Curve Matching Algorithm Applied to Fabric Automatic Register Detection

Abstract

With the register detection problems in the rotary screen printing process, the curve matching strategy was proposed to get the position vector of off-pattern-topping, on the basis of the proposed rotary screen printing machine automatic register detection system based on machine vision. In order to ensure the capture positions of the target edge image and the standard edge image is the same, Fourier-Mellin curve matching was used to match them to eliminate the shooting error;then the curve matching method, based on the Fourier transform shift theorem,was used to match the well-tuned target edge image and the standard edge image to detect the register error. The matching strategy eliminated the error, and the computation is small.The experiments show that its detection accuracy and the computing speed can meet the requirements of the fabric register detection on line.