In order to achieve the fast and accurate perception of the shortwave spectrum, a shortwave spectrum sensing algorithm based on digital channelization and dual- threshold stochastic resonance detection is proposed under low SNR. A dual power detection threshold is set for one or more sub-bands extracted from the digital channelized output subchannel to screen out certain channels and suspicious channels, and a good detection performance under low SNR is achieved with the method of stochastic resonance energy detection. Then, the suspicious channel is further detected so as to achieve fast and accurate sensing of the short-wave spectrum. Theoretical derivation and simulation results show that the algorithm can effectively reduce the impact of noise variance uncertainty on the spectrum sensing performance and improve the flexibility and adaptability of the algorithm. Compared with traditional energy sensing, the detection performance is greatly improved.