Scale Invariant Feature Transform (SIFT) is mainly invariance to various image transformation including rotation,scale and illumination changes. First, the video is converted into frames. key frames are extracted video. SIFT is mainly used for object detection.- Scale Invariant Feature Transform (SIFT) is mainly invariance to various image transformation including rotation,scale and illumination changes. First, the video is converted into frames. key frames are　extracted　video. SIFT is mainly used for object detection.