Image processing for in vitro cell tracking

Abstract

In this paper, we propose an image processing method able to detect cells trough in vitro phase-contrast video microscopy. Bodies of the cells and image background are very similar; it does not allow having a good performance of tracking process. The proposed method normalizes original image to obtain better object-background contrast, image is equalized to highlight the object to be tracked, and image is threshold in order to obtain a better object-background contrast. Gray Level Morphological Gradient is applied to obtain more defined contours in the areas surrounded by Halos. Dilatation and Erosion were combined to achieve this goal. A subtraction process was applied after dilation of image to obtain borders. © Springer-Verlag Berlin Heidelberg 2007.