A new approach for cells detection and classification on blood smear images is considered. Benefit of 4-connected over 8-connected component labeling for cell detection is shown. Color and intensity histogram clustering are proposed to extract common features for cells classification. A new approach for k-means initial centroids detection proposed. The algorithms effectiveness was tested
and estimated for some blood smear images. The algorithm examples, figures and result table to illustrate the approach are presented
[1] C. Hc sliding windows: Object localization by efficient subwindow search”, CVPR, 2008.
[2] Pham, Dzung L.; Xu, Chenyang; Prince, Jerry L., "Current Methods in Medical Image Segmentation". Annual Review of Biomedical Engineering 2: 315–337, 2000.
[3] Luigi Di Stefano, Andrea Bulgarelli, “A Simple and Efficient Connected Components Labeling Algorithm,” ICIAP, 10th International Conference on Image Analysis and Processing, pp.322, 1999.
[4] N. Otsu, ‘‘A threshold selection method from gray level histograms,’’ IEEE Trans. Syst. Man Cybern. SMC-9, 62–66, 1979.
[5] MacKay, David, "Chapter 20. An Example Inference Task: Clustering". Information Theory, Inference and Learning Algorithms. Cambridge University Press. pp. 284–292. ISBN 0-521-64298-1. MR 2012999, 2003
[6] Orchard M, Bouman C, “Color quantization of images”. IEEE Trans Signal Process 39(12):2677- 2690, 1991.
[7] P. Maslak, “Normal peripheral blood smear - 1.” http://imagebank.hematology.org/AssetDetail.aspx?A ssetID=3666&AssetType=Asset, September 2008.