3D SKELETON ESTIMATION AND MORPHING OF DOGS FROM A SET OF IMAGES
Volodymyr Husiev- Yaroslav Tereshchenko
Опубліковано 15.08.2025
Як цитувати
Завантаження
Авторське право (c) 2025 Volodymyr Husiev; Yaroslav Tereshchenko (Науковий керівник)
Ця робота ліцензується відповідно до Creative Commons Attribution-ShareAlike 4.0 International License.
Анотація
We propose a unified pipeline for estimating and morphing 3D dog meshes from a set of images. Our pipeline consists of four main stages: dataset preparation and standardization of 3D dog meshes with texture mapping and multi-view screenshot generation, 2D keypoint detection using trained YOLOv8 models followed by 3D reconstruction via Perspective-n-Point algorithms, statistical shape modeling through PCA-based morphological analysis , and iterative mesh morphing with skeleton-guided optimization.
Посилання
- 1. Yaseen М. (2024) What is YOLOv8: An In-Depth Exploration of the Internal Features of the Next-Generation Object Detector. Retrieved from: https://arxiv.org/abs/2408.15857.
- 2. Lepetit V., Moreno-Noguer F. & Fua P. (2009) EPnP: An Accurate O(n) Solution to the PnP Problem. International Journal of Computer Vision, (81), 155–166. Retrieved from: https://link.springer.com/article/10.1007/s11263-008-0152-6.
- 3. Blanz V. & Vetter T. (1999) A morphable model for the synthesis of 3D faces. Proceedings of the 26th Annual Conference on Computer Graphics and Interactive Techniques (SIGGRAPH '99), 187–194. Retrieved from: https://dl.acm.org/doi/10.1145/311535.311556.