Abstarct: Identifying and determining various geometries of structural and laxyered geology (such as faults, salt domes, and buried channels) in reflection seismic data has always been one of the biggest challenges for interpreters. According to the physical properties of these events, they can often be identified using seismography. The use of seismic attributes makes it possible to reveal geological phenomena that are not usually identifiable in seismic sections. Each of the seismic attributes, according to the nature and theoretical foundations that govern it, has the ability to identify only a part of the geological events in the seismic data. Therefore, multi-attributes analysis and the use of the combination of multi-attributes information simultaneously can lead to the identification and determination of the geometry of geological structures in seismic data with more accuracy. Therefore, the combination of seismic attributes to reduce the uncertainty in identifying and determining the geometry of subsurface structures is the main goal of this thesis. The data used in this thesis are two synthetic models of the salt dome with simple and complex structures, the real seismic data of the salt dome is related to the Strait of Hormuz, and the real seismic data of the buried channel and the seismic data with a fault are from the dataset of block F3. To identify the structures of a salt dome, buried channel, and fault in the indicated data, first, the seismic attributes related to each geological structure were extracted from the data. Since some of the extracted attributes may have similar information to each other and their simultaneous presence will only cause confusion and increase the number of calculations in the information combination algorithms, the Laplacian score method is used to select the appropriate attribute set. used. After that, the selected set of suitable attributes was considered as the input of information combination algorithms. In the ordered weighted averaging method, at first, the attributes are sorted in descending order using two mechanisms of optimistic and pessimistic weighting, then they are combined using the relevant relationships of the attributes. In the similarity map method, by choosing a baxse point in each attribute and determining the appropriate control radius, each attribute is classified and finally, all attributes will be aggregated together. using the logistic function, the seismic attributes will be transferred to the fuzzy space, where they will be combined together using the fuzzy gamma method and the geometric averaging method. The algorithm of the graph method is baxsed on the fact that each attribute is defined as a graph and then by preparing the similarity matrix baxsed on the similarity of the corresponding nodes in each graph, the edges of the graph will be weighted and finally, the graphs will be combined through the corresponding points that have the most similarity with each other. After implementing the information combination algorithms, two methods of pixel-to-pixel comparison and the F1-Score method were used to quantitatively evaluate the obtained results. For this purpose, at first, by using Otsu’s thresholding method and morphological methods, binary models of each of the results obtained from information combination methods were prepared. Then, the binary models in synthetic data are compared with the binary models that were used in the construction of the model, and the real data are compared with the binary models obtained from the average interpretation of three experienced interpreters, and the accuracy of each method is determined. baxsed on the results obtained and the seismic data used in this study, in order to identify the geological structures of the salt dome and the buried channel, the graph method is more accurate than the other three methods, and in identifying the position of faults the error of the ordered weighted averaging method with a pessimistic weighting mechanism is more accurate than other methods. Also, in order to compare the accuracy of the proposed methods in this thesis with the common methods of the salt dome boundary, real seismic data was determined by using the multilxayer perceptron method, that the four proposed methods, despite the lack of training data They have higher accuracy in identifying the boundary and geobody of the salt dome.