Image geodesic reconstruction by connected components

In this work, a reconstruction operator is defined using geodesic distance. As a consequence a method of segmentation of gray level digital images based on connected classes theory is presented. Considering that a digital image can be associated to a connected class, the connected theory of components emerges like a powerful tool in the Digital Processing Images. Based on this theory, an algorithm is developed. Possible application fields are enunciated. Segmentation of objects simply related or textured images is performed efficiently by this novel approach.