To obtain more prominent target information and retain more textural details in infrared and visible light fusion images, an infrared and visible light image fusion algorithm based on the non-subsample shearlet transform (NSST) domain combined with a spiking cortical model (SCM) and improved fuzzy C-means clustering model (FCM) is proposed. First, the infrared target information in the source infrared image is extracted by the FCM. Subsequently, the NSST is used to decompose the target and background areas of the infrared and visible images to obtain their own high- and low-frequency sub-band images. Subsequently, different fusion strategies are adopted for different regions, and the SCM and improved time matrix are adopted for high-frequency background regions. The final fused image is obtained by using the NSST inverse transform. Simulation experiments show that, compared with other methods, the fusion image obtained by this algorithm has a prominent infrared target and intricate texture details in subjective vision, and its information entropy and edge retention factor are optimal for objective evaluation.