The invention discloses a plastic
bottleneck excess material detecting method and device based on
machine vision. Detection objects are semitransparent and nontransparent
plastic bottle products formed by hollow
blow molding, and a detection defect type is excess materials in bottlenecks. The method comprises the following steps of: conveying molded plastic bottles to a detection
station by a conveying device, and lighting up light sources through a
position sensor, and triggering cameras to photograph; establishing an imaging
system according to appropriate light sources, cameras and distances between the light sources and the plastic bottles and between the cameras and the plastic bottles, wherein the image effect only highlights details in the bottlenecks and shields the interference of other details; transmitting images to a computer, and determining whether excess materials are in the bottles through
image processing software; and removing bottles with excess materials, and recording the removed bottles. According to the invention, whether excess materials exist in plastic bottlenecks can be detected rapidly and accurately on line; and the plastic
bottleneck excess material detecting method and device based on
machine vision have the characteristics that the imaging scheme can highlight details in bottlenecks, so that the subsequent
image processing method is simple, the computation is less, and moreover, the whole detection
system is high in stability.