Computer Network System for Fire Flame detection from Digital Video using Fractal Geometry

Abstract

This research aims to build a computer network system that works on detecting and locating flame by using fractal geometry in the digital video through building the network system in a server-client structure through which the characteristics of both the server and the clients are determined. A terminal stations were adapted to receive the digital video, and then divided into set of frames to be treated later on individually using fractal geometry (2 dimension variation algorithm to create fractal dimensions matrix). Studying and analyzing the value of the fractal dimensions of each frame, to identify the flame region by recognizing the common properties. Applying the suggested algorithm on Various samples exceeding 300 colored digital images containing flames with different spectra have been selected. Those images are taken by digital cameras or downloaded from the internet. On the other hand, the application on digital video applied and exceed 100 sample. Results show that using fractal dimension is an active and encouraging way in determining the characteristics of flame with its all spectra. No any negative evidence was recorded on all samples which exceed 100 sample from the digital video during the interval of the application test. Results also show the harmony and consistency in transferring process through the protocols that have been dealt with.

Keywords

• digital video processing
• fractal geometry
• fire flame
• computer network