Abstract
This paper introduces a method for cellular sequence recognition based on visual information processing. In this method, by using a simple cell model in the primary visual cortex, cellar sequence images are transformed into enhanced line images. These images are adopted for the input features in learning phase by using boosting algorithms, and the outline of cellar sequence is obtained. The advantages of the method are simplification of parameter tuning for line enhancement and utilization of multidimensional features for learning.
