Automated on-line inspection for glass fiber forming

Abstract

Glass fiber forming is a complicated process in which many factors could affect the measuring accuracy of fiber diameters. In the forming machine there are many tubes close to each other, which results in improper lighting and unwanted video signals. This paper presents the employment of a new filter called anti-causal zero-phase was to remove noise without distortion. In this work, the unwanted video signals constantly moved from one place to another, which created a major problem in image analysis. This paper presents a technique to identify the unwanted signals by developing a model for an object, and training the modeled experimental data using a neural network to classify patterns. Only the patterns that met the expectation were used for fiber diameter measurement. The entire inspection process was automated with the aid of a PLC (programmable logic controller). The results for noise removal and pattern classification are included.