Shadow-Calibrated Stereo Vision for Colorimetric Sweat Analysis

The image recognition technologies, such as facial recognition and object detection, have found widespread application across diverse fields. However, conventional monocular camera systems capture only 2D information, rendering the accurate reconstruction of 3D morphological features challenging and thereby limiting their utility in precision measurement applications. To address this limitation, this study proposes a novel shadow-assisted 3D calibration methodology. This approach reconstructs 3D morphology from a monocular camera viewpoint by introducing a controllable light source and leveraging shadow geometry. In this work, we establish a mathematical model for calibrating 3D structures using shadows and employ Convolutional Neural Network-based 2D image analysis to achieve volumetric calibration. Applied to 3D hydrogel swelling colorimetric analysis for biomarker concentration determination, this method achieves a coefficient of determination (R²) exceeding 0.998. The utilization of shadow-calibrated 3D structural analysis yields a substantial enhancement in analytical accuracy. By design, this "shadow-calibrated" strategy overcomes the dimensional constraints inherent to traditional 2D colorimetric analysis.