Fig. 1. Flowchart of fire point detection method

Fig. 2. Scatter plots of surface brightness temperature before and during the disaster in Qinyuan, Shanxi province

Fig. 3. Comparison chart of fitted background brightness temperature with different interpolation methods

Fig. 4. Histogram and probability distribution of background brightness temperature fitting in potential fire points
(a): Histogram and probability distribution AT fitting; (b): Histogram and probability distribution IDW fitting;(c): Histogram and probability distribution OK fitting; (d): Histogram and probability distribution SK fitting

Fig. 5. Results of forest fires detection of New Barhu Right Bannerin Inner Mongolia on April 21, 2019
(a): Results of AT; (b): Results of IDW; (c): Results of simple Kriging; (d): Results of ordinary Kriging

Fig. 6. Results of forest fires detection of Qinyuan, Shanxi province on March 31, 2019
(a): Results of AT; (b): Results of IDW; (c): Results of simple Kriging; (d): Results of ordinary Kriging

Table 1. Parameters of GaoFen-4 satellite

Table 2. Statistical table of adaptive threshold method and three interpolation methods about fitting bright temperature values

Table 3. GF-4/PMI Fire point detection of dual phase methods and single phase methods in New Barhu Right Bannerin Inner Mongolia

Table 4. GF-4/PMI Fire point detection of dual phase methods and single phase methods in Qinyuan, Shanxi Province

Table 5. The validation result of forest fire points detection in different methods