<p>In this report, we present a prominence observed by New Vacuum Solar Telescope (NVST) in H&#945; wavelength. &#160;We use morphological approach to identify the rising or descending knots in the prominence. The rising knots are often found on the top of the prominence, while more knots are seen to descend from anywhere of the prominence.</p><p>The optical flow, referring to the apparent proper motion of a feature across the image plane, may be used to determine the velocity field from two images. The technique of local correlation tracking (LCT) optical flow has been widely used in the solar research. The Demon algorithm, which has been &#160;used to match medical image, performs image-to-image matching by determining the optical flow between two images. We have examined the performance of the two methods applying the H&#945; images, and we noted that the Demon algorithm outperforms traditional LCT &#160;methods.</p><p>The result of Demon optical flow allows us to estimate the velocity and acceleration of the moving knots. The descending speed of the knots near the solar surface is higher than that far away from the solar surface. This indicate that most of knots are more possible to descend across the horizontal magnetic field.</p>