Implementation of 3D Shapes by Using an Optimal and Efficient Shape Recovery Algorithm

Abstract

This research work presents an optimal and efficient implementation of 3D shape recovery of objects based on image focus analysis. A modified SML operator known as Cross Sum Modified Laplacian (XSML) operator is employed which contains diagonal component and improves the quality performance as compared to conventional SML operator. The Gaussian interpolation technique is used to model the focus measure function. In order to analyze, accomplish and validate the proposed algorithm several synthetically refocused real objects with varying textures are used. An intensive analysis in carried out based on quantitative and qualitative evaluation by using conventional methods and statistical metrics. Both evaluated quantitative and qualitative frameworks ensure and validate the optimality and efficiency of suggested algorithm.