Advanced Quality Control
in Can Production
Advanced Quality Control
in Can Production
The challenge:
testing beverage cans
An inline inspection system is designed to fully analyze beverage cans and detect and reject faulty decorations and misprints.
The solution:
Multi-camera system
Six or twelve intelligent cameras enable image recognition. They transfer the data to the CanInspect software and to a blow-off device.
Inline inspection system for can production lines – Insensiv GmbH
Depending on the version, the inspection unit generates 360° images of the cans with 6 or 12 cameras and transfers them to the CanInspect software
Cost-efficient, Smart Camera-based Quality Control
Smart cameras are already being used as machine vision solutions for quality assurance and production control in numerous industrial sectors. These powerful units perform all image processing routines independently, making them much more compact and robust than conventional cameras that require an external PC for evaluation. In the following, a high-performance application is presented in more detail in the form of an inline inspection system for beverage cans that can evaluate 50 images per second using twelve intelligent cameras.
insensiv GmbH, which specializes in hardware and software solutions for machine vision tasks, developed the inline inspection system for installation above conveyor belts. Together with the company's proprietary software CanInspect, it enables the detection and analyzation of beverage cans as well as other features (barcode or OCR).
Faulty decorations and misprints are reliably identified: After detecting errors, the software sends a signal to the inspection system, which then rejects the cans with a connected blow-off device.
The inspection system consists of six or twelve board cameras from Vision Components, a customized lighting system, a light barrier, a housing with mounting brackets for installation above all standard conveyor belts, I/Os for the blow-off device and an external PC for further image processing and visualization.
The CanInspect software includes functions for decor recognition, OCR code recognition, barcode recognition, defect image storage, log files, configuration options and statistics evaluation. Operation is via a graphical user interface for Windows PCs.
Decor analysis including OCR function and barcode scan
In the basic version, the inline inspection unit has six cameras arranged in a circle at 60° around the can to be inspected. A light barrier registers cans passing through the unit. All cameras take a picture of the can in the center of the system. The six images are combined to form an overall 360° image, which is then compared with a template database of the current production to identify deviations.
In the CanInspect software, five threshold values can be set for detection sensitivity. If these are exceeded, a corresponding defect is registered. Users can also select in which defect cases the system should initiate a rejection by compressed air blow-off and whether the associated images should be saved.
If an OCR function is also required, the inline inspection unit is equipped with six additional intelligent color cameras for reading codes in text form. These cameras take images of parts of the can at the same time as the decor cameras. Sequentially, the code is searched for in these six individual images. The individual characters in detected codes are recognized using an OCR function and compared with the specified reference code. In the barcode version, six intelligent gray-scale cameras are installed in addition to the decor cameras. These are used to read barcodes and take partial images of the can at the same time as the decor cameras.
The systems are suitable for all standard beverage cans with heights of up to 160 mm and, depending on the version, inspect up to 50 cans per second; the maximum belt speed is 5 m/s.
Key Facts
Technology
Inline inspection with decor analysis, OCR & barcode scanning
High speed
VC Camera
VCSBC board camera with remote head
High performance in most compact design: VCSBC nano-RH
Image acquisition is performed by a global shutter CMOS sensor with a resolution of 752 x 480 pixels, which delivers razor-sharp images even in extremely fast applications.
About insensiv GmbH
The Bielefeld-based insensiv GmbH develops and programs image processing technology for process monitoring and quality control in numerous industries. Intelligent cameras and components or software from insensiv are used, for example, in deposit systems, vending machine technology, printing plants and agriculture.
Advanced Quality Control
in Can Production
Inline quality control for beverage can manufacturers