An image processing system for track-etch detectors

Abstract An automated recognition system has been developed which identifies, counts and classifies nuclear tracks on solid state nuclear track detectors (SSNTD). Special care is taken for overlapping tracks, allowing analysis of films with high track densities up to 60,000 tracks/cm 2 . The performance of the proposed method, with a counting error of 3% for a track density of 30,000 tracks/cm 2 , is superior to conventional track counting techniques when evaluating either CR-39 or Lexan SSNTDs. The recognition system evaluates each image in less than 2 s with a PC-486 at 66 MHz, allowing its integration in an on-line automated track detection set-up.