The quality of ultrasonic images is usually influenced by speckle noises. To reduce the speckle noise, compounding techniques have been widely applied to improve image quality through signal averaging. In this paper, a 2-D ultrasonic motion compounding system is proposed that uses a hierarchical adaptive feature weighted motion estimation method to realign the frames before compounding. Each frame is first registered to the reference frame and then compounded to reduce the speckle noises. Several compounding strategies are then used to remove the unwanted degradation. Performance has been tested on both synthetic and in vivo clinical ultrasonic images. The experimental results show that the motion can be assessed accurately and the compounded images achieve good signal-to-noise ratio for improving the quality of ultrasonic images.