WatchPose: A View-Aware Approach for Camera Pose Data Collection in Industrial Environments


Yang, Cong and Simon, Gilles and See, John Su Yang and Berger, Marie Odile and Wang, Wenyong (2020) WatchPose: A View-Aware Approach for Camera Pose Data Collection in Industrial Environments. Sensors, 20 (11). p. 3045. ISSN 1424-8220

[img] Text
Restricted to Repository staff only

Download (10MB)


Collecting correlated scene images and camera poses is an essential step towards learningabsolute camera pose regression models. While the acquisition of such data in living environmentsis relatively easy by following regular roads and paths, it is still a challenging task in constrictedindustrial environments. This is because industrial objects have varied sizes and inspections areusually carried out with non-constant motions. As a result, regression models are more sensitiveto scene images with respect to viewpoints and distances. Motivated by this, we present a simplebut efficient camera pose data collection method, WatchPose, to improve the generalization androbustness of camera pose regression models. Specifically, WatchPose tracks nested markers andvisualizes viewpoints in an Augmented Reality- (AR) based manner to properly guide users tocollect training data from broader camera-object distances and more diverse views around the objects.Experiments show that WatchPose can effectively improve the accuracy of existing camera poseregression models compared to the traditional data acquisition method. We also introduce a newdataset, Industrial10, to encourage the community to adapt camera pose regression methods for morecomplex environments.

Item Type: Article
Additional Information: This article belongs to the Special Issue Advanced Computational Intelligence for Object Detection, Feature Extraction and Recognition in Smart Sensor Environments
Uncontrolled Keywords: augmented reality, pose estimation, deep learning, industrial environments, data acquisition
Subjects: Q Science > QA Mathematics > QA71-90 Instruments and machines > QA75.5-76.95 Electronic computers. Computer science
Divisions: Faculty of Computing and Informatics (FCI)
Depositing User: Ms Rosnani Abd Wahab
Date Deposited: 13 Dec 2020 12:24
Last Modified: 13 Dec 2020 12:24


Downloads per month over past year

View ItemEdit (login required)