This article describes a longitudinal dataset of Salicornia plants cultivated in the marine pilot of the BeBlue project’s recirculating aquaponic system at Ca’ Foscari University of Venice, Italy. Five ESP32-CAM modules equipped with OV2640 sensors captured RGB images of potted plants at nominal 15-minute intervals. A total of 61,313 raw images were received by the server between May and October 2025. The published dataset comprises 38,539 images acquired between 29 May and 31 August 2025, covering the monitored plants’ growth cycle up to approximately the end of the cycle. These images were retained after quality filtering and preprocessing. The dataset covers two acquisition phases: mid-growth (Phase I) and near-complete plant development (Phase II). The image data include rotation-corrected red, green, and blue (RGB) frames showing up to three monitored plant positions (left, center, and right) within each camera’s field of view, together with segmented single-plant images generated using the PixelLib library with the Mask Region-based Convolutional Neural Network (Mask R-CNN) instance segmentation model. Morphological features, including projected area, convex-hull area, width, height, aspect ratio, solidity, skeleton path length, and branch and tip counts, were extracted using PlantCV. Image-derived measurements were converted to millimeters using calibration based on a pattern with known dimensions. The dataset also contains water-quality telemetry, including temperature, pH, dissolved oxygen, oxygen saturation, and salinity, together with manual biomass measurements. Water-quality records are available for Phase II, whereas Phase I contains image data without corresponding telemetry because of a sensor malfunction. The data are provided in image collections and structured CSV files, containing image-derived features at a 15-minute resolution and daily records with biomass values interpolated from manual measurements, along with metadata for 17 plant image sequences. The dataset can be reused for image-based plant phenotyping, non-destructive biomass estimation, morphological analysis, growth modeling, and the development of data-driven digital twins for controlled-environment agriculture and aquaponic production systems.

Longitudinal Color Image, Water-Quality Telemetry, and Biomass Data for Salicornia Cultivated in a Marine Aquaponic Pilot in Venice, Italy (May–August 2025)

Sharifi, Ashraf;Tarif, Mehran;Migliorini, Sara
;
Quaglia, Davide;
2026-01-01

Abstract

This article describes a longitudinal dataset of Salicornia plants cultivated in the marine pilot of the BeBlue project’s recirculating aquaponic system at Ca’ Foscari University of Venice, Italy. Five ESP32-CAM modules equipped with OV2640 sensors captured RGB images of potted plants at nominal 15-minute intervals. A total of 61,313 raw images were received by the server between May and October 2025. The published dataset comprises 38,539 images acquired between 29 May and 31 August 2025, covering the monitored plants’ growth cycle up to approximately the end of the cycle. These images were retained after quality filtering and preprocessing. The dataset covers two acquisition phases: mid-growth (Phase I) and near-complete plant development (Phase II). The image data include rotation-corrected red, green, and blue (RGB) frames showing up to three monitored plant positions (left, center, and right) within each camera’s field of view, together with segmented single-plant images generated using the PixelLib library with the Mask Region-based Convolutional Neural Network (Mask R-CNN) instance segmentation model. Morphological features, including projected area, convex-hull area, width, height, aspect ratio, solidity, skeleton path length, and branch and tip counts, were extracted using PlantCV. Image-derived measurements were converted to millimeters using calibration based on a pattern with known dimensions. The dataset also contains water-quality telemetry, including temperature, pH, dissolved oxygen, oxygen saturation, and salinity, together with manual biomass measurements. Water-quality records are available for Phase II, whereas Phase I contains image data without corresponding telemetry because of a sensor malfunction. The data are provided in image collections and structured CSV files, containing image-derived features at a 15-minute resolution and daily records with biomass values interpolated from manual measurements, along with metadata for 17 plant image sequences. The dataset can be reused for image-based plant phenotyping, non-destructive biomass estimation, morphological analysis, growth modeling, and the development of data-driven digital twins for controlled-environment agriculture and aquaponic production systems.
2026
Digital twin, Plant phenotyping, Instance segmentation, Precision agriculture, Internet of Things (IoT) sensing, Plant growth modeling, Time series analysis, Non-destructive monitoring
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11562/1209433
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact