Abstarct: Evaluation, development, management of reservoirs has a strong relationship to knowing of reservoir traits. Indeed texture and porosity of reservoir have an important and basical role in evaluation petrophysical parameters. As the majority of the Iran reservoirs are carbonate, and also the high heterogeny of these types of reservoirs. Studying of these type of reservoirs, is more important than the other ones. In these type of reservoirs knowing of taxonomy of Dunham class can have a good influence in better estimation of petrophysical parameters of reservoirs. To date, the petroleum industry has tried to determine porosity by injecting helium gas to core samples (plugs) and to determine texture by examining thin sections under the microscope. Laboratory methods are usually time consuming and costly and is not possible in all the circumstances. In today's world oil industry it is dealing with a large number of difficult problems does not meet all the needs of engineers and experts. In recent years, with advances in computer hardware and software, the use of artificial intelligence techniques and image analysis in the petroleum industry expanded. Thus, in order to reduce costs and time in reservoir studies, this study was divided into two parts: At first the technique of artificial neural networks used to calculate the porosity of cores. The results show that intelligent techniques have been successful in estimating porosity. In the second part of this thesis, Carbonate rocks by use of an automatic algorithm are classified according to Dunham's classification. Studies show that the algorithm is acceptable and capable of high precision resolution in Dunham classification. The main limitation of this method is that the texture and porosity data must be synchronized to a harvest field.