Abstract: The presence of small geological structures is the main cause of coal and gas outbursts, making the precise detection of these structures an urgent need. High-precision detection of coal-rock interfaces is essential for identifying small structures and achieving transparent mining faces. The acoustic remote reflection logging technology, with a large detection range, a high resolution, and imaging capability, can accurately identify coal-rock interfaces. In this context, this study proposed a technique for coal-rock interface detection based on acoustic remote reflection logging within crossing boreholes. Specially, by placing acoustic detectors in crossing boreholes, the array waveforms generated by the coal-rock interfaces around the boreholes were collected. Then, the coal-rock interface images were obtained through the inversion of the reflected wave information. This technology, combined with the borehole group in the mining face, allows for the overall exploration of the mining face. The steps are as follows: (1) A numerical model of monopole acoustic remote reflection logging for coal seams was established using the COMSOL Multiphysics software. (2) Through forward modeling, the entire spatio-temporal evolutionary laws of full waveform signals and wavefield snapshots were analyzed. (3) The inversion of the acoustic data for acoustic remote reflection logging was performed, enabling the migration imaging of coal-rock interfaces. The forward modeling results indicate that the compressional wave velocity in a coal seam was approximately 1.2 km/s slower than that in its roof and floor. Acoustic waves exhibited faster energy decay when propagating in a coal seam and showed dominant frequency shifting when spreading through coal-rock interfaces. When a measurement point approached the position where an acoustic detector shifted from the rock layer on a coal seam’s floor to the coal seam, direct waves exhibited a sharp decrease in the amplitude and an increase in sonic time, while reflected waves from the interfaces displayed changes in the slopes of the inclined events in the time-depth domain. Inversion was completed through four steps: filtering, wavefield separation, reflection wave enhancement, and migration imaging. The imaging results closely resembled the original model, with the coal-rock interface dip angles and coal thickness exhibiting errors of 0.6° and 0.212 m, respectively. Therefore, the acoustic remote reflection logging with crossing boreholes based on the finite element method allowed for the effective inversion of the positions and morphological characteristics of coal-rock interfaces. This study will provide fundamental theoretical support for the application of acoustic remote reflection logging technology in the coal-rock interface identification through crossing boreholes.