A method of extracting 3D patch information from object in image sequences: Volume Editor

A method for the extraction of 3-D patch information from an object in image sequences on a VTR is proposed. The stationary object is filmed on the VTR. The system, in cooperation with a human operator, helps to estimate 3-D patch information of the object accurately and flexibly. The 3-D patch information of the object is expressed as both a rough 3-D shape and fine 2-D texture in the object image. As the operator can 3-D edit the object in image sequences directly, the system is named the Volume Editor.<
>