Avtor/Urednik		Kodek, Timotej; Pernuš, Franjo
Naslov		Zaznavanje razpok v injekcijskih stekleničkah z računalniškim vidom
Prevedeni naslov		A computer vision system for detecting cracks in vials
Tip		članek
Vir		In: Zajc B, editor. Zbornik 9. elektrotehniške in računalniške konference ERK 2000. Zvezek B. Računalništvo in informatika, umetna inteligenca, robotika, razpoznavanje vzorcev, biomedicinska tehnika, močnostna elektrotehnika, didaktika, študentski članki; 2000 sep 21-23; Portorož. Ljubljana: Slovenska sekcija IEEE,
Leto izdaje		2000
Obseg		str. 257-60
Jezik		slo
Abstrakt		In this paper an automatic system for inspecting the quality of vials (little glass bottles) filled with antibiotics in powder is presented. Such vials are used in the pharmaceutical industry where every single vial has to be inspected for quality standards. At present this task is still performed manually, which is a tedious and eye-straining process. A defective bottle has some sort of an anomaly (usually little cracks) inside the glass that must be reliably detected from the acquired images. Images were taken throughout the whole vial revolution and afterwards processed in order to filter out only the useful information. Algorithms of confining the field of interest, finding minimim-value points, closing and image subtraction were used for this purpose.
Deskriptorji		IMAGE PROCESSING, COMPUTER-ASSISTED PATTERN RECOGNITION DRUG PACKAGING ANTIBIOTICS DRUG INDUSTRY GLASS