A novel recursive algorithm for area location using isothetic polygons

In this paper, a novel recursive algorithm is presented for binary image area location. The result of an image segmentation task is usually a set of connected components that belong to certain categories, such as texts, images, fines, etc. A challenging task is to restrict all connected components inside simple geometric shapes. For this purpose, isothetic polygons with minimum number of vertices are used in order to achieve simplicity of description and efficiency of storage. These polygons are defined by a recursive formula, where the resulting areas are calculated from successive additions and subtractions of simple rectangular blocks. The proposed area location method can be applied to the majority of image segmentation tasks. The effectiveness of the proposed method has been tested for tracing all page components of a newspaper page decomposition module as well as for tracing objects in the field of a view of a digital camera