3-Layer specific pedestrian continuous tracking framework
Fang Baofu, Wang Zaijun, Jinjin Chen, Hao Wang, Fang Lu · 2017
Pedestrian tracking is an important research direction and hot spot in the field of machine vision. In this paper a 3-layer specific pedestrian continuous tracking framework based on secondary feature extraction is proposed. Firstly, we give the image features set used in secondary feature extraction, as well as feature set difference measurement method and online feature weight Adjustment method. Secondly, a 3-layer continuous frame framework is described in detail: detection layer, tracking layer and management layer, and the pedestrian tracking tasks can be executed quickly and efficiently. Experiments show that this framework can complete the pedestrian tracking task in realtime and robustly.