Back to articles
Articles
Volume: 31 | Article ID: art00011
Image
People Recognition and Position Measurement in Workplace by Fisheye Camera
  DOI :  10.2352/ISSN.2470-1173.2019.7.IRIACV-459  Published OnlineJanuary 2019
Abstract

In workplace of factory or office, measuring positions of workers is important to make the workplace visible for improving working efficiency, avoiding miss operation and accidents. Monocular fisheye camera is used by conventional methods to find and track people in workplace. But it is difficult for these methods to get 3D position of workers. Stereo fisheye cameras were used to measure 3D position. But calibrating these fisheye cameras is a hard and time-consuming work. We propose a new method to measure 3D position of people in workplace by only one fisheye camera. One 360-degree fisheye image is projected to a unit sphere and several perspective projection images are generated to correct fisheye distortion. People recognition is made in the corrected images by machine learning method. Recognition results are used to calculate 3D position of people in workplace. Only very few markers are set on floor of workplace to make calibration between perspective projection images and plane of workplace floor. People recognition and position measurement experiments were made in workplace by Ricoh R 360-degree monocular fisheye cameras. People recognition rate of 92.5%, false positive rate of 0.2%, people position measurement accuracy of 6.6% were obtained. The evaluation results demonstrate effectiveness of our proposed method.

Subject Areas :
Views 9
Downloads 0
 articleview.views 9
 articleview.downloads 0
  Cite this article 

Haike Guan, Makoto Shinnishi, "People Recognition and Position Measurement in Workplace by Fisheye Camerain Proc. IS&T Int’l. Symp. on Electronic Imaging: Intelligent Robotics and Industrial Applications using Computer Vision,  2019,  pp 459-1 - 459-7,  https://doi.org/10.2352/ISSN.2470-1173.2019.7.IRIACV-459

 Copy citation
  Copyright statement 
Copyright © Society for Imaging Science and Technology 2019
72010604
Electronic Imaging
2470-1173
Society for Imaging Science and Technology
7003 Kilworth Lane, Springfield, VA 22151 USA