Tissue-Selective Contrast Enhancement in Cervical Imaging Using Hemoglobin Mapping Informed by Optical Modeling

White-light imaging (WLI)–based cervical colposcopy is often limited by strong surface reflections, low tissue contrast, and unstable color appearance, making blood vessels and bleeding regions difficult to visualize. Existing image enhancement techniques largely rely on global processing and lack selectivity to hemoglobin absorption, which may introduce pseudo-contrast and reduce clinical interpretability. This study proposes an image enhancement approach based on hemoglobin index (iHb) mapping that integrates cross-polarized imaging, a two-layer Kubelka–Munk (K–M) optical model, and adaptive color enhancement. Cross-polarization suppresses surface reflections, while the optical model establishes the relationship between the red to green intensity ratio (R/G) and hemoglobin absorption. The derived iHb information is adaptively mapped to the RGB image, enabling tissue-selective contrast enhancement driven by hemoglobin content rather than global intensity changes. The proposed method was evaluated using two-layer optical phantoms with graded fractional red blood cell concentrations and representative clinical cervical images, including polyp, Nabothian cyst, and inflammatory tissues. Phantom experiments demonstrated monotonic, non-linear contrast trends consistent with K–M predictions. In clinical images, the method achieved hemoglobin-dependent improvements in lesion-to-background contrast, with the contrast ratio increasing by more than 100%, while preserving image fidelity, as confirmed by high structural similarity (SSIM > 0.98) and low color distortion (ΔE2000 < 2.2). These results suggest that the proposed iHb-guided enhancement provides a physically grounded and clinically interpretable framework for visualizing hemoglobin-rich cervical lesions, with potential for integration into real-time colposcopic imaging systems.