SOTAVerified

Burr detection and classification using RUSTICO and image processing

2021-11-17Journal of Computational Science 2021Code Available0· sign in to hype

Virginia Riego, Lidia Sánchez-González, Laura Fernández-Robles, Alexis Gutiérrez-Fernández, Nicola Strisciuglio

Code Available — Be the first to reproduce this paper.

Reproduce

Code

Abstract

Machined workpieces must satisfy quality standards such as avoid the presence of burrs in edge finishing to reduce production costs and time. In this work we consider three types of burr that are determined by the distribution of the edge shape on a microscopic scale: knife-type (without imperfections), saw-type (presence of small splinters that could be accepted) and burr-breakage (substantial deformation that produces unusable workpieces). The proposed method includes RUSTICO to classify automatically the edge of each piece according to its burr type. Experimental results validate its effectiveness, yielding a 91.2% F1-Score and identifying completely the burr-breakage type.

Tasks

Reproductions