一维核磁共振（NMR）是一种高分辨率的非侵入式检测技术，广泛应用于化合物结构解析与组分分析．然而，在复杂体系检测中，化学位移接近和J偶合导致的谱峰重叠，以及组分浓度差异引起的弱信号淹没，使得低浓度化合物检测面临挑战．为提高拥挤谱区弱信号的检测灵敏度，本文提出了一种基于纵向多自旋阶、Hadamard编码和二量子滤波的融合方法——Hadamard-DQF-LMO．该方法通过二量子滤波器实现纵向多自旋阶检测，显著抑制强信号对弱信号干扰；结合多色跃迁脉冲和Hadamard编码翻转脉冲，实现了多频点的并行采样，进一步提升灵敏度和检测效率．实验结果表明，该方法可有效分离橙汁和混合氨基酸样品中的重叠谱峰，其选择性和信噪比提升效果显著，为复杂体系的NMR分析提供了新策略．

One-dimensional proton nuclear magnetic resonance (NMR) spectroscopy is a high-resolution, non-invasive technique widely used for structure elucidation and composition analysis. However, when applied to complex systems, its effectiveness is often hampered by overlapping peaks from similar chemical shifts and J-coupling, along with concentration variations obscuring weak signals from low-abundance compounds. To enhance the detection sensitivity for weak signals in crowded spectral regions, this work proposes Hadamard-DQF-LMO, integrating longitudinal multi-spin orders (LMOs), Hadamard encoding, and double quantum filtering (DQF). The approach utilizes DQF to selectively detect LMO signals while suppressing strong interference. The incorporation of polychromatic transition pulses and Hadamard-encoded 180^° pulses enables parallel acquisition of multiple frequencies, which significantly improves both sensitivity and detection efficiency. Experiments on orange juice and mixed amino acid samples demonstrate peak separation with enhanced selectivity and signal-to-noise ratio (SNR), offering a novel and effective strategy for the NMR analysis of complex systems.