preview

Essay On Smoke Detection

Decent Essays

Arun Singh Pundir i [21] presents a novel and robust approach for smoke detection that employs Deep Belief Networks. The proposed technique is separated into three stages. In the pre-processing stage, the region of high movement is removed by background subtraction technique. During the next stage smoke pixel intensities are removed from the Red, Green and Blue and Luminance; Chroma: Blue; Chroma: Red colour spaces for foreground regions. Consequently, second characteristic which is based on texture is calculated for identifying smoke regions in which Local Extrema Co-occurrence Pattern, an improved version of local binary patterns are removed from dissimilar foreground regions which calculate not only texture of smoke but also power and …show more content…

Arduino Uno (R3) Board is utilized as the hardware platform of the system. It is very important to have the apps currently because its mobility and convenience to consumers. Moreover that, it is very importance as early detection of smoke which can avoid cost-effective, ecological and human-life endangering threats from fire hazard. In order to create this system more important and precious, several recommendations can be executing in the future. Zhiqiang Zhou et al. [24] proposed for long-distance wildfire smoke detection. While the long-distance wildfire smoke frequently moves slowly and lacks salient characteristics in the video, the detection is still a challenging issue. Unlike numerous traditional video based techniques that usually rely on the smoke colour or movement for initial smoke region segmentation, we apply the Maximally Stable Extremal Region (MSER) detection technique to remove local extremal regions of the smoke. This builds the initial segmentation of probable smoke region minimum dependent on the movement and colour data. Potential smoke regions are then chooses from all the probable regions with applying some static visual characteristics of the smoke, helping to reduce the non-smoke regions as several as possible. Once a potential smoke region is established, we keep tracking it by investigating the best-matched Extremal regions in the successive frames. At the same time, the propagating movements of the

Get Access