Shape Recovery of a Levitated Aspherical Droplet From 2D Image Information

Abstract

This paper presents an image processing technique in order to predict the shape of a levitated aspherical droplet. The technique is of great importance to containerless materials processing. A majority of the electromagnetic levitation techniques utilizes two cameras at right angles to observe both transversal and frontal views. This allows obtaining two images of the droplet at instant time. In many cases, the portion of the frontal image is missing due to the heating coil. The newly developed technique allows restoration of the missing portion of the image information. The through image can be reconstructed by combining the recovered shapes. A special computer program is generated to simulate a normalized volume of the droplet.