Abstract: In order to solve the problem of underwater antenna detection with complex sea surface interference, a sea-sky line detection method based on boundary priori double model Bayesian decision in infrared image is proposed. Firstly, the infrared image in sea-sky background is divided into sub-image blocks, and the main cosine spectrum of the discrete cosine transform of sub-image blocks is used as the sub-image block feature. Then, based on the boundary priori, the Bayesian decision model of sea-sky image blocks is established to roughly divide the sea-sky image blocks. Then, the Bayesian decision of fine partition is established by using the set of rough partitioned sub-image blocks. Finally, the parameters of the sea-sky line model are obtained by using RANSAC algorithm to realize the detection of the sea-sky line. The experimental results show that the method can effectively detect the sea antenna in complex sea-sky background, and has better detection effect and faster operation speed for complex infrared images with more interference such as sea clutter and bright spot.