3D Orientation Estimation of Industrial Parts from 2D Images using Neural Networks

Abstract : In this paper we propose a pose regression method employing a convolutional neural network (CNN) fed with single 2D images to estimate the 3D orientation of a specific industrial part. The network training dataset is generated by rendering pose-views from a textured CAD model to compensate for the lack of real images and their associated position label. Using several lighting conditions and material reflectances increases the robustness of the prediction and allows to anticipate challenging industrial situations. We show that using a geodesic loss function, the network is able to estimate a rendered view pose with a 5 degrees accuracy while inferring from real images gives visually convincing results suitable for any pose refinement processes.
Liste complète des métadonnées

https://hal.archives-ouvertes.fr/hal-01681124
Contributor : Harold Mouchère <>
Submitted on : Thursday, January 11, 2018 - 1:21:55 PM
Last modification on : Tuesday, March 26, 2019 - 9:25:22 AM

Identifiers

  • HAL Id : hal-01681124, version 1

Collections

Citation

Julien Langlois, Harold Mouchère, Nicolas Normand, Christian Viard-Gaudin. 3D Orientation Estimation of Industrial Parts from 2D Images using Neural Networks. International Conference on Pattern Recognition Applications and Methods, Jan 2018, Madeira, Portugal. ⟨hal-01681124⟩

Share

Metrics

Record views

280