Towards a Smarter CAD: Application of the Integration of Artificial Vision and Neural Networks in Aided Design

In this study, a tool based on neural networks was developed for the processing and recognition of primitive entities in images, with a specific focus on the field of Computer Aided Design (CAD). The tool demonstrated effectiveness in identifying and categorizing entities, optimizing detection and editing processes in CAD systems. The most immediate and realistic application of this tool lies in the simplification of technical drawings, in designs and the facilitation of the con-version of images to vectors. This innovation has the potential to significantly improve efficiency and accuracy in the CAD workflow.