Deep Learning can be used to reliably read expiration dates on yogurt containers. AI enables reliable inspection.
Reading the Best Before Date on Yogurt Cups
Our Solutions
- AI-based OCR tool for reading text and symbols on difficult surfaces, especially for direct-marked content.
- OCR and OCV applications on labels, packaging, paper
- Code reader for inline manufacturing
- Code reader for lasered, needled or
printed codes - Validation system for inline code evaluation
- Code verification according to AIM/DPM or ISO 15415
- Print image control and comparison against ZPL files
-
Reading on an Inhomogeneous Background
Segmentation has always been the biggest challenge with previous OCR systems. Fluctuating brightness or variable backgrounds could quickly lead to false reads. AI systems, on the other hand, are able to segment and classify characters even on inhomogeneous backgrounds. This significantly increases reliability and robustness in many industries.
-
Weak Contrast
Similar to inhomogeneous backgrounds, weak contrast can also quickly bring classic font reading systems to their knees. Here, too, Deep Learning shows its strengths. An AI system can even read raised lettering on rough surfaces such as cast aluminum robustly and reliably.
-
Varying Characters
AI-based reading tools are adaptable and can easily handle different fonts, sizes, and layouts without the need for complex manual adjustments. Especially in the case of stamped or embossed fonts and inkjet printing (e.g. sell-by dates on food packaging), the character representation can vary considerably while the text remains legible to the human eye. This is where the strengths of AI come into play.
-
Learning New Fonts or Font Sizes
Thanks to its training, an AI system has a diverse collection of fonts that allows it to effortlessly handle a wide range of characters, regardless of their size, style, or stroke width, without prior training. If necessary, additional characters can be integrated into the training to further increase the robustness of the system.
Detect Defects in a Printed Image at the Highest Level
We detect reliably:
- Print offset
- Print position of individual print elements
- Incorrectly printed characters
- Incorrectly printed characters
- Blurred characters
- Graphic symbols for completeness
- Frames and lines for completeness and possible interruption
- Barcodes: Completeness, broken lines, contrast, conformity to standards
- Datamatrix codes: AIM-DPM, ISO 15415
- Background for contamination
- and many others.
Fully automated ZPL-based print inspection - is it possible?
Yes, thanks to senswork ZPL Inspector!
senswork's ZPL-Inspector enables direct interpretation of ZPL (Zebra Printer Language) print jobs and automatic generation of inspection programs. This means that our ZPL-Inspector camera solution automatically adapts to the print layout without having to change the inspection program. The contents of the ZPL printer file are read and interpreted, and an appropriate inspection program is generated - fully automatically and without user intervention.
Reading 1D and 2D Barcodes
Read and evaluate the quality of directly marked and printed codes
For the identification and traceability of products and components, codes are applied to them using various methods, such as laser marking, needling or printing.
To ensure the quality and readability of barcodes, there are several standards such as AIM/DPM or ISO/IEC. These define partly different quality characteristics for the evaluation of codes. AIM/DPM is aimed at directly marked barcodes, while ISO/IEC 15415 deals with printed 2D codes.
Benefits - that's why with senswork
The planning of OCR and OCV systems requires extreme care. In addition to the correct camera resolution, suitable illumination of the reading area is crucial. In particular, the choice of illumination wavelength, whether visible light or infrared, can have a significant impact on background suppression. Different printing methods also require specific reading instruments. While conventional methods may be sufficient in some cases, we increasingly recommend the use of our AI-based tools for reliable inspection results.
You can rely on us! As a competent partner, we will make your OCR or OCV project a success.
OCR System for Turbocharger Guide Vanes
A dual camera system with mirror deflection enables OCR verification of part numbers on turbocharger guide vanes.
Inline Label Inspection with Smart Camera
The smart camera from senswork enables reliable inspection of texts, codes and symbols - even at a great distance.