International Journal of Video&Image Processing and Network Security IJVIPNS-IJENS Vol:13 No:01 7
2015
Abstract — Through this paper we mention a new cogent method for 3D fitting ellipsoid and conic from scattered data, damaged by noise and outliers. This 3D statistical approach fitting builds on 3D Zernike moments. We prefer to estimate the probability density function p.d.f with the Zernike moment theory overseen by maximum entropy principle MEP. The purposed algorithm was successfully used in simulated noisy 3D conic forms images and 3D fitting a tumor for magnetic resonance imaging MRI. A deeper analysis of 3D fitting process is demonstrated to show its rapidness as well as its superior performance related noise immunity.