Flame Detection Algorithm Based on Image Processing Technology

Fig. 2. Comparison of results obtained by single-color model and proposed method. (a)(f)(k) Original images, (b)(g)(l) images obtained by RGB model; (c)(h)(m) images obtained by HSI model; (d)(i)(n) images obtained by Fuzzy Re-fusion of double-color spatial model; (e)(j)(o) images obtained by proposed method

Download full size

Fig. 3. Proportions of Cb and Cr component values. (a)(d) Flame; (b)(e) flame analogue; (e)(f) without flame

Download full size

Fig. 4. Basic structural diagram of PNN

Download full size

Fig. 5. ECM-PNN training fitting result

Download full size

Fig. 6. ECM-PNN test result

Download full size

Table 1. Comparison of test data of different algorithms
View table
Table 1. Comparison of test data of different algorithms
Algorithm T_PR /% T_NR /%
R-FCN+ResNet^[2] 98.25 96.3
Naive Bayes^[4] 94.41 98.15
FCM+SVM^[5] 94.78 98.68
BP Neural Network^[7] 99.1 97.3
GMHI+SVM^[9] 97.02 100
Traditional PNN 98 98
ECM-PNN 99.33 99.33

Tools

Get Citation

Copy Citation Text

Yong Tan, Linbo Xie, Hongwei Feng, Li Peng, Zhengdao Zhang. Flame Detection Algorithm Based on Image Processing Technology[J]. Laser & Optoelectronics Progress, 2019, 56(16): 161012

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites