By supporting a wide variety of products and automating inspection setting operations, we have achieved significant efficiency in the inspection process.
There are a wide variety of labels in terms of shape, size, imposition, material, printing method, etc. Therefore, the inspection settings for print quality inspection are complicated and time-consuming. Moreover, as the needs for printing small quantities and a wide variety of products increase, the frequency of setup changes also increases, and the burden of inspection settings on field workers is increasing.
SiriusVision’s Machine Vision (Iimage Inspection) software “AsmilVision” is an epoch-making software that has succeeded in achieving both simple operation and high-precision inspection.
AsmilVision significantly reduces setup change times by simplifying and automating inspection settings for a wide variety of labels. Moreover, even if the inspection level is raised to detect minute defects, it will not be over-detected. As a result, field inspectors can rest assured that AsmilVision will inspect and ship high-quality products.
Protecting quality in the upstream process of printing
There is an increasing need to introduce automated inspection machines in the printing process to meet stricter quality requirements, but it is said that many of the causes of printing defects are in the printing plate. Therefore, there is an increasing need to use a plate inspection machine for automatic inspection in the printing plate inspection process before entering the printing process. However, printing plates come in a variety of materials and sizes, and it is necessary to select the most suitable plate inspection machine for your printing plate.
SiriusVision has a lineup of plate inspection machines that supports various types of plate materials and printing plates of different sizes. In addition, we have a plate inspection system that strictly checks the quality of plate making; performing a comparative proofreading check of the design data provided by the customer and the plate making data, checking if the printed matter that was test-printed after plate making matches the design data, and checking for defects such as plate scratches and foreign matter.
Our unique technology eliminates the distortion, expansion and contraction of the captured image, which is peculiar to bottle inspection.
We support a wide variety of products by selecting the transport method according to the shape and material of the bottle, the printing method, and the optimum camera and lighting. Containers range from cylindrical to elliptical cylinders, rectangles, polygons, irregular shapes, tubes, and so on. Printing methods also varies from direct printing including screen printing, hot stamping, and inkjet printing to shrink film and labeling. To inspect the print quality of these various containers, take a picture with a line camera while rotating the container, or capture the containers with a multiple area cameras from multiple directions to cover the area required for inspection.
Due to the three-dimensional shape of the container, the captured image will be distorted and stretched regardless of the method. Even when taking an image with a line camera while rotating a cylindrical container, large distortion and expansion and contraction occur in the captured image due to the eccentricity of the axis of rotation and the molding error of the container. In a normal inspection machine, this distortion, expansion and contraction become the difference between the good product master image and the inspected product image, and the difference causes over-detection that the good product is regarded as a defective product.
SiriusVision’s bottle inspection software “FlexVision” has a unique image correction technology that detects only defects with high accuracy without over-detection while correcting such large distortion, expansion and contraction. In particular, FlexVision’s latest technology, “Smart Fit Technology,” can be used to correct even non-linear local distortion.
With conventional Machine Vision (Image Inspection) system, even if high-definition defect detection is possible, over-detection occurs in which a non-defective product is mistaken for a defective product.
People visually judge subtle expansion and contraction, variation, register and misalignment that cause over-detection as good products.
The Machine Vision (Image Inspection) algorithm originally developed by our company is an epoch-making technology that achieves both over-detection suppression equivalent to visual inspection and high-definition automatic inspection.
Marginal Fuzzy Logic
Register Movement Correction
Sample of the Measurement Function
Simultaneous defect detection on both variable and fixed parts
With this unique function, inspection of smudges and print quality of non-variable print (break in a character, blur, scratch, pinhole) and inspection of variable printing with different contents for each sheet such as barcodes, 2D codes and alphanumeric characters on the same surface can be performed at the same time.
For variable printing, it is possible to inspect the print quality at the same time as the readability inspection (reading, recognition, serial number check, etc.).
Variable Inspection (Partial)
Full Variable Inspection (Variable Printing Inspection)
Lighting technology for special substrates
Inspection of foil, holography, transparent labels, etc. was difficult due to problems such as uneven brightness and low contrast. In addition, depending on the lighting conditions, it was possible to support with multiple units or to build a compromised lighting condition. We have solved this problem with multi-lighting technology.
Smart Fit Technology (SFT) to solve problems
Smart Fit Technology is a technology that corrects the registration misalignment between two images by using our our original nonlinear distortion removal algorithm.
With this technology, distortion due to eccentricity and molding tolerances is corrected and only defects are detected.
Smart Fit Technology
Our original non-linear distortion removal processing enables printed matter inspection which has large distortion.
Color measurement function (ΔE measurement)
By using SiriusVision’s color measurement function, it is possible to measure the color information of any part specified by the operator (example of color measurement location: 4 red frames in the lower right figure), and output it as a numerical value.
There are 17 types of color information that can be measured: RGB, SH Gray, CMYK, XYZ, L * a * b *, and ΔE.
Utilization of AI function
With the latest AI technology, we approach the area of inspection that could only be reached by humans with advanced skills and experience.
By fusing the existing machine vision inspection key technology with the latest AI technology, we will achieve the same pass or fail judgment and defect classification automation as skilled inspectors.