Abstract
Forest fires can destroy millions of acres of land at shockingly fast speeds. The forest fire points identification algorithm is the most critical step in the forest fire monitoring process. Most traditional forest fire monitoring methods use fixed thresholds, ignoring background pixels, and have low recognition rates, which could lead to many problems, such as false reporting and low recognition rate. This paper proposes and tests an adaptive forest fire points identification algorithm using Himawari-8 data. By calculating the three-dimensional histogram of brightness temperature, an adaptive threshold that can automatically identify potential forest fire points is obtained. Based on this three-dimensional Otsu method, the contextual test algorithm has also been adopted to specify forest fire points. The experimental results show that the omission rate of the improved algorithm is about 10% lower than that of the previous algorithm in small-scale fire incidents. The improved algorithm can quickly and effectively extract fire point information, and it is also sensitive to small and low-temperature fires, which provides an efficient means for monitoring fire disasters.
Highlights
Protecting and developing forest resources are the signs of human needs and the progress of social civilization [1]
When forest fire points are identified in a traditional way, which might be influenced by regions and seasons, there are possible omissions or wrong classifications
In order to reduce the omission caused by fixed forest fire points pixel recognition threshold, this article first set up windows to monitor potential forest fire points according to different regions and different environmental conditions
Summary
Protecting and developing forest resources are the signs of human needs and the progress of social civilization [1]. Considering that the fire point information and the neighboring area information would bring a low false alarm rate, Giglio et al [3] put forward a method that the contextual fire point detection algorithm that applied to MODIS data This algorithm still uses a fixed potential threshold to identify potential fire points, which will greatly affect the results of forest fire points monitoring. This paper uses image segmentation technology — three-dimensional Otsu method — to automatically select the threshold of potential fire pixels, and uses the potential dynamic threshold to identify potential fire pixels, which is more adaptive and optimizes the existing contextual fire detection algorithm This algorithm solves the problem of wrong forest fire points detection caused by the fixed threshold, which is not suitable for certain areas. The detailed calculation process can be divided into four main steps: data preprocessing, identification of potential forest fire points, confirmation of potential forest fire points, and false forest fire point filtering
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
