Deep learning-based road crack segmentation can achieve high benchmark performance on public datasets, but its reliability in UAV-acquired field imagery remains uncertain because real road scenes contain various crack-like non-crack features. This study assessed the field transferability of a public-dataset-trained crack segmentation model using UAV-acquired road images. Nine semantic segmentation model configurations were trained and compared using integrated public crack datasets, and U-Net++ with a ResNet50 backbone was selected for UAV field application. After confidence-based data refinement and training data augmentation, the final model achieved precision, recall, Dice, and IoU values of 0.802, 0.787, 0.794, and 0.659, respectively, on the public validation dataset. Field evaluation was conducted on 148 UAV images and manually corrected UAV ground-truth patches generated from non-overlapping 8 × 8 image grids. The total UAV field evaluation produced precision, recall, Dice, and IoU values of 0.856, 0.806, 0.830, and 0.710, respectively. Major errors were associated with crack-like non-crack factors, including lane markings, manholes, repaired boundaries, shadows, surface contamination, and reflections. Threshold sensitivity analysis indicated that approximately 0.4–0.5 was a practical operating range under the tested UAV conditions. These results show that idealized public-dataset evaluation can overestimate real-world UAV performance and that field transferability evaluation is essential for reliable UAV-based pavement inspection.

Unmanned Aerial Vehicles (UAVs) provide high-resolution imagery for detailed object detection, but wide-angle lenses can introduce radial and tangential distortions that affect object geometry and detection performance. This study evaluated the effects of lens distortion correction and image preprocessing on UAV-based vehicle detection. Four preprocessing conditions were compared: Original Image (OI), Modified Original Image (MOI), Bilinear, and Nearest, where Bilinear and Nearest denote lens-corrected images generated using bilinear and nearest-neighbor interpolation. Two convolutional neural network (CNN)-based two-stage object detectors, Faster R-CNN and Cascade R-CNN, were combined with three backbone networks, ResNet-50 (R50), ResNet-101 (R101), and ResNeXt-101 (X101). In total, 24 experimental conditions were evaluated using COCO-based AP and AR metrics. At the final checkpoint, Cascade R-CNN achieved AP and AR values of 0.644 and 0.724, respectively, compared with 0.619 and 0.706 for Faster R-CNN. Among the backbones, R50 showed the most favorable balance between detection accuracy and training time. For image preprocessing, MOI achieved the highest final AP and AR values of 0.662 and 0.737, followed by OI, Bilinear, and Nearest. Bilinear and Nearest retained only 0.944 and 0.938 of the final AP of OI, respectively. These findings indicate that lens distortion correction does not necessarily improve UAV-based vehicle detection and that preprocessing strategies should be selected by considering both geometric correction and downstream detection performance.

Portable sun photometers are useful for atmospheric characterization at satellite calibration and validation (CAL/VAL) sites where permanent refFerence instruments are unavailable, but conventional MICROTOPS II operation requires manual solar pointing and is not well suited to unattended, temporally regular observations. This study developed an automated solar tracking and supervisory control system for MICROTOPS II and evaluated reference-based calibration transfer against a collocated AErosol RObotic NETwork (AERONET) CIMEL CE-318 for CAS500-4. Two Model 540 instruments were operated during three field days at the Jeonju_RDA AERONET site. A custom Python (version 3.10.20) controller automated serial communication, location and time synchronization, fixed interval scanning, data retrieval, calibration constant logging, and local network monitoring. Calibration parameters were estimated from two complete observation days and then frozen before evaluation on the remaining day in a threefold leave-one-day-out (LODO) design. The pooled out-of-fold (OOF) dataset contained 147 pooled OOF matched pairs for each aerosol channel and 147 matched pairs for precipitable water. CE-318-based calibration reduced the RMSE for NASCM540 to 0.0088–0.0188 across 340–870 nm and reduced PKNUM540 RMSE to 0.0117 at 440 nm and 0.0080 at 500 nm; the largest reduction was 89.8% for PKNUM540 500 nm aerosol optical depth (AOD). Residual performance was weakest at 870 nm for PKNUM540 (RMSE = 0.0282, R² = 0.280). Cross-validated temperature compensation improved 1020 nm AOD estimates (RMSE 0.0234 to 0.0199) and precipitable water estimates (0.1177 to 0.1100 cm) but did not improve AOD estimates at 870 nm. The results demonstrate that automated acquisition and independent-day-wise calibration transfer can substantially reduce systematic error while revealing wavelength specific limitations that must be resolved before broader seasonal or spatial transfer for satellite CAL/VAL.