Performance of Image Enhancement Techniques on the 6D Pose Estimation Method under Complex Lighting Conditions
DOI:
https://doi.org/10.11594/nstp.2025.4755Keywords:
Intelligent robots, 6D pose estimation, complex lighting conditions, image enhancementAbstract
An approach for estimating 6D pose, which identifies location and orientation of the objects, plays an important role for intelligent robots to make appropriate responses when interacting with surrounding objects. In reality, intelligent robots often face complex lighting conditions, where the light intensity on objects is too high or too low. However, there are still few studies on 6D pose estimation under extreme lighting conditions. Therefore, we focus on this issue. Regarding complex lighting problems, the techniques of image enhancement are often added in the preprocessing stage to produce better quality input images. In this study, we apply several image enhancement techniques in the preprocessing stage of a 6D pose estimation method and empirically tests them on the standard dataset to determine which technique has the most influence on the performance of the method to estimating 6D pose. We apply the techniques for image enhancement, namely gamma correction (GC), contrast-limited adaptive histogram equalization (CLAHE), and histogram equalization (HE). We use the EfficientPose method which is state of the art in the study scope. The experimental results exhibit that the addition of image enhancement techniques in the preprocessing improves the performance of the EfficientPose method by up to 6.80% in high lighting condition and 8.30% in low lighting condition, where GC technique performance outperforms other techniques in high lighting condition and the CLAHE technique shows the best performance in low lighting condition. The results in the study provide important information that preprocessing techniques have a significant influence on the performance of the method for estimating 6D pose. Specifically, GC and CLAHE techniques provide the most optimal influence on the method in estimating 6D pose.
Downloads
References
Bukschat, Y., & Vetter, M. (2020). EfficientPose: An efficient, accurate and scalable end-to-end 6D multi object pose estimation approach. ArXiv. https://api.semanticscholar.org/CorpusID:226281757
Di, Y., Manhardt, F., Wang, G., Ji, X., Navab, N., & Tombari, F. (2021). SO-Pose: Exploiting self-occlusion for direct 6D pose estimation. 2021 IEEE/CVF International Conference on Computer Vision (ICCV), 12376–12385.
Finlayson, G. D. (2018). Colour and illumination in computer vision. Interface Focus, 8(4), 20180008.
He, Y., Sun, W., Huang, H., Liu, J., Fan, H., & Sun, J. (2019). PVN3D: A deep point-wise 3D keypoints voting network for 6DoF pose estimation. 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 11629–11638.
Hinterstoisser, S., Holzer, S., Cagniart, C., Ilic, S., Konolige, K., Navab, N., & Lepetit, V. (2011). Multimodal templates for real-time detec-tion of texture-less objects in heavily cluttered scenes (pp. 858–865). Proceedings of the 2011 International Conference on Computer Vision, ICCV 2011. IEEE Computer Society.
Hinterstoisser, S., Lepetit, V., Ilic, S., Holzer, S., Konolige, K., Bradski, G. R., & Navab, N. (2013). Model Based Training, Detection and Pose Estimation of Texture-Less 3D Objects in Heavily Cluttered Scenes. Asian Conference on Computer Vision (ACCV). https://api.semanticscholar.org/CorpusID:17334031
Hinterstoisser, S., Lepetit, V., Rajkumar, N., & Konolige, K. (2016). Going further with point pair features. European Conference on Com-puter Vision. https://api.semanticscholar.org/CorpusID:20877803
Kehl, W., Manhardt, F., Tombari, F., Ilic, S., & Navab, N. (2017). SSD-6D: Making RGB-Based 3D Detection and 6D Pose Estimation Great Again. 2017 IEEE International Conference on Computer Vision (ICCV), 1530–1538. https://doi.org/10.1109/ICCV.2017.169
Kim, J., Pyo, H., Jang, I., Kang, J., Ju, B., & Ko, K. (2022). Tomato harvesting robotic system based on Deep-ToMaToS: Deep learning net-work using transformation loss for 6D pose estimation of maturity classified tomatoes with side-stem. Computers and Electronics in Agriculture, 201, 107300. https://doi.org/https://doi.org/10.1016/j.compag.2022.107300
Michel, F., Kirillov, A., Brachmann, E., Krull, A., Gumhold, S., Savchynskyy, B., & Rother, C. (2016). Global hypothesis generation for 6D object pose estimation. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 115–124.
Pereira, N., & Alexandre, L. A. (2019). MaskedFusion: Mask-based 6D Object Pose Estimation. 2020 19th IEEE International Conference on Machine Learning and Applications (ICMLA), 71–78.
Tremblay, J., To, T., Sundaralingam, B., Xiang, Y., Fox, D., & Birchfield, S. (2018). Deep object pose estimation for semantic robotic grasp-ing of household objects. ArXiv, abs/1809.10790. https://api.semanticscholar.org/CorpusID:52893770
Wu, D., Zhuang, Z., Xiang, C., Zou, W., & Li, X. (2019). 6D-VNet: End-to-end 6DoF vehicle pose estimation from monocular RGB images. Proceedings of IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW).
Wu Yangzheng, Zand, M., Etemad Ali, & Greenspan Michael. (2022). Vote from the center: 6 DoF pose estimation in RGB-D images by radial keypoint voting. In G. and C. M. and F. G. M. and H. T. Avidan Shai and Brostow (Ed.), Computer Vision – ECCV 2022 (pp. 335–352). Springer Nature Switzerland.
Xiang, Y., Schmidt, T., Narayanan, V., & Fox, D. (2018). PoseCNN: A convolutional neural network for 6D Object pose estimation in clut-tered scenes. Robotics: Science and Systems. https://doi.org/10.15607/RSS.2018.XIV.019
Xu, Y., Lin, K.-Y., Zhang, G., Wang, X., & Li, H. (2022). RNNPose: Recurrent 6-DoF object pose refinement with robust correspondence field estimation and pose optimization. 14860–14870. https://doi.org/10.1109/CVPR52688.2022.01446
Downloads
Published
Conference Proceedings Volume
Section
License
This work is licensed under a Creative Commons Attribution 4.0 International License.
Authors who publish with this proceedings agree to the following terms:
Authors retain copyright and grant the Nusantara Science and Technology Proceedings right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this proceeding.
Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the proceedings published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this proceeding.
Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See the Effect of Open Access).
