Nowadays the quality inspect of soft capsules is mainly by manual. Despite the intensive of this work, the accuracy of inspection by manual is very low. This paper proposed soft capsules online sorting system based on machine vision. The inspection process are following: (1) soft capsules were placed on rollers are rotating while moving. The image of each soft capsule was grabbed. (2) automatic threshold based on ostu was used to segmentation capsule image from background, and morphological filter was used to eliminate noise and regional markings. (3) 4 features were extracted which were perimeter, area, girth, altitude diameter and latitude diameter. Support Vector Machine (SVM) and was used to analyze these features. 15460 soft capsules were tested by the online sorting system. The overall grading accuracy was up to 94.1%. Furthermore, the grading speed of the sorting line resches10 capsules per second. © 2009 Springer Science+Business Media, LLC.
CITATION STYLE
Ge, F., Shi, J., Xu, Y., Zou, X., & Zhao, J. (2009). Machine vision on-line detection quality of soft capsules base on SVM. In IFIP International Federation for Information Processing (Vol. 294, pp. 1369–1378). Springer Science and Business Media, LLC. https://doi.org/10.1007/978-1-4419-0211-5_65
Mendeley helps you to discover research relevant for your work.