Torso detection in static images is a crucial and challenging problem. Torso means the trunk portion of the human body. Locating the torso region of a human body is complex due to its cluttered backdrop and various postures. Existing works on torso detection typically insist a prolonged training phase for complex shape-matching processes. In this work, a new methodology for detecting the torso portion of the human body is introduced. In this approach, a face detection algorithm is used to identify the face region as a priori and then the trunk region is extracted by effectively combined the input image segments and the foreground detection algorithm along with the detected face.