Abstract: This paper studies the small moving target detection in the background of the sky,the ground surface or field.Wavelet analysis and inverted transform to image can remove the low-frequency components and speckle noise.We can obtain good results by adopting adaptive thresholding and using the maximum non-zero pixels segmentation,interpolation and region filling. Experimental results show that this method can effectively detect and locate small target and has high anti-noise performance.

Key words: remove noise, adaptive threshold, image segmentation, small target detection