Application of video image understanding to measurement of pedestrian flow
Xiaowei Xu · Jisuanji gongcheng yu sheji · 2008
Video image understanding technology,as a hotspot in computer vision,places emphasis on interpreting video sequences,which involves the space property of images and the timing feature of video sequences.Pedestrian flow measurement is an important ap-plication in the field of video image understanding.First,a layered framework for video image understanding is presented,including video image segmentation,object recognition and scene/behavior understanding.Meanwhile,the operational objectives,tasks and related technical fields of each layer are discussed.Then,the current research status of pedestrian flow measurement is summarized.Finally,the problem of pedestrian flow measurement is divided into two problems: Pedestrian counting and crowd density estimation,and solved by using the proposed layered framework.