A Trous wavelet real-time image fusion algorithms is presented based on DM642 fusion System. The fusion system is based on TI DM642 as its kernel processor，a latest high performance fixed-point DSP designed for video and image processing. It is built through the high operation speed of 600MHz

the high data transfer rates with EDMA and EMIF

the configurable video ports capable of glueless I/F to video decoder/encoder

and the dual-band image fusion system platform is constructed. A Trous wavelet transform is used to perform a decomposition of visible image in several scales in the system. The details extracted from which are fused into the infrared image. The fused images show the contribution of the visible image for discriminating lines and edges

while the infrared image provides an information on the detection of point targets. The simulated results show that dual-band image fusion algorithms cost 39.46ms

and fulfil 25 frame/s.