In order to quickly and accurately detect ship targets in long-range horizontal observation under sea-sky background, a ship detection method based on the sea-sky line extraction and mixed gray difference is proposed in this paper. Firstly, the image is divided into sky, sea-sky line and sea surface by the edge detection algorithm. The linear fitting method is adopted on the connection points of sea and sky to determine the parameters of sea-sky line. Then the sea-sky line area is extracted as the target area of interest, and the interferences such as sea clutter and noise are removed. The mixed gray difference horizontal and vertical projection are used to locate the target. Finally, the statistical value of the projection curve is analyzed to obtain the coordinates and size of the target. The experimental results show that this method can effectively detect ship targets in long-range horizontal observation under sea-sky background，and it is simple in calculation and fast in processing speed. So it can meet the real-time processing requirements of ship target detection.