Abstarct: The CRS stacking method is a new automatic method in seismic data processing, that is macro model independent and has been born since late 90’s. Application of the CRS stack technique is advantageous in complex areas, since it involves information about the shape of seismic reflectors, i.e., dip and curvature, into processing. Moreover, a multi-parameter formula allows to sum up more traces during the stack. All together, this leads to better imaging results, especially to an improvement of the signal-to-noise (S/N) ratio. Reflection events in the CRS stack section appear clearer and more continuous, compared to the conventional CMP stack section. But, one of the problems of the CRS method, is neglecting the conflicting dips problem in these sections. It means that if two or more events conflict in a same point in the section, the CRS stack method could not handle this problem well.The conflicting dips problem is an important issue in the matter of imaging in complex structures. The CDS method was proposed to solve this problem and in fact it was extension and improvement of the previous method. So in this thesis, I investigate the functionality and results of the conventional, CRS and CDS methods in a complex structure region of IRAN and then implement my method to solve this issue beside a comparison with results of the cited methods. Using migration as a complementary process in seismic imaging, especially Pre Stack Time Migration (PSTM), because it’s short time to perform and also application in regions with complex structure, have made this technique, a useful and applicable method in industry. So we could prepare a high quality data to be input for CRS method by migrating the data before stacking in the time domain. PSTM in complex structures, produces better and clearer images with conflicting dips issues, rather than conventional methods. The CRS method in this case, uses data with better quality, instead of only preprocessed data in usual CRS implementation. Using this high quality data as input for CRS, leads to get a final optimized section with more detected events in late times, more quality and more resolution. This section could be trusted in interpretation, more than previous methods.