불꽃 감지를 위한 임베디드 시스템에 적합한 딥러닝 구조

Abstract

In this paper, we propose a deep learning structure suitable for embedded system. The flame detection process of theproposed deep learning structure consists of four steps：flame area detection using flame color model, flame imageclassification using deep learning structure for flame color specialization,  × cell separation in detected flame area, flameimage classification using deep learning structure for flame shape specialization. First, only the color of the flame isextracted from the input image and then labeled to detect the flame area. Second, area of flame detected is the input ofa deep learning structure specialized in flame color and is classified as flame image only if the probability of flame classat the output is greater than 75%. Third, divide the detected flame region of the images classified as flame images lessthan 75% in the preceding section into  × units. Fourthly, small cells divided into  × units are inserted into theinput of a deep learning structure specialized to the shape of the flame and each cell is judged to be flame proof andclassified as flame images if more than 50% of cells are classified as flame images. To verify the effectiveness of theproposed deep learning structure, we experimented with a flame database of ImageNet. Experimental results show that theproposed deep learning structure has an average resource occupancy rate of 29.86% and an 8 second fast flame detectiontime. The flame detection rate averaged 0.95% lower compared to the existing deep learning structure, but this was theresult of light construction of the deep learning structure for application to embedded systems. Therefore, the deep learningstructure for flame detection proposed in this paper has been proved suitable for the application of embedded system.