2518-1092Research result. Information technologies2518-109210.18413/2518-1092-2024-9-3-0-63560ARTIFICIAL INTELLIGENCE AND DECISION MAKING<strong>APPROACH TO VECTORIZATION OF DRAWINGS OF&nbsp;DESIGN DOCUMENTATION ON PAPER</strong><strong>APPROACH TO VECTORIZATION OF DRAWINGS OF&nbsp;DESIGN DOCUMENTATION ON PAPER</strong>BasovOleg OlegovichBasovOleg Olegovicho.basov@acti.ruDeminOleg DmitrievichDeminOleg DmitrievichNoskovDmitry AleksandrovichNoskovDmitry Aleksandrovich20249300

The work proposes a solution to the problem of vectorization and machine interpretation of drawings of design documentation on paper, which provides the ability to automate the transfer of images of products, parts and assembly units into CAD systems. A number of neural network architectures have been proposed for detecting and recognizing the main elements of a drawing (frame, title block, specification, views, projections and sections), inscriptions, dimension and extension lines, as well as primitives that directly describe the product. For hierarchical and interconnected vectorization, a mechanism for semantic segmentation of drawings based on a graph neural network is proposed. The results of the implementation of the main stages of solving the problem of vectorization of design drawings are presented.

The work proposes a solution to the problem of vectorization and machine interpretation of drawings of design documentation on paper, which provides the ability to automate the transfer of images of products, parts and assembly units into CAD systems. A number of neural network architectures have been proposed for detecting and recognizing the main elements of a drawing (frame, title block, specification, views, projections and sections), inscriptions, dimension and extension lines, as well as primitives that directly describe the product. For hierarchical and interconnected vectorization, a mechanism for semantic segmentation of drawings based on a graph neural network is proposed. The results of the implementation of the main stages of solving the problem of vectorization of design drawings are presented.

design documentationdrawingproductvectorizationgraph neural networkdeep reinforcement learningdesign documentationdrawingproductvectorizationgraph neural networkdeep reinforcement learningСписок литературыLiu C., Wu J., Kohli P., Furukawa Y. Raster-to-vector: revisiting floorplan transformation. Proceedings of the IEEE International Conference on Computer Vision. 2017: 2195&ndash;2203. DOI: 10.1109/ICCV.2017.241.Ellis K., Ritchie D., Solar-Lezama A., Tenenbaum J. Learning to infer graphics programs from hand-drawn images. Advances in neural information processing systems. 2018: 6059&ndash;6068.Guo Y., Zhang Z., Han C., Hu W.B., Li C., Wong T.T. Deep line drawing vectorization via line subdivision and topology reconstruction. Comput. Graph. Forum 38, 2019: 81&ndash;90. DOI: 10.1111/cgf.13818.Gao J., Tang C., Ganapathi-Subramanian V., Huang J., Su H., Guibas L.J. Deepspline: Data-driven reconstruction of parametric curves and surfaces. 2019. arXiv preprint arXiv:1901.03781.Ha D., Eck D. A neural representation of sketch drawings. 2018. URL: https://openreview.net/pdf?id=Hy6GHpkCW.Zhou T., Fang C., Wang Z., Yang J., Kim B., Chen Z., Brandt J., Terzopoulos D. Learning to doodle with stroke demonstrations and deep q-networks. BMVC. 2018: 13.Kaiyrbekov K., Sezgin M. Stroke-based sketched symbol reconstruction and segmentation. 2019. arXiv preprint arXiv:1901.03427.Zheng N., Jiang Y., Huang D. Strokenet: A neural painting environment. ICLR (Poster) 2019.Ribeiro L.S.F., Bui T., Collomosse J., Ponti M., Sketchformer: Transformer-based representation for sketched structure. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 2020: 14&thinsp;153&ndash;14&thinsp;162.Lopes R.G., Ha D., Eck D., Shlens J. A learned representation for scalable vector graphics. Proceedings of the IEEE/CVF International Conference on Computer Vision (CVPR). 2019: 7930&ndash;7939.Carlier A., Danelljan M., Alahi A., Timofte R. Deepsvg: A hierarchical generative network for vector graphics animation. Advances in Neural Information Processing Systems. 2020, vol. 33: 16&thinsp;351&ndash;16&thinsp;361.Azadi S., Fisher M., Kim V.G., Wang Z., Shechtman E., Darrell T. Multi-content gan for few-shot font style transfer. Proceedings of the IEEE conference on computer vision and pattern recognition (CVPR). 2018: 7564&ndash;7573.Gao Y., Guo Y., Lian Z., Tang Y., Xiao J. Artistic glyph image synthesis via one-stage few-shot learning. ACM Transactions on Graphics (TOG). 2019, vol. 38, no. 6: 1&ndash;12.Li T.-M., Luk&aacute;č M., Gharbi M., Ragan-Kelley J. Differentiable vector graphics rasterization for editing and learning. ACM Transactions on Graphics (TOG). 2020, vol. 39, no. 6: pp. 1&ndash;15.Hao S., Xuefeng L., Jianwei N., Jiahe C., Ji W., Xinghao W., Nana W. MARVEL: Raster Gray-level Manga Vectorization via Primitive-wise Deep Reinforcement Learning. IEEE Transactions on Circuits and Systems for Video Technology. 2023: 2677&ndash;2693. DOI: 10.1109/TCSVT.2023.3309786.