TITLE:
Development and implementation of an automated system to aid laboratory diagnosis using image processing
AUTHORS:
Álvaro Manoel de Souza Soares, Marco Rogério da Silva Richetto, João Bosco Gonçalves, Pedro Paulo Leite do Prado
KEYWORDS:
Image Processing; Robotics; Automation; Pattern Recognition
JOURNAL NAME:
Journal of Biomedical Science and Engineering,
Vol.6 No.5,
May
27,
2013
ABSTRACT: The
objective of this work is to provide an automatic system to count white blood
cells in a blood smear. To do so an experiment was assembled, composed by a
standard microscope with two step motors coupled to its knobs in order to move
the microscope in x and y directions and a web cam which was mounted in the top
of the microscope responsible for to acquire images from the smear. The step
motors and the web cam are controlled by a microcomputer PC standard via
software developed inDelphi. The motors use
the parallel port to communicate with the PC and the camera use the USB port.
The main idea is to set an initial point into the smear and the automated
system will carry over the smear acquiring images (frames with 640 × 480
pixels) and counting the white blood cells encountered. The double histogram
threshold technique is implemented to initially exclude the red cells from the
image leaving only the white ones. Preliminaries results are obtained and
show that the system is quite fast and has a good capacity of selection, even
when different kinds of smear are used.