One bit transforms have been proposed for low-complexity motion estimation by reducing the representation order to a single bit, and employing binary matching criteria. As a single bit is used to represent the image information it is important that the transformation captures local detail successfully to enable reasonable motion estimation accuracy. This paper elaborates on one-bit transformation approaches proposed in the literature by utilizing a wavelet domain transformation. The proposed approach is shown to improve the motion estimation accuracy of one-bit transform based low-complexity motion estimation.