Learning and Classifying Motions of Construction Workers and Equipment Using Bag of Video Feature Words and Bayesian Learning Methods
Jie Gong, Carlos H. Caldas · 2011
Automated motion classification of construction workers/equipment from videos is a challenging problem, but has a wide range of potential applications in construction. These applications include, but are not limited to, enabling rapid construction operation analysis and ergonomic studies. This research explores the potential of an emerging motion analysis framework, bag of video feature words, in learning and classifying workers and heavy equipment motions in challenging construction environments. We developed a test bed that integrates the bag of video feature words with a Bayesian learning method, and evaluated the performance of this motion analysis approach on two video data sets. For each video data set, a number of motion models are learned from the training video segments and applied to the testing video segments. Compared to previous studies of construction worker/equipment motion classification, this new approach can achieve good performance in learning and classifying multiple motion categories while robustly coping with the issues of partial occlusion, view point and scale changes.