earticle

영어

This paper proposes a novel method of connected component labeling and extraction that segments and removes dirt and interphase cells from the chromosome images. This technique was tested with a standard clinical database of human karyotype and excellent segmentation results were achieved. The process involves identifying the various connected components in the input image and assigning labels to create a Label Matrix using the Color Map for these connected components. The connected components or objects that have fewer than the predefined amount of pixels are removed from the image that produces another image where the chromosome except the dirt, stain and interphase cells are uniquely identified and removed. The segmented image is subtracted from the original image, leaving behind only the chromosomes with no dirt, stain and interphase cells, facilitating accurate karyotyping procedures. This technique is extremely helpful when the unwanted object to be segmented and removed shares common intensity levels with the desired information where traditional threshold based procedure will fail to accomplish precise segmentation results.