Abstract: Encasement counting is an important part of edible oil production line.How to use the advanced technology in the high-speed assembly line to improve the efficiency,reduce cost,reduce the error detection rate and miss detection rate is the key.A complete set of intelligent encasement counting system based on machine vision is designed,including the hardware parts,such as industrial camera light source,industrial personal computer and so on,and the software parts,such as mechanical control,image processing and so on.A series of algorithms such as image preprocessing denoising,bottle cap positioning and lifting ring positioning are proposed to deal with the quantity of oil bottles and the defects of the lifting ring.Experimental results show that the proposed method can effectively improve the detection rate and efficiency,and meet the requirements of practical applications.

Key words: machine vision, image processing, automatic counting system