New Introduction: Search Method for Similar Images from Representative Video Frame Images
Daewon Jung · The Smart Computing Review · 2013
With rapid advances in computer and communications technology, the internet and video information is being used in a variety of areas?and the amount of this information is increasing exponentially. However, existing information analysis systems mainly depend on text, so it is not easy to differentiate video information. In order to solve this problem, we propose an algorithm that efficiently analyzes large amounts of video information and searches for similar images. We extract representative frames from a video via histogram analysis of the video information through low computational complexity of frame intervals. Then, we search for images with high similarity using information about colors and shapes. For obtaining color feature information in images and shape feature information, we use both hue?saturation?intensity and moment invariants.