Research on Flame Image Edge Detection Algorithm Based on Mobile Robot Platform

The mine rescue robot can use the visual system to obtain the information of fire site, and real-time transmit the information and image of site. This paper builds the mobile robot platform, and obtained the fire image using the visual system of this mobile robot. Then an edge detection algorithm of flame image based on wavelet module maxima is proposed. This method is calculating the gradient module maxima of wavelet transform of flame image, and removing false edge by comparing with local gradient module maxima. This algorithm can implement availably edge detection from flame image with noise, which is a little amount of calculation, and better robust and real-time. The experimental results show this method could improve the accuracy and meet real-time requirements of flame image edge detection.