<p>As a critical channel for sea-air energy exchange, the precise detection of Arctic leads is essential for understanding the mechanism of climate change and ensuring the safety of Arctic navigation. In this study, we address the limitations of existing remote sensing products, such as insufficient spatial resolution, high sensitivity to cloud interference, and limited adaptability of detection methods by leveraging the visible and thermal infrared multispectral synergistic observation capability of China’s FY-3D/MER-SI-II. We introduce the visible channel of FY-3D/MERSI-II, constructing a synergistic observation system with the thermal infrared channel to overcome the inherent limitations of single-channel observation, addressing cloud interference in Arctic lead detection by establishing a pixel credibility model, and propose an adaptive iterative fuzzy classification (AIFC) method to effectively surmount the technical bottleneck of poor adaptability of traditional threshold methods when applied to complex underlying surfaces. The results indicate that from April 1 to 30, 2021, the proposed method effectively suppresses false alarms due to thin cloud cover through the incorporation of pixel credibility, thereby achieving a lower false detection rate than the Willmes lead product. The average probability of lead detection from this method reaches 0.912±0.041, which is 10.4% higher than that of the iterative threshold method, and the probability of background detection is increased to 0.987±0.035. In addition, the average precision rate reaches 0.872, and the <i>F</i>1-score reaches 0.913. The proposed method effectively identifies leads of varying scales across different Arctic regions, demonstrating that the combination of multispectral synergy and fuzzy classification significantly enhances the detection accuracy of Arctic leads.</p>