The control of forest fire has formed into an autonomous and complex science. The employment of present day strategies of correspondences, quick air and ground transport and new sorts of fire fighting devices are decreasing the quantities of hectares of timberlands consumed every year. Anticipating fire conduct is a workmanship as much as it is a science. Indeed, even prepared fire fighters experience difficulty perusing fire conduct and anticipating flame's potential risk to property and lives. When they cannot, the outcome might just prompt catastrophe. The proposed system uses Computer Vision (CV), Machine Learning and Data Analytics to an effective usage. The video feed taken from the camera will be used. Fire in the feed will be quantized using feature extraction technique and segmentation is applied with machine learning algorithms, and by keeping the wind data in mind the fire coverage and likelihood of spreading can be found using data analytics.
Key words: Forest fire, Fire fighting, Grid searches CV, Image segmentation, Masking, Support vector machine algorithm.
|
