Classifying White Blood Cells from Peripheral Blood Smear Image Using Histogram of Oriented Gradient Feature of Nuclei Shape

Anas Mohd Noor, Haniza Yazid, Aishah Mohd Noor


Researchers developed various methods and algorithms to classify white blood cells (WBC) from a blood smear image to assist the hematologist or for an automatic system to analyze pathological and hematological conditions related to disease quickly, efficiently and accurately. In this work, we proposed a simple technique for the WBC classification from peripheral blood smear image based on the types of nucleus cell. The developed algorithms utilized Histogram of oriented gradient feature (HOG) typically known use for human shape recognition or classification. The Segmentation of WBC nuclei is based on a simple YCbCr color space and K-means clustering. The HOG feature contains information of the nucleus cell shape which then is classified using support vector machine (SVM) and back propagation artificial neural network (ANN). The experimental results establish that the proposed HOG feature is effectively and suitable for the WBC classification based on the shape of nucleus. We are able to categorize the type of WBC based on the nuclei shape with more than 95% of average accuracy.


Histogram of oriented gradient; K-means clustering; YCbCr color space; WBC Classification

Full Text: PDF


  • There are currently no refbacks.


Indonesian Journal of Electrical Engineering and Informatics (IJEEI)
ISSN 2089-3272

Creative Commons Licence

This work is licensed under a Creative Commons Attribution 4.0 International License.

web analytics
View IJEEI Stats