TY - JOUR AU - Tran, Bao Quang AU - Hoa, Nguyen Thi PY - 2017/10/30 Y2 - 2024/03/28 TI - Approach to Detecting Forest Fire by Image Processing Captured from IP Cameras JF - European Journal of Applied Sciences JA - EJAS VL - 5 IS - 5 SE - Articles DO - 10.14738/aivp.55.3690 UR - https://journals.scholarpublishing.org/index.php/AIVP/article/view/3690 SP - 01 AB - <p>In this paper, the results show an algorithm to detect the presence of smoke and flame using image sequences captured by Internet Protocol (IP) cameras is represented. The important characteristics of smoke such as color, motion and growth properties are employed to detect fire. For the efficient smoke and fire detection in the captured images by the IP camera, a detection algorithm must operate directly in the Discrete Cosine Transform (DCT) domain to reduce computational weigh, avoiding a complete decoding process required for algorithms that operate in spatial domain. In order to assess the possibility and the accuracy of proposed algorithm, the author used the video sequences which are captured by IP camera from control forest fire at different spatial location and levels of fire intensity. Evaluation results illustrated the efficiency of the proposed algorithm in effectively detecting forest fires with accuracy at 97%.</p> ER -