Segmentation of microscope cell images via adaptive eigenfilters (MA-L4)

Author(s) : Saravana Kumar (National University of Singapore, Singapore) Sim Heng Ong (National University of Singapore, Singapore) Surendra Ranganath (National University of Singapore, Singapore) Fook Tim Chew (National University of Singapore, Singapore) Tan Ching Ong (National University of Singapore, Singapore)

Abstract : This paper presents the use of adaptive eigenfilters to solve the problems of uneven illumination and lighting variation in segmenting microscope cell images. The eigenfilters adapt to the image being considered and is therefore able to segment accurately for a broad range of image conditions. Principal component analysis (PCA) is first applied to the RGB color bands of the image. The image corresponding to the principal component has significantly better contrast over the original image, and a set of eigenfilters is obtained by applying PCA to local neighborhoods of this image. The correlation of this image with its unique set of eigenfilters results in a corresponding set of eigenimages. The eigenimages with small entropy values and large eigenvalues are used to define the edge pixels. The proposed segmentation technique is accurate and robust against uneven illumination, lighting variation and noise.

Menu