Submission: Drost-CVPR10-Edges/TYO-L

Submission name
Submission time (UTC) Aug. 12, 2019, 12:25 p.m.
User Berti
Task Pose estimation (BOP 2019-2022)
Dataset TYO-L
Training model type Default
Training image type None
Description
Evaluation scores
AR:0.613
AR_MSPD:0.600
AR_MSSD:0.586
AR_VSD:0.652
average_time_per_image:36.256

Method: Drost-CVPR10-Edges

User Berti
Publication Drost et al., Model globally, match locally: Efficient and robust 3D object recognition, CVPR 2010.
Implementation
Training image modalities None
Test image modalities RGB-D
Description

An implementation of the paper, including - ICP for post-processing - using 3D surface and 3D edges for voting - using 3D surface, 3D edges and 2D images for refinement

The parameters are set to a slower, but more accurate mode (relative sampling distance of 0.03 for model creation and matching). CPU only, no GPU is used.

Implementation: HALCON 19.05 progress.

Computer specifications Max. 12 Threads on intel Xeon CPU E5-2690 v4 @ 2.60GHz, 2601 Mhz