Answers are grounded in this paper and cite their sources.
As society develops further, the demand for more complex information increases. With this in mind, the use of the naked eye becomes unsatisfactory, as the desired information may present itself in a manner that requires more advanced technology in order to capture it. Therefore, we need to understand how to capture and confirm what it is. For the reason, in this paper, we demonstrate first how to capture the video information, deal, and store each frame as an image then by comparing the current frame with its previous frame, finally return a histogram matching algorithm value (Bhattacharyya Coefficient) for their similarity, meanwhile determine whether it is a fast-moving object. As an object appears on a dynamic video, the first frame differential dichotomy processes the image and disregards the background. The object then appears again in a second image via the same processing while a background image is taken simultaneously. These steps allow for the corresponding regions to be analysed so that the object is identified via a collaboration of the two separate images. Afterwards, face recognition and skin pigmentation are extracted and detailed. By using Face ++ interface for face and skin recognition experiment, after some tests, in the case of a good picture resolution, the recognition rate is about 70%; the skin color extraction is one kind of relevant color extracting, when it has a well-lit background, so the effect will be better.
No takes yet. Share an insight, caveat, or question.
Li‐Hong Juang (2018) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: