Automatic Measurement System of Section Geometry Parameters for Composite Filter Rods Based on Machine Vision

Abstract

This article aims to develop a reliable automatic measurement system to measure the geometry parameters of the cross section of cigarette filter rods. To achieve this purpose, an automatic measurement system based on machine vision is constructed. This measurement procedure is divided into three stages: initial stage, cutting stage, and measurement stage. In the measurement stage, a robust and accurate measurement approach is proposed to measure the depth, number of grooves, and area ratio of grooves to the section for different kinds of cigarette filter rods. First, a dynamic location algorithm is proposed to solve the position error of the mechanical system based on template matching. Subsequently, considering the uncertainty of the shape of the grooves, the iterative groove identification based on convex hull (IGICH) algorithm is constructed to detect correctly the grooves using three features: the area, width, and distance of groove distribution for the connected regions. Next, we have presented the section geometry parameter measurement method according to quality control requirements. Moreover, the system calibration algorithm is developed to improve the system accuracy based on the telecentric lens principle. Finally, based on the experimental setup, detailed characterizations including efficiency, precision, and robust analysis were performed. The results show that the proposed system can adapt to the automatic, precise, and robust measurement of geometry parameters for the section of different kinds of cigarette filter rods, and satisfy the measurement accuracy and speed required by quality control.