Image Processing of Edge and Surface Defects

1. Front Matter

   Pages 1-10

   PDF

2. Introduction

   • Roman Louban

   Pages 1-8

3. Edge Detection

   • Roman Louban

   Pages 9-29

4. Defect Detection on an Edge

   • Roman Louban

   Pages 31-45

5. Defect Detection on an Inhomogeneous High-Contrast Surface

   • Roman Louban

   Pages 47-70

6. Defect Detection on an Inhomogeneous Structured Surface

   • Roman Louban

   Pages 71-91

7. Defect Detection in Turbo Mode

   • Roman Louban

   Pages 93-97

8. Adaptive Edge and Defect Detection as a basis for Automated Lumber Classification and Optimisation

   • Roman Louban

   Pages 99-128

9. Object Detection on Images Captured Using a Special Equipment

   • Roman Louban

   Pages 129-134

10. Before an Image Processing System is Used

    • Roman Louban

    Pages 135-160

11. Back Matter

    Pages 1-7

    PDF

The human ability to recognize objects on various backgrounds is amazing. Many times, industrial image processing tried to imitate this ability by its own techniques. This book discusses the recognition of defects on free-form edges and - homogeneous surfaces. My many years of experience has shown that such a task can be solved e?ciently only under particular conditions. Inevitably, the following questions must be answered: How did the defect come about? How and why is a person able to recognize a speci?c defect? In short, one needs an analysis of the process of defect creation as well as an analysis of its detection. As soon as the principle of these processes is understood, the processes can be described mathematically on the basis of an appropriate physical model and can then be captured in an algorithm for defect detection. This approach can be described as “image processing from a physicist’s perspective”. I have successfully used this approach in the development of several industrial image processingsystemsandimprovedupontheminthecourseoftime.Iwouldlike to present the achieved results in a hands-on book on the basis of edge-based algorithms for defect detection on edges and surfaces. I would like to thank all who have supported me in writing this book.

• PED
• PES
• REM
• STEM
• ceramics
• image processing
• image processing system

Aus den Rezensionen: “... Die HeIligkeitsverhältnisse an der Kante einer Materialbeschädigung werden als Gauß'sche Verteilung einer Strahlung interpretiert und in einem physikalischen Modell erfasst. Basierend auf diesem Modell wurden neue Methoden entwickelt, mit denen unterschiedliche Fehlertypen unabhängig von den Helligkeitsbedingungen eines aufgenommenen Bildes ermittelt werden können. ... Zahlreiche Anwendungsbeispiele veranschaulichen die theoretischen Ausführungen.“ (in: QZ Qualität und Zuverlässigkeit, March/2010, Issue 3, S. 13)

• Thermosensorik GmbH, Erlangen, Germany

  Roman Louban

Born on February 26, 1954 in Kiev, Ukraine.

1971 – 1976 Study at the State University of Woronezh, Russia. Graduation: Qualified physicist with award.

1977 – 1992 Scientific research at the Institute of Material Problems of the Science Academy, Kiev, Ukraine. Area of studies: plasma physics, composed materials.

1987 Promotion with the grade: Doctor of Engineering Sciences

1992 – 2006 hema electronic GmbH, Aalen, Department: machine vision, Development engineer, Project director. Main topic: Algorithmics for defect recognition on edges and surfaces

Since April 2006 Thermosensorik GmbH, Erlangen, Manager Software Engineering, IP Expert

Policies and ethics