A DCT-Based Local Contrast Enhancement Algorithm in SAR Image Target Detection

Synthetic aperture radar (SAR) has become an indispensable remote sensing technology for maritime surveillance. Due to the influence of sea clutter, ship targets may be submerged in background noise, making it difficult for SAR ship target detection. In order to solve this problem, a discrete cosine transform (DCT)-based local contrast enhancement algorithm (DCT-LCE) is proposed in this article. By integrating DCT with sliding window, this algorithm innovatively transforms the SAR image into the DCT domain for processing. A weighted alternating current coefficients calculation method is designed to characterize statistical features within the sliding window, providing a quantitative method for distinguishing between targets and backgrounds. In addition, as optimization and improvement of DCT-LCE, multiscale DCT local contrast enhancement (MDCT-LCE) is proposed to enhance the detailed morphological information of ship targets. Experimental simulations demonstrate that the proposed algorithms can effectively enhance ship targets. Moreover, compared with other sliding window-based algorithms, the proposed algorithms have better detection performance both in accuracy and morphological features under different levels of complexity background.