<p>Under the binocular vision, the maritime search and rescue target is disturbed by dense fog and uncertain morphological features of sea conditions, which makes it difficult to locate the target accurately. This paper puts forward an accurate location method for maritime search and rescue targets based on virtual reality visual location technology. The virtual imaging model of maritime search and rescue target under binocular vision is established, the edge contour feature values of maritime search and rescue target imaging under binocular vision are extracted, the distribution area and image pixels of maritime search and rescue target is fused by corner location detection method, the correlation feature quantity of maritime search and rescue target under binocular vision is extracted, and the background interference points of maritime search and rescue target image under binocular vision are separated by steady feature point location fusion method. Combining multi-source beam positioning and feature point positioning methods, the azimuth estimation and dynamic positioning of maritime search and rescue target images under binocular vision are carried out, and the precise positioning and recognition of maritime search and rescue targets under binocular vision are realized in the virtual scene simulation model. The simulation results show that the accuracy of maritime search and rescue target positioning based on binocular vision is high, the resolution of feature extraction is good, and the ability of dynamic positioning, positioning detection and recognition of maritime search and rescue targets is improved.</p>