AUTOMATIC LASER ENGRAVING READER – OCR

CTC


AUTOMATIC LASER ENGRAVING READER – OCR

ISSUE

Among its objectives, the Centre Technique du Cuir (CTC) wishes to ensure the traceability of calf hides. To this end, it undertakes R&D actions to develop laser marking and optical character recognition systems. In this context, the CTC relied on Neovision and its expertise in artificial intelligence.

CUSTOMER

CTC Lyon (FRANCE)

SOLUTION PROVIDED

After assisting the CTC in the acquisition of usable data, Neovision designed and developed an OCR technology capable of reading laser engravings on calf skins. This robust technology remains capable of reading characters despite significant deformation.

CUSTOMER BENEFIT

The benefits induced by this technology are twofold for the CTC. The CTC gets a custom-made and complete solution. The identification by operators of each skin, at first tedious, repetitive and above all too slow for production constraints, is now automated. This results in a significant improvement in traceability. Finally, the CTC plans to market this technology on a large scale (international target) which will enable it to establish and extend its leadership and expertise.

REALIZATION

The CTC – Centre Technique du Cuir – provides public service missions for the French leather industries. CTC’s service offer addresses a wide range of customers (manufacturers, distributors and brands) and includes all the services required for design, manufacturing, sourcing and distribution.

Neovision and the CTC began their collaboration by studying the data the CTC had. These data correspond to images of laser engravings on leather. Following this audit, Neovision supported the CTC in the acquisition of new data, more suitable for use by an AI. The characters to be read are strongly distorted by the industrial process of leather creation, leaving traditional OCR (automatic character reading) in failure.

Neovision developed an OCR (Optical Character Recognition) technology capable of reading these engravings. To address this issue, the unique solution developed by Neovision combines several Deep Learning algorithms capable of detecting and recognizing characters despite strong distortions.

Neovision then developed a complete software solution for the CTC, which was put into production in real conditions to validate its performance in the Annonay and Haas tanneries. The solution is now being optimized and the prospects for its use in other French tanneries and abroad are promising.

Exemple OCR sur gravures laser
Exemple OCR sur gravures laser
Exemple OCR sur gravures laser
Exemple OCR sur gravures laser

ASSOCIATED TECHNOLOGIES AND EXPERTISE

COMPUTER VISION, DEEP LEARNING, R&D

TESTIMONIAL

Cédric Vigier - Responsable Département Innovation

« Neovision was able to work efficiently to meet our initial requests and to adapt to the last-minute constraints related to the R&D projects we entrusted to them, which by their very nature evolve and adapt as the study progresses. This adaptability and agility allowed us to get convincing results for our project.
The team is dynamic, innovative, organized, punctual and attentive to its customers in order to bring solutions adapted to our industrial stakes and constraints.
Besides, Neovision showed they work with quality in mind to produce results in line with their customers’ needs. »                    

  Cédric Vigier, Head of Innovation Department CTC

Date

16 October 2020

Category

Computer Vision, Deep Learning, R&D

Neovision © 2020