Camera calibration, AUV navigation and index of the SmartBay seabed mapping campaign (11 AUV surveys), SmartBay test site, Galway Bay (53.2283° N, 9.2670° W, 24 m), Graal Tech AUV with DIGI4ECO stereo camera, 2026-07-07/08, DIGI4ECO D4.2 DS2 v1.0

Index record of the DIGI4ECO SmartBay seabed mapping campaign: 11 AUV surveys flown on 7 and 8 July 2026 over the SmartBay test site in Galway Bay, Ireland (53.2283° N, 9.2670° W; about 24 m depth), with the DIGI4ECO stereo camera on a Graal Tech AUV. This record holds the camera calibration, the full AUV navigation of both days and the documentation of the campaign, and lists the 25 records in which the images and the photomosaic are published. The campaign is the source dataset DS2 of DIGI4ECO Deliverable D4.2 "Seabed mapping: optic, laser, and acoustic 2D maps and turbidity maps", with two derived products: the enhanced images (DS2P1) and the 2D photomosaic (DS2P2). Because of Zenodo's size limits, and so that users can download single surveys, the raw and the enhanced images are published as one record per survey (DS2-S01…S11 and DS2P1-S01…S11; the enhanced images of survey S11 are split into two records), and each version of the mosaic as its own record. Every per-survey record includes the calibration and the navigation of its survey, so it can be used on its own. Records of the campaign ID Survey Content Size DOI DS2 all This record: calibration, full navigation, documentation ≈ 0.1 GB 10.5281/zenodo.23572251 DS2-S01 20260707_112455 Raw images (245) 0.73 GB 10.5281/zenodo.23572255 DS2-S02 20260707_113631 Raw images (1 508) 4.63 GB 10.5281/zenodo.23572257 DS2-S03 20260707_121223 Raw images (1 587) 3.02 GB 10.5281/zenodo.23572259 DS2-S04 20260707_130256 Raw images (6 264) 12.62 GB 10.5281/zenodo.23572261 DS2-S05 20260707_134600 Raw images (6 180) 12.52 GB 10.5281/zenodo.23572263 DS2-S06 20260707_144952 Raw images (1 954) 4.23 GB 10.5281/zenodo.23572269 DS2-S07 20260707_151038 Raw images (6 281) 13.08 GB 10.5281/zenodo.23572271 DS2-S08 20260707_154913 Raw images (1 960) 4.11 GB 10.5281/zenodo.23572273 DS2-S09 20260708_104341 Raw images (4 072) 5.46 GB 10.5281/zenodo.23572275 DS2-S10 20260708_111743 Raw images (206, RGB) 0.83 GB 10.5281/zenodo.23572277 DS2-S11 20260708_112513 Raw images (10 450) 25.52 GB 10.5281/zenodo.23572279 DS2P1-S01 20260707_112455 Enhanced images (245) 1.37 GB 10.5281/zenodo.23572281 DS2P1-S02 20260707_113631 Enhanced images (1 508) 8.30 GB 10.5281/zenodo.23572283 DS2P1-S03 20260707_121223 Enhanced images (1 587) 8.96 GB 10.5281/zenodo.23572285 DS2P1-S04 20260707_130256 Enhanced images (6 264) 34.27 GB 10.5281/zenodo.23572287 DS2P1-S05 20260707_134600 Enhanced images (6 180) 33.63 GB 10.5281/zenodo.23572289 DS2P1-S06 20260707_144952 Enhanced images (1 954) 10.17 GB 10.5281/zenodo.23572291 DS2P1-S07 20260707_151038 Enhanced images (6 281) 33.47 GB 10.5281/zenodo.23572295 DS2P1-S08 20260707_154913 Enhanced images (1 960) 10.40 GB 10.5281/zenodo.23572297 DS2P1-S09 20260708_104341 Enhanced images (4 072) 27.28 GB 10.5281/zenodo.23572299 DS2P1-S10 20260708_111743 Enhanced images (206) 1.33 GB 10.5281/zenodo.23572303 DS2P1-S11a 20260708_112513 Enhanced images, frames 1–5225 32.01 GB 10.5281/zenodo.23572309 DS2P1-S11b 20260708_112513 Enhanced images, frames 5226–10450 32.60 GB 10.5281/zenodo.23572313 DS2P2-NC 9 surveys 2D photomosaic, natural colour 24.2 GB 10.5281/zenodo.23572317 DS2P2-CLAHE 9 surveys 2D photomosaic, contrast-enhanced 33.9 GB 10.5281/zenodo.23572315 Surveys Survey Date Images (UTC) Images Planned / median altitude Median depth Pattern 20260707_112455 2026-07-07 09:24:56 – 09:29:00 245 2.0 / 2.01 m 24.1 m Visibility test, 1 fps 20260707_113631 2026-07-07 09:36:31 – 09:44:53 1 508 2.0 / 2.01 m 24.3 m 4 transects, north–south 20260707_121223 2026-07-07 10:12:23 – 10:21:12 1 587 2.0 / 2.01 m 24.2 m 5 transects, north–south 20260707_130256 2026-07-07 11:02:56 – 11:37:40 6 264 2.0 / 2.01 m 23.8 m East–west transects, same area 20260707_134600 2026-07-07 11:46:01 – 12:20:20 6 180 2.0 / 2.01 m 23.5 m East–west, many transects 20260707_144952 2026-07-07 12:49:52 – 13:00:43 1 954 2.0 / 1.96 m 23.1 m North–south, wider spacing 20260707_151038 2026-07-07 13:10:38 – 13:45:31 6 281 2.0 / 1.96 m 22.8 m East–west transects 20260707_154913 2026-07-07 13:49:13 – 14:00:06 1 960 2.0 / 1.96 m 22.5 m North–south, loop closing 20260708_104341 2026-07-08 08:43:41 – 09:06:22 4 072 2.0 / 2.01 m 23.7 m Aborted early because of the conditions 20260708_111743 2026-07-08 09:17:43 – 09:18:51 206 1.5 / 1.50 m 24.4 m Lawnmower, recorded as TIFF 20260708_112513 2026-07-08 09:25:13 – 10:23:16 10 450 1.5 / 1.50 m 24.6 m Lawnmower Altitudes: planned / median measured by the DVL. Survey names are the start time in the camera computer's local time (UTC+2). Campaign and acquisition The SmartBay campaign of DIGI4ECO took place on 7 and 8 July 2026 at the SmartBay test site in Galway Bay, Ireland, centred near 53.2283° N, 9.2670° W, at about 24 m depth. A Graal Tech AUV carried the DIGI4ECO stereo camera (model PDH050S-C, serial number 261300517): two 2448 × 2048 px colour sensors (Bayer BG, 8 bit), a baseline of about 100 mm and one flat glass port, 5.75 mm thick, shared by both cameras. The camera looks down, tilted 16° forward from the vertical. Only the left camera's images are published, since they are all that is needed to build the photomosaic. Most surveys were lawnmower patterns at 2 m altitude, 3 frames per second and 0.3–0.5 m/s. On day 2, more particles in the water and less natural light led to a lower altitude (1.5 m) and higher exposure and gain. Artificial lights were not used, because they made the suspended particles visible. In total, the campaign has 11 surveys and 40 707 images. Files of this record File Content Size DIGI4ECO_SMARTBAY_20260707-08_DS2_calibration_v1.0.zip calibration/SN261300517_20260521/: in-air calibration, Pinax seawater maps and virtual cameras, README, SHA256SUMS ≈ 0.1 GB DIGI4ECO_SMARTBAY_20260707-08_DS2_navigation_v1.0.zip navigation/: full navigation of both days, per-survey excerpts, camera mounting, README, SHA256SUMS ≈ 0.02 GB README_images.md Description of the raw and enhanced images 6 kB README_mosaics.md Description of the photomosaic 5 kB Calibration: in air on 21 May 2026 with a ChArUco board (89 stereo pairs), pinhole model with radial terms k1, k2; RMS reprojection error 0.39 px (left), 0.36 px (right), 0.52 px (stereo). Refraction at the flat port (5.75 mm glass, index 1.492; seawater index 1.34) is handled with the Pinax model (Łuczyński et al., 2017): per-pixel maps turn a debayered raw image into an image from an ideal pinhole camera. The calibration covers both cameras, although only the left camera's images are published. Navigation: 20260707-08_galway_full_navigation.csv holds 239 274 rows of the AUV's navigation-filter estimate at about 5 Hz, from 2026-07-07 08:53:45 to 2026-07-08 16:26:13 UTC: time (Unix epoch, UTC), latitude and longitude (WGS84), z = −depth, roll, pitch, yaw and DVL altitude. per_survey/ holds the verbatim excerpt for each survey, and auv_to_left_camera.yaml the pose of the left camera in the AUV body frame (0.88 m forward, 0.05 m to port, optical axis tilted 16° forward from the vertical). Image file names use the camera computer's local time (UTC+2): subtract 2 hours to match the navigation. Usage notes The full navigation file contains 4 772 bad position fixes, more than 2 km from the site, during transits, descents and time at the surface. None falls inside a survey, so the per-survey files contain none of them; users of the full file should filter by distance to the site. Underwater, the position is the estimate of the AUV navigation filter, not a GPS fix, and its absolute accuracy has not been characterised. On day 1 it was 4–10 m off. The camera and AUV clocks were not synchronised by hardware. Only the left camera's images are published, so stereo depth cannot be computed. Survey 20260708_104341 was aborted early because of the conditions, and 20260708_111743 was recorded as debayered TIFF files, so its raw images are RGB. Neither is part of the photomosaic. References Łuczyński, T., Pfingsthorn, M., Birk, A. (2017). The Pinax-model for accurate and efficient refraction correction of underwater cameras in flat-pane housings. Ocean Engineering, 133, 9–22. https://doi.org/10.1016/j.oceaneng.2017.01.029 Responsible partners CORONIS, Graal Tech, Marine Institute. Licence Creative Commons Attribution 4.0 International (CC BY 4.0). Acknowledgement This dataset was produced within the DIGI4ECO project (Digital Twin-sustained 4D ecological monitoring of restoration in fishery depleted areas). Funded by the European Union under grant agreement No 101112883 (DIGI4ECO). Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Climate, Infrastructure and Environment Executive Agency (CINEA). Neither the European Union nor the granting authority can be held responsible for them.