AI, the new focus of image processing

In the automotive industry and also in areas where small parts are pressed, the demands on the components are becoming ever greater.

Dimensional accuracy and flawless surfaces are a minimum requirement.

With new optics and cameras, it’s no longer just a matter of getting the lighting right, AI is also being used more and more to give customers the results they want.

The company evopro systems engineering AG has developed its own smart camera, the evoVIU. In the course of this, a department for image processing and AI was also founded, which now comprises 8 employees and has already been able to generate orders from well-known automobile manufacturers.

Why image processing is so important?

Image processing can help your company increase throughput by optimizing quality in the form of reducing costly rework. Likewise, image processing increases efficiency by reducing costs.

Our portfolio includes the following methods and solutions:

Development of image processing systems for the following tasks.

  • Code Recognition:
    Identification of an object by drawings (barcode, DMC, QR code, customer specific code)

  • Object Recognition:
    Identification of an object by characteristic features such as shape/geometry, dimension, color, structure/topology, texture

  • Positional Recognition:
    Determination of the pose (position + orientation) of an object in a reference system by calculating features or matching method

  • Completeness check:
    Categorization of workpieces as correctly or incorrectly assembled

  • Form and dimensional inspection:
    Determination of geometric quantities with emphasis on precise and accurate measurements

  • Surface Inspection:
    Determination of topographical features (quantitative inspection) or surface defects (qualitative inspection)

In the future, we plan to:

Implementing AI methods in areas where inspection features are not definable in a traditional way, such as code recognition and surface inspection.

Other contributions:

Press contact