This study proposes a novel adaptive image enhancement technique in infrared band for sea-surface targets based on local frequency cues. Fourier transforms of the image blocks are computed and clustering is done according to the number of expected regions to be enhanced. Based on the variations of the elements in any cluster and the differences between the cluster centers, two gain matrices are computed for the enhancement of mid and high frequency images. We provide enhancement results for images obtained at different field trials and compare the performance of the proposed algorithm using subjective and quantitative tests with four different methods. The algorithm shows better performance in the detail visibility of the target.