Forest Fire Detection based on Color Spaces Combination
Manual surveillance of inaccessible areas covered in trees and foliage is a difficult, labor-expensive and challenging task. On the other hand, early discovery of forest fires signs is essential in the preservation of the environment, therefore, the need for fast and efficient methods is very important. The characteristic color of the forest fire can be classified with the specific image processing techniques, especially during the late spring and summer when the color of the fire is easily distinguished from the color of trees and foliage, or generally flora. In this paper, a new method for the detection of forest fires is proposed. The method combines several predetermined and fuzzy criteria for image segmentation. A hybrid technique consists of a combination of different components of different color spaces.