TY - JOUR AU - Saenz, Jose Gabriel AU - Camargo, Luz Helena AU - Camargo, Esperanza PY - 2014/12/31 Y2 - 2024/03/28 TI - Color descriptors for cytology smear images analysis JF - Visión electrónica JA - Vis. Electron. VL - 8 IS - 2 SE - Visión Investigadora DO - 10.14483/22484728.9873 UR - https://revistas.udistrital.edu.co/index.php/visele/article/view/9873 SP - 67-73 AB - <p>The technique Pap conventional cytology is used as a means of screening to identify normal and abnormal cells, which can prevent cervical cancer. However, a high level of expertise by the cytopathologist performing screening is required, and considering the number of samples to be examined in a working day, increasing the possibility of error. Therefore, several strategies have been implemented to automate this process by taking the original image and converting to grayscale, but this way the information related to the color is lost, a feature that provides a high discriminative power of the elements of interest (nucleus and cytoplasm).</p><p>In this work images obtained in a database of public domain cell cervical cytology, a step of preprocessing was applied to correct lighting and eliminate image noise, then color two descriptors were implemented; the Dominant Color Descriptor (DCD) and the Descriptor of the Distribution of Color (DDC) for characterizing the content of the images. As a result of this study the implementation of a preprocessing stage in the cell image analysis of cervical cytology combined with the use of information related to the color achieved effectively make detection nucleus and cytoplasm, able to develop a method automatic detection of abnormal cells and thus prevent cervical cancer.</p> ER -