Over the past decade, there has been increased interest in quantifying cell populations and morphological structures in tissue sections using image analysis systems. Automated analysis is now being used in limited pathological applications, such as PAP smear evaluation, with the dual aim of increasing the accuracy of diagnosis and reducing the review time. Applications such as these primarily use gray scale images and deal with cells that are well separated. Quantification of routinely stained tissue represents a more difficult problem in that objects can not be separated in gray scale and the images are more complex. Many of the existing semi-automated algorithms are specific to a particular application and are computationally expensive. Hence, we investigated general adaptive automated color segmentation approaches which alleviate these problems.
The first stage of this research concentrated on separating tissue color clusters in a color space histogram of the image.