Microscobic Imaging with Automatic Scanner

Yaldiz H., EKİNCİ M.

21st Signal Processing and Communications Applications Conference (SIU), CYPRUS, 24 - 26 April 2013 identifier identifier

  • Publication Type: Conference Paper / Full Text
  • Volume:
  • Doi Number: 10.1109/siu.2013.6531380
  • Country: CYPRUS
  • Karadeniz Technical University Affiliated: Yes


First and fundamental function to make an automatic image analysis on microscobic images is automatic implementation of imege sharpness before scanning function of sample, all area of which is analyzed, by scanning on the X-Y-Z plane and automatic continuation of image sharpness during the image scanning process. When automatic scanning, for bacterial analyses, within the boundaries of the sample area on the X-Y plane, automatic detection of outside field area from image sequence is one of the other fundamental functions. Implementation and protection of focus in imaging on all the stage of of the sample area scanning, on the Z axis direction, must be provided. In this work, for imaging provided with CMOS camera from light microscopy motorized system making scanning process automatically by detecting the boundaries of the sample area with visual data on X-Y-Z plane, automatic focus and the success of auto-focusing continuity during scanning are presented. Focusing microscopic imaging is based on optimizing based on analyses of variance image sequences distance in the Z axis direction between lens and sample. After determination of optimization for imaging with automatic focusing in the starting position platform, which placed the sample on, is moved in the X-Y plane for scanning. Focus distortion caused by possible distance changes on the Z axis during scanning process of this movement is automatically detected. Image enhancement for which in low degree of distortion is provided with software, and for which high level is provided by success of auto control of platform motion, in Z axis direction. Review of in/out the field based on multi-layer artificial neural network to detect coming to the limit of scan area is succeed with processing of image sequences.