An image segmentation approach of forest fire area based on aerial image

被引：0

作者：

Li, Sanqi ^{[1
]}

Li, Wenbin ^{[1
]}

Kan, Jiangming ^{[1
]}

Wang, Yutan ^{[1
]}

机构：

[1] School of Technology, Beijing Forestry University, Beijing 100083, China

来源：

Journal of Theoretical and Applied Information Technology | 2012年 / 46卷 / 01期

关键词：

Fire hazards - Deforestation - Antennas - Fire protection - Fires - Image segmentation;

D O I：

暂无

中图分类号：

学科分类号：

摘要：

In order to determine the forest fire danger rating quickly while finding it during the aviation of forest protection, it is important to calculate the fire area. Until recently the area has been mostly calculated by people's experience. With the mature of digital image processing technology, an assumption based on the forest fire area calculation of aerial images has been made, and the key point is how to identify and extract the fire area. In this paper, a extraction method on forest fire area based on the combination of color and gray level information has been proposed. First, the adaptive threshold should be found by applying the Otsu threshold method to a gray image. Then, the image analysis should be made by the RGB space component's characteristics of the target images' color features. Compared with traditional image segmentation methods, experiments show that the proposed method has low false points rate, good area consistency, accurate and clear edge, and can be used to extract and segment the fire area of aerial images. © 2005 - 2012 JATIT & LLS. All rights reserved.

引用

页码：207 / 211