is pre-estimated; each acquisition device acquires fire scene data for performing position identification on a fire scene; the area of the fire scene is recognized based on a feedforward neural recognition network, feature extraction is performed on the recognized area of the fire scene, and the range of fire movement is estimated. According to the method, the fire scene is identified by adopting the search ring, and nonlinear fitting is carried out by combining with the moving path of the fire behavior, so that the fire passing range of the fire scene can be accurately identified; meanwhile, analysis is carried out in combination with the eccentricity and the circularity, so that the ignition analysis of the fire scene is more accurate.

Description

technical field [0001] The invention relates to the technical field of fire rescue, in particular to a fire detection algorithm for cross-media analysis and reasoning. Background technique [0002] Most of the existing image-based fire detection methods are based on fixed cameras for fire detection and recognition. The scene is relatively fixed, and it has a certain detection effect on general indoor and large space fires. So far, it still cannot meet the actual application requirements. [0003] Such as CN101493980A a video flame detection method based on multi-feature fusion and Chinese patent CN101587622A forest fire detection and recognition method and equipment based on intelligent analysis of video images, use the method based on the mixed Gaussian model to segment the flame movement area, this method cannot Adapting to Fire Region Segmentation in Rapidly Changing Scenes. Chinese Patent CN102163358A A Smoke / Flame Detection Method Based on Video Image Analysis Althoug...

AI Technical Summary

Problems solved by technology

[0002] Most of the existing image-based fire detection methods are based on fixed cameras for fire detection and recognition. The scene is relatively fixed, and it has a certain detection effect on general indoor and large space fires. So far, it still cannot meet the actual application requirements

[0003] Such as CN101493980A a video flame detection method based on multi-feature fusion and Chinese patent CN101587622A forest fire detection and recognition method and equipment based on intelligent analysis of video images, use the method based on the mixed Gaussian model to segment the flame movement area, this method cannot Adapting to Fire Region Segmentation in Rapidly Changing Scenes

Chinese Patent CN102163358A A Smoke / Flame Detection Method Based on Video Image Analysis Although this patent adds RGB color restrictions when detecting flame color, it tries to detect a complete fire area, but it ignores the correlation between RGB channels and the richness of flame colors, it is easy to mistakenly detect some objects with similar colors to the fire area as forest fires

In addition, the Chinese patent CN102496234A is a forest fire early warning method, which uses infrared cameras to detect forest fires. This patent does not involve specific image processing algorithms

Chinese patent CN201885804U and Chinese patent CN201844880U use dual-band infrared to detect flames, Chinese patent CN202195883U and Chinese patent CN201191222 ultraviolet band fire detection, and Chinese patent CN202306757U belongs to multi-band compliance detection technology. Compared with visible light band detectors, they increase infrared , ultraviolet or multi-band sensors, the system is relatively complex and expensive, which limits its promotion

[0004] In order to solve the common problems in the field that fusion across media, prediction, poor recognition accuracy and easy misjudgment, etc., the present invention is made

Images