Abstract: Based on the analysis of the existing methods of image update, this paper proposes a new method for image update and fusion in an incremental and dynamic way. The method firstly locates the changing map sheets and areas with grid index maps and incrementally updates the local parts of original data with images of different phrases through mosaic technique. Then it conducts color balance manipulation to reduce color variation of different images. The image smooth process based on convolution and distance-inverse weighted method is also applied to solve the pixel mutation effects existed in the transition zone. This paper illustrates the method with a case study and corresponding analysis. The result shows that the method has high computational efficiency and good mosaic effect. The method exhibits satisfied practicability in local scope incremental image updating and successfully solves the dynamic update problems caused by limitation of capital, deficiency of technology and fast changing rates in local areas.