Parallel methods for preliminary image processing in nondestructive testing

We consider parallel methods for preliminary processing of grayscale images under nondestructive testing of materials and structures. By virtue of the proved theorem, we show that grayscale images belong to the class of ℜ-functions which can be found and allow deep parallelization. We propose a new approach to geometric transformations of images, which can be considered as a single operation of transformation of coordinates. Karpenko Physicomechanical Institute, Ukrainian Academy of Sciences, L'viv. Translated from Fizyko-Khimichna Mekhanika Materialiv, Vol. 33, No. 5, pp. 107–112, September–October., 1997.