The visible foreign matter visual detecting algorithm research in oral liquid

Subtitle:

Article type: Research Article

Authors: Bai, Feng | Zhang, Minglu | Zhang, Xiaojun^* | Huang, Xiaochen

Affiliations: School of Mechanical Engineering, Hebei University of Technology, Tianjin, China

Correspondence: [*] Corresponding author: Xiaojun Zhang, School of Mechanical Engineering, Hebei University of Technology, Tianjin, China. Tel.: +86 13821003733; Fax: +86 022 60204187; E-mail:bfhebut@163.com

Abstract: Aiming at poor precision of the visible foreign body visual inspection in oral liquid, This paper adopt improved image processing algorithm to achieve the accurate detection of foreign body. Using joint difference way to separate the background at the same time improve the energy value of target. Using slope correlation determine whether filling possible fracture and hole in the target. Using self-adaption search algorithm independent choice expansion direction to realize the virtual Foreign body target accurate search. Through the change of form and movement direction difference to eliminate the interference of bubbles, eventually identify the real foreign body target. Experimental results show that the improved foreign body target visual detection steps enhance the detection accuracy of tiny hair and fiber target, the average error rate is less than 2%. It has important guiding significance for foreign body in oral liquid monitoring on the actual production and ensure the safety use of related products.

Keywords: Visual inspection, joint difference, slope correlation, self-adaption search

DOI: 10.3233/JCM-150557

Journal: Journal of Computational Methods in Sciences and Engineering, vol. 15, no. 3, pp. 447-455, 2015