FAQs: Inspection features

Perceptor Knowledge Base > FAQs > Inspection features

Perceptor IoT can inspect and verify the following 1D symbologies:

  • Code 128
  • EAN-13
  • EAN-8
  • UPC-A
  • UPC-E
  • ITF (Interleaved 2 of 5)
  • Code 39
  • Code 39 Extended
  • Code 93
  • DataBar
  • DataBar Limited
  • Pharma Code

Perceptor IoT can inspect and verify the following 2D symbologies:
  • Data Matrix Code
  • QR Code
  • Dot Code

Yes, the user can choose to set a minimum pass grade for barcodes of 0.0 to 4.0.

Yes - both options are available in Perceptor IoT i.e barcode reading and barcode verification.

Please check what your licence allows as each option is licensed separately. Barcode verification also requires barcode reading. This process will happen without any input from the operator.

The short answer is yes. But this requires clarification, as follows.

Different light colours can enhance the contrast between label objects and the label itself. The greater the contrast the easier it is for Perceptor IoT - as with any machine vision system - to detect, inspect and verify the areas of interest specified.

The Perceptor IoT inspection camera uses red (630nm) illumination to comply with ISO 15415/15416 code verification. Red light is often used in machine vision as it minimizes ambient reflections and glare. But red light is also readily absorbed by materials including label media and so, depending on the colour of the label/label content, the contrast will vary.

Red label/content: using red light on a red label or with red content means the red label or content will not absorb any light. The light will be reflected and the Perceptor IoT sensor will receive all the light back and the resulting image will be white.

Any black content absorbs the red light so that will appear as black because no light is returned to the sensor.

White content or label media won't absorb the light and so it will reflect and appear as white in the image.

Green label/content: using red light on a green label/green content means the light will be absorbed and the resulting image will be dark, making it impossible to distinguish between the black printed information and the background label media.

Blue label/content: as with the green label, using red light on a blue label/content means the light will be absorbed and the resulting image will be dark, making it impossible to distinguish between the black printed information and the background label media.

Green and blue media are very unresponsive to red light and so will also produce very dark images.

Images produced: 

Perceptor IoT always returns a mono image with different grayscale values according to the colour of the media and the content.

A custom image sensor can be provided if a project or production line requires it - an InfraRed, Green and Blue light solution can be implemented with your Perceptor IoT hardware kit. Please contact us to discuss.

InfraRed light: beyond the visible spectrum, it is used as it is beneficial for applications that require imaging where there would otherwise be very little contrast between media, content and background.

Perceptor IoT detects:

  • miscount
  • incorrect content
  • missing information
  • wrong artwork or version
  • print defects (smudges/blotches, streaks, broken characters, punctuation)
  • misalignment
  • missing, partial, or obscured label content
  • substrate anomalies (foreign substances, bubbles, wrinkles)
  • colour variations; and physical damage

Yes. The system can link human-readable OCR fields to a GS1 barcode to check accuracy and compare content.

The Linked Fields function can be used to ensure data such as GTIN, LOT, EXPIRY, and SERIAL match between human-readable text and machine-readable GS1 DataMatrix or GS1 128 barcode.

Yes. Perceptor IoT is capable of performing real-time duplicate checking.

Each product reference has its own duplication check; if two different product references are used, only duplication within that specific product reference will be detected. Ensure the product reference is configured correctly for accurate duplicate detection.

The duplicate field checker must be enabled for every field in your label vision template that needs to be checked for duplication across any of your labels.

Follow these steps to ensure duplicates are checked for accurately:

1. Enable the duplicate field checker feature for every field that must be unique.

2. Select a product reference.

- A product reference could be your identity number, an internal part number, your customer’s part number etc.

- Whatever product reference is chosen, Perceptor will check for duplicates against that reference.

- Additional product references can be applied so for checking at multiple levels.

3. Save these settings to your label vision template.

With this feature enabled and a product reference applied for every field to be checked, Perceptor will be able to check for, identify and overstrike any duplicates. 

All fonts can be used in your label templates and inspected by Perceptor IoT.

The system reads a minimum font size equivalent to Arial 4pt. Currently supported fonts are Arial, Cambria, Consolas, and OCR-B.

Contact Perceptor if alternative font support is required as these can be added.

When a failure is detected, the selected label voiding mode in settings determines the action. Options include:

  1. Backfeed Overstrike and Continue - the failed label is backfed under the printhead, which overprints VOID across it for the configured distance (up to 6 inches), then printing continues
  2. Backfeed Overstrike and Pause - as above but printing pauses for operator action
  3. Pause - the system does not backfeed but pauses and the operator is prompted on screen to confirm the action they have taken before printing continues.

Rejected labels are displayed on screen, and the live view keeps a configurable history of recent labels. The User can configure in the settings for ALL, FAIL, PASS, NONE inspection images to be saved to the database and included in the Batch Reconciliation Report.

An indicator light flashes red when a printed label fails inspection. Operators with appropriate permissions may perform an Operator Override to force-pass a label.

 

Current inspection speed is approximately 6 inches per second, however speed can be affected by the number of areas of interest being inspected. A simple barcode inspection will not have an impact on speed but if multiple, complex inspections are required, the speed may be compromised.

Perceptor IoT can check for a numerical sequence - e.g. 0001, 00002, 00003 - in an ascending order.
Two consecutive numbers are needed to build the sequence e.g.10001, 10002. If the next number in this sequence is 10004 then it will fail 10004 and also 10005.
However, it will continue with 10006 because that follows a sequence after 10005 and the preceding 10004.
The user also has the facility to put in an incremental step of 1,2,3,4,5 or say 10.

As OCR does not have any error correction - e.g with a DataMatrix code - there is no guarantee that the OCR is 100% correct every time.

If the application is dependent on the OCR being 100% correct then the OCR should be linked to a machine-readable code that does have error correction and so can be relied upon to be 100% correct.

The purpose of the linked field function is to check that two fields are identical so if, for example, a 3 is read as an 8 using the OCR then the difference will be detected in the linked field.

Yes.

The Operator Override function allows an operator to force-pass a failed label. When 21 CFR Compliance Mode is enabled, the operator must select a justification reason (configured by the End User in Settings), enter their email address, and enter their 4-digit PIN. Without a PIN, the operator cannot override and would need assistance from a user with the necessary permission.

An image of the full label is captured and saved to the database along with the Operator email and justification reason is included in the Batch Reconciliation Report as an audit trail.

Browse Solution Categories