Deep Learning - Based Shape Recognition and Classifications of Conic Geometries in Engineering Drawing
Abstrak
Engineering drawings frequently contain conic geometries such as circles, ellipses, parabolas, and hyperbolas, which are fundamental to mechanical design and industrial applications. Accurate identification and classification of these shapes are therefore essential for computer-aided design (CAD) systems, automated inspection, and intelligent design analysis. However, conventional geometry-based or rule-based approaches often perform poorly when drawings are noisy, complex, or partially incomplete. This study proposes a deep learning-based approach using convolutional neural networks (CNNs) to automatically extract features and classify conic shapes in engineering drawings. By learning discriminative visual representations directly from input data, the proposed method enhances classification accuracy, improves robustness, and reduces the need for manual intervention. The study concludes that CNN-based conic shape recognition offers a reliable and efficient solution for engineering and industrial contexts, with practical implications for improving automation and intelligent analysis in design-related applications.
Penulis (5)
Rajnandani Das
Neha Shah
Dilip Kumar Sah
K. Sahani
S. K. Sahani
Akses Cepat
- Tahun Terbit
- 2026
- Bahasa
- en
- Sumber Database
- Semantic Scholar
- DOI
- 10.58578/ajstea.v4i2.9335
- Akses
- Open Access ✓