Please use this identifier to cite or link to this item:
|Title:||Malaria infection detection in color blood cell images using local regional features|
|Source:||Xiong, W.,Ong, S.H.,Lim, J.H. (2010). Malaria infection detection in color blood cell images using local regional features. APSIPA ASC 2010 - Asia-Pacific Signal and Information Processing Association Annual Summit and Conference : 753-756. ScholarBank@NUS Repository.|
|Abstract:||Large variations in coloration and illumination are found in Giemsa-stained malaria-infected blood smear images. Traditional automatic approaches to infection detection are sensitive to these variations as they rely on ad hoc and unsatisfactory segmentation and classification techniques. We propose a robust unsupervised approach to infection detection and classification without using segmentation. This is based on a specific morphological pattern that corresponds to infected regions containing blob-like structures within cells. We first detect blobs using a blob detector robust to color and illumination changes. From the detected blobs, all the blob contexts of the morphological pattern are identified and, subsequently, color distributions of the respective infected and uninfected regions are derived. Finally, we apply Bayesian rules to estimate the infection status of other blobs not in the patterns. Good performance has been demonstrated in validation experiments using 200 images with large inter- and intra-variations in coloration and illumination.|
|Source Title:||APSIPA ASC 2010 - Asia-Pacific Signal and Information Processing Association Annual Summit and Conference|
|Appears in Collections:||Staff Publications|
Show full item record
Files in This Item:
There are no files associated with this item.
checked on Dec 9, 2017
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.