Estimating 3D Volume of Dirt Particles Using Depth from Shadow
Measuring the volume of dirt particles within the context of technical cleanliness is a crucial task, because the measured 3D shape yields important information concerning the risks those particles might cause while operating the technical component in the intended application. The production process therefore needs to be optimized to take care, that such particles are not introduced into the final product during fabrication. Measuring 3D volume in the range of a few microns concerning the area of a particle implies expensive microscope measuring devices. In this paper we describe a hardware approach combined with software algorithms using eight LED lights for illuminating the specimen from various known directions. The LEDs are used to produce cast shadows of the specimen to be inspected. The circular layout of the LEDs takes care to illuminate the particles out of different directions to yield shadows surrounding the whole object. The brightness of the LEDs can be manipulated to adapt for various specimen with different reflection properties. Our model driven approach uses a grid, which is aligned to the image data in order to sample the surface for estimating the height at those discrete positions. The model-driven approach incorporates the measurement results given by established state-of-the-art microscopes for initial calibration. The estimated heights are illustrated in a 2.5D pseudo-color representation and evaluated from a quality point of view. The approach described in this paper yields a practical solution for estimating 3D shape and volume with little hardware requirements and therefore low maintenance costs.
KeywordsDepth from shadow Height estimation 3D volume estimation Adaptive circular LED illumination
- 1.VDA, Band 19 Prüfung der Technischen Sauberkeit - Partikelverunreinigung funktionsrelevanter Automobilteile (2014)Google Scholar
- 2.Frühberger, P., Stephan, T., Beyerer, J.: Integrating microscopic analysis into existing quality assurance processes. In: 2nd International Multidisciplinary Microscopy and Microanalysis Congress: Proceedings of InterM, 16–19 Oct 2014. Springer International Publishing, Cham, pp. 57–64. doi: 10.1007/978-3-319-04639-6
- 3.Bradski, G.: Dr. Dobb’s J. Softw. Tools (2000)Google Scholar
- 4.Beyerer, J., Puente León, F., Frese, C.: Automatische Sichtprüfung: Grundlagen, Methoden und Praxis der Bildgewinnung und Bildauswertung. Springer Vieweg (2012). doi: 10.1007/978-3-642-23966-3
- 5.Gonzalez, R.C., Woods, R.E.: Digital Image Processing. Pearson Education (2008)Google Scholar