In ideal conditions, the use of seismic data can provide a lot of underground information about the structural condition of oil wells and reservoir properties of reservoir rocks. Buried channels are considered as significant geological phenomena, which, in addition to having reservoir potential, could have risks in drilling, which must be identified on seismic data before drilling. Therefore, the identification of buried channels can be important and necessary. Continuous wavelet transform is one of the spectral decomposition methods for non stationary signal analysis in order to achieve better time and frequency resolution of the signal. In this research, continuous wavelet transform with frequencies of 25, 35, and 45 Hz was applied to the real seismic data, and then on each of these frequency maps, appropriate edge detection attributes were used to identify the buried channel, such as: Energy, Instantaneous amplitude and Curvature gradient are applied. Then, by using the methods of Principal Component Analysis (PCA), multiplication and summation of scales to combine different attributes for several frequencies. The internal texture of the buried channel was observed distinctly from its external texture with all three techniques. In the method of PCA, the boundaries of the buried channel can be more clearly distinguished from other points than the multiplication and summation of scales. In addition, the sub-branches of the buried channel can be seen with a higher degree of clarity. The proposed approach can be considered as an alternative to the conventional method of identifying the buried channel.