Automatic Recognition of Internal Features of Axisymmetric Parts from 2-D Images

Objectives: This paper presents the development of a software package for Automatic Extraction of Geometrical Data (AEGD) and Recognition of Internal Turning Features (RITF). Methods/Analysis: The developed algorithms used to extract geometrical data and recognize internal features from the 2-D gray scale images and analyzed the recognized features. The expert system is developed using Java Advanced Image packages. The extracted test examples geometrical data analyzed with known geometrical data of a CAD model of the same axisymmetric component. Findings: The present research work deals with the development of a feature-based approach to automatically extract geometrical data and recognize the internal features of the workpiece from an image database without any human intervention. The geometrical data extracted from the sample images and the extracted data used to recognize turned features with the developed algorithms. It is found that the average percentage variation of extracting geometrical data not more than 0.09. Applications/Improvement: The proposed system is novel approach to extract geometrical data from the image files and use this data to recognize turned features. This information is required for downstream applications i.e. generation of process plans, production schedules, NC code generation etc.