Digital Image Processing Techniques for Leukemia Detection

Mukesh Kumar Saini, Arun Saini, Sachin Gupta

Abstract: This paper presents a methodology for the detection of leukemia using various digital image processing techniques. By identifying and analyzing red blood cells and immature white cells, diseases such as anemia, leukemia, malaria, and vitamin B12 deficiency can be diagnosed. The objective is to detect and quantify leukemia- affected cells, determining whether the condition is chronic or acute. Techniques such as histogram equalization, linear contrast stretching, morphological operations (area opening, area closing, erosion, dilation), watershed transform, K- means clustering, and shape-based features are employed. The accuracy of these methods is 72.2%, 72%, 73.7%, and 97.8%, respectively.

Keywords: Leukemia detection, image processing, K-means clustering, watershed transform, histogram equalization, shape-based features, red and white cell counting