Aiming at the problems of long-distance sea-scanning small targets with low contrast, easy interference and difficult extraction, an automatic detection and threat calculation method for infrared sea-scanning small targets was proposed. By modeling and analyzing the infrared imaging system, combining the position of the image sea-sky line to determine the target threat area, and then using morphological filters to suppress the background of the image, and finally counting the characteristic information of the suspicious target, evaluating the threat level of the suspicious target, and selecting the threat level, the highest target is tracked. The experimental results show that the algorithm has a good detection effect on targets near the sea antenna. The threat coefficient of the sea surface target is distributed between 0 and 5. 104, and it is related to the target brightness, scale and position. The calculated threat coefficient is consistent with human subjective judgment. The degree is very high, and the algorithm has important reference value.