发明名称 Information processing device, information processing method, and recording medium
摘要 An image inputter inputs image data in which a user has been captured. An extractor extracts a region corresponding to a hand of the user included in the image data input by the image inputter. A determinator determines whether or not the region corresponding to the hand of the user extracted by the extractor is in a stationary state. If it is determined by the determinator that the region corresponding to the hand of the user is in a stationary state, a recognizer recognizes the shape of the region corresponding to the hand of the user extracted by the extractor. Otherwise, if it is determined by the determinator that the region corresponding to the hand of the user is not in a stationary state, the recognizer recognizes the movement of the region corresponding to the hand of the user extracted by the extractor. An inputter inputs data associated with the shape or the movement recognized by the recognizer.
申请公布号 US9367732(B2) 申请公布日期 2016.06.14
申请号 US201214112294 申请日期 2012.04.27
申请人 NEC SOLUTION INNOVATORS, LTD. 发明人 Kurokawa Takafumi;Tanimura Ryohtaroh
分类号 G06K9/00;G06F3/00;G06F3/01;G06F3/03;G06T7/20;G06K9/44;G06K9/52 主分类号 G06K9/00
代理机构 Finnegan, Henderson, Farabow, Garrett & Dunner LLP 代理人 Finnegan, Henderson, Farabow, Garrett & Dunner LLP
主权项 1. An information processing device comprising: an image inputter configured to input image data in which a user has been captured; an extractor configured to extract a region corresponding to a hand of the user included in the image data input by the image inputter; a determinator configured to determine whether or not the region corresponding to the hand of the user extracted by the extractor is in a stationary state; a recognizer configured to, if the region corresponding to the hand of the user is determined by the determinator as being in a stationary state, recognize the shape of the region corresponding to the hand of the user extracted by the extractor; and an inputter configured to input data associated with the shape recognized by the recognizer, wherein the extractor is further configured to set a rectangle that circumscribes a movement region obtained from difference data between frames in the image data input by the image inputter, specify a region corresponding to a wrist of the user, on the basis of a width of the hand included in the movement region obtained by scanning the inside of the rectangle, and extract the region corresponding to the hand of the user, on the basis of the specified region corresponding to the wrist of the user.
地址 Tokyo JP