The sixth-generation mobile communication system is expected to provide sensing services, and thus the integrated sensing and communication waveform has emerged as a hot research topic. In typical sensing systems, such as radar, sensing resolution is often constrained by available wireless resources. Achieving super-resolution sensing within the resource constraint poses a significant challenge. To this end, this work proposed a calibration algorithm for orthogonal frequency division multiplexing waveforms to conduct barycenter calibration (BC) of the multiple signal classification (MUSIC) spectrum that performs two-dimensional sensing. Simulation results demonstrated that the BC algorithm could significantly reduce sensing errors. Even at the edge of the MUSIC spectrum, it still exhibits robust performance. In addition, to comprehensively evaluate the performance gains, we introduce the capacity-to-error ratio and provide simulation results and discussions.