Track: Artificial Intelligence

Abstract

Texture analysis and evaluation would help in predicting a component’s functionality. An attempt has been made in the current research work to explore the texture properties of machined surfaces in detail. Pieces made from milling, shaping, grinding, EDM, and sandblasting processes are used for texture analysis. The method uses a charge-coupled device (CCD) camera connected to the vision system to acquire images for texture analysis.  The experimental investigation revealed that texture features extracted from the grey level co-occurrence matrix (GLCM) help study texture features. The experimental investigation revealed that when distance =1, the classification accuracy was less. However, the classification accuracy improved when distance =2, but a further increase in the distance would not significantly improve the classification accuracy. Instead, it would add up to the computational time. Thus, it was concluded that d=2 gave maximum classification accuracy of 96%.