METHOD AND A DEVICE FOR DIVIDING A SIGNIFICANT OBJECT WHICH HAS A MEANING FROM SOUND BY USING SOUND LOCATION INFORMATION

摘要

PURPOSE: A method and a device for diving a significant object from sound are provided to divide objects which have a meaning from multi-channel audio signal by using virtual sound location information.;CONSTITUTION: A virtual acoustic image location information and an audio signal are extracted from a bit stream(210). Based on virtual sound location, an object included in an audio signal is divided(220). Objects of previous frame and objects of current frame which is existing in a virtual sound location are mapped(240). The mapped objects are extracted between continued frames(250). The virtual sound location information is extracted based on the multi channel signal size or is extracted from additional information of a beat stream.;COPYRIGHT KIPO 2011