Curvelet transform for estimation of machining performance

Automation of the processes from design to final component manufacturing poses numerous challenges. The problems that need to be addressed are tolerances for geometry, dimensional accuracy and surface quality of the component. The first two can be addressed by integrating the Computer aided design, computer aided manufacturing and machining systems. The real time monitoring and performance estimation systems for surface quality need to be developed. As such approach in this work, front end software is developed where the input to software will be the surface image of the machined workpiece taken by a camera. The image is further processed in image processing module in mat lab for the texture feature extraction namely the curvelet transforms data. The software in turn classifies the machining processes undergone according to the trend in the curvelet transform data and then sends it to the corresponding process model for the estimation of the machining performance. In this work, the software is tested with sample test data and it displayed the machining process accurately and estimated the machining performance. The results are encouraging and this method can be used in real time estimation of machining process and machining performance.