Abstract: To elucidate the mechanism by which shock and obstacle interactions induce local detonation initiation, an experimental investigation was conducted on flame acceleration of premixed hydrogen–air mixtures in an obstacle-laden tube. Experiments were performed over a range of equivalence ratios from 0.8 to 1.3, spanning both fuel-lean and fuel-rich conditions relative to stoichiometry. Particular emphasis was placed on the interaction between the Mach stem of the leading shock and a single obstacle, and on how this interaction governs the formation of localized hot spots and the subsequent onset of local detonation. At an equivalence ratio of 0.8, initiation was not observed. Under this condition, the reflected shock interacted with the flame front and induced Richtmyer–Meshkov instability, generating pronounced wrinkling of the flame surface. This interaction further densified the pre-existing flame folds and increased the flame surface area, thereby accelerating the flame; however, no local detonation was initiated. In contrast, if the equivalence ratio was within the range from 0.9 to 1.3, successful initiation was observed. Shock reflection in the obstacle vicinity generated localized regions of elevated temperature that acted as hot spots, providing favorable conditions for rapid energy release and the onset of a locally detonative event. However, the locally initiated detonation did not develop into a self-sustained stable detonation. During subsequent diffraction, expansion waves imposed pronounced cooling and attenuation, causing progressive decoupling between the leading shock and the reaction zone, and thereby suppressing further development into a stable detonation wave. Through analysis of the critical initiation characteristics, it is found that increasing either the shock strength or the equivalence ratio increases the critical initiation parameter, and the critical initiation parameter is more sensitive to shock strength than to equivalence ratio. Furthermore, considering the discrepancy between the Thomas critical initiation model and the incident shock wave in the present experiments, the shock reflection zone was divided into sections for calculation. The analysis reveals that, at an equivalence ratio of 1.2, the lower sonic velocity following reflection of the leading shock wave prolongs the time taken for the expansion wave to reach the base of the obstacle, thereby favoring initiation. Conversely, at an equivalence ratio of 0.8, the lower equivalence ratio reduces the reactivity of the mixture, leading to a longer ignition delay time and consequently reducing the likelihood of initiation.