Innovative surface inspection with multispectral technology and artificial intelligence

Purpose and goal: The objective is to develop and evaluate a concept for a vision-based system for automatic detection- and

classification of surface defects on substances that arise in the manufacturing process in an industrial environment as well as acquire new knowledge about the emerging AI technology and evaluate its advantages and disadvantages in this particular field.

The purpose is to obtain new knowledge about the process and the origin of the defects and to carry out an investigation of new methods for material and process development to minimize surface defects. Expected results and effects:

The project provides a major increase in competence regarding the occurrence of surface defects on long products in the steel industry. The collected information together with the new knowledge will together form a tool to reduce the number of surface defects and systematically work to prevent their occurrence. One result is an on-line camera system, adapted for each application, which automatically detects and visualizes surface defects.

Approach and implementation:

The camera system is to be optimized and used at all five industrial partners. In parallel, optimization of image analysis for automatic detection of defects will take place. In WP1, among other things, suitable measurement locations, detection requirements and robustness will be defined.

In WP2, design for vision systems and AI is developed but also preparation for the industrial trials. WP3. includes industrial trials at the various companies. In WP4. evaluation of the AI technology and its pros and cons. WP5 includes dissemination of results such as reporting and meetings.