Abstract: This paper proposes a mosaic method for UAV images based on filtering, which can eliminate ghosting and seams in the mosaicking process. The UAV images are decomposed into high-frequency components and low-frequency components by a Gaussian low-pass filter. Various mosaic schemes are designed to accomplish the stitching process. For the low-frequency components, weighted blending rule is utilized to achieve smoothing transition. An improved optimal seam searching strategy based on dynamic programming is introduced to guide the stitching of high-frequency components. The mosaic result is produced by linearly composing all stitching results from different components. Experimental results demonstrate the effectiveness of the proposed method in avoiding visible stitching seams and ghosting effects, especially in the cases of intensity differences.