Describes the theory behind, special features and application of flame image inspection system, which makes the accurate flame inspection on a single flame nozzle possible so that the operator can see in the central control room the picture of real flame over every burner thereby providing a reliable inspection means to ensure safe operation of every power station boiler.

Judging boiler combustion stability based on flame image and fuzzy neural network;

Skeleton extraction of boiler flame images based on mathematical morphology;

DSP embedded type furnace flame image pickup-processing system;

Based on the characteristics of actual forest fire images,a flame segmentation algorithm of forest fire images based on segmentation strategy of color images is presented.

The paper presents the method of distinguish of fire image based on lifting wavelet transform,aiming at the real-time character of fire image and many fire characters in fire edge,combined image processing technology with wavelet analysis theory.