Fire Detection based on Smoke Image using Convolutional Neural Network (CNN)

Abstract

Forest fire is a natural disaster that is difficult to control and has a very wide scope that threatens forest ecosystems [1]. In Indonesia itself, forest area decreases every year, one of the causes of the reduction in forest area in Indonesia is forest fires and illegal logging. On this basis, a remote fire detection system is designed and can monitor large forest areas so that problems caused by fires can be minimized. The existing technology in computer vision is a combination of image processing and pattern recognition. The results show that the convolutional neural network has good performance in the field of image processing and obtains architectural optimization in its area. In the smoke image detection research, the accuracy results are very good, namely 99.72% using the Convolutional Neural Network method.