Abstarct: Many important geological events and structures relevant to reservoir studies are not easily discernible during seismic sections and in the usual seismic forms. Therefore, efforts are underway to employ various methods and invent new techniques in seismic studies to identify other crucial features of hydrocarbon reservoirs that are significant in the fields of geology and petroleum engineering. One of the critical structures in the pursuit of hydrocarbon resources is the identification and detection of oil-bearing channels. Consequently, the effort to identify these channels holds great importance. In this research, seismic texture attributes have been utilized to identify and detect branches of a buried channel. The aim of this study is to use the Minimum Redundancy Maximum Relevance (MRMR) algorithm along with principal component analysis and linear discriminant analysis methods to select the best subset of seismic attributes and rank them. After selecting the best attributes, buried channels are identified using the linear discriminant analysis classification algorithm. The use of the MRMR algorithm for attribute selection is crucial because the performance and quality of all seismic attributes may not be uniform. The results of the analysis indicated that this method, through the combination of seismic information and feature optimization, demonstrated a significant improvement in the accuracy of seismic channel detection. The MRMR algorithm proved to be an effective tool for selecting seismic attributes, and the use of PCA and LDA was optimized in a way that the optimal combination of features led to a considerable increase in the discrimination of seismic channel bodies in seismic data. The proposed method not only contributes to a quantitative improvement in the accuracy of seismic body detection in channels but also, due to the utilization of analytical and optimization approaches for feature combination, possesses a high interpretability. This approach is presented as an efficient and generalizable solution for seismic body detection in seismic channels.