Abstract: In-Seam Seismic is a key method for exploring geological structures within deep coal seams and has the potential to overcome the challenge of accurately detecting hidden, small-scale faults.This method primarily utilizes the velocity and amplitude attenuation characteristics of channel waves. However, the stability and accuracy of inversion results are significantly affected when geophone coupling is suboptimal or geological anomalies are pronounced.To enhance the stability of transmitted channel wave exploration results and expand its application potential, this study introduces a multi-attribute fusion exploration method.Channel waves have unique frequency and waveform characteristics, enabling the derivation of a comprehensive set of proprietary attributes.The application of multi-attribute analysis and fusion can assist in identifying coal seam structures.The methods for calculating four channel wave attributes—centroid frequency, arc length, bandwidth and spectral ratio—are presented, and a weighted data fusion approach was developed for these attributes.Upon analyzing the numerical simulation results of the model featuring a fault, it becomes evident that the four-attribute fusion data exhibits a closer-to -linear attenuation trend with distance and demonstrates greater stability and sensitivity in responding to fault geological anomalies.By processing and analyzing tomographic imaging results of field measured data, the four attributes and fusion of transmitted channel waves were successfully obtained.The fusion of centroid frequency, arc length, and spectral ratio provides superior characterization of large-scale faults, while the combination of centroid frequency,bandwidth and spectral ratio shows enhanced sensitivity to small faults and other anomalies. The four-attribute fusion of channel wave collectively responded to varying sizes of six faults in the coal seam working face.The multi-attribute fusion tomography method of transmitted channel wave was proposed and implemented. Its effectiveness for geological structure exploration in coal seams was verified. Moreover, it has demonstrated greater stability and accuracy in field applications, presenting a novel idea for channel wave data processing.