One of the main problems in artificial vision is the need to increase the spatial resolution of the object images that are studied. We propose a system which is enable to detect zones of interest for the analysis inside an image and to increase the resolution in such areas. We considered as major relevance zones for the analysis, those which constitute joints or fences of the objects in the scene, as the fine obtaining of these contours is the base of any previous operation. The procedure of resolution increase is based on the modelling of the mentioned joints from the information in grey scale of the image.