Consumer cameras, particularly onboard smartphones and UAVs, are now commonly used as scientific instruments. However, their data processing pipelines are not optimized for quantitative radiometry and their calibration is more complex than that of scientific cameras. The lack of a standardized calibration methodology limits the interoperability between devices and, in the ever-changing market, ultimately the lifespan of projects using them. We present a standardized methodology and database (SPECTACLE) for spectral and radiometric calibrations of consumer cameras, including linearity, bias variations, read-out noise, dark current, ISO speed and gain, flat-field, and RGB spectral response. This includes golden standard ground-truth methods and do-it-yourself methods suitable for non-experts. Applying this methodology to seven popular cameras, we found high linearity in RAW but not JPEG data, inter-pixel gain variations >400% correlated with large-scale bias and read-out noise patterns, non-trivial ISO speed normalization functions, flat-field correction factors varying by up to 2.79 over the field of view, and both similarities and differences in spectral response. Moreover, these results differed wildly between camera models, highlighting the importance of standardization and a centralized database.

Pine Wilt Disease is one of the most destructive pests affecting coniferous forests. After being infected by the harmful Bursaphelenchus xylophilus nematode, most trees die within one year. The complex spreading pattern of the disease and the tedious hard labor process of diagnosis involving field wood sampling followed by laboratory analysis call for alternative methods to detect and manage the infected areas. Remote sensing comes naturally into play owing to the possibility of covering relatively large areas and the ability to discriminate healthy from sick trees based on spectral characteristics. This paper presents the development of machine learning classification algorithms for the detection of Pine Wilt Disease in Pinus pinaster, performed in the framework of the European Commission’s Horizon 2020 project “Operational Forest Monitoring using Copernicus and UAV Hyperspectral Data” (FOCUS) in two provinces of central Portugal. Five flight campaigns have been carried out in two consecutive years in order to capture a multitemporal variation of disease distribution. Classification algorithms based on a Random Forest approach were separately designed for the acquired very-high-resolution multispectral and hyperspectral data, respectively. Both algorithms achieved overall accuracies higher than 0.91 in test data. Furthermore, our study shows that the early detection of decaying trees is feasible, even before symptoms are visible in the field.

Detailed knowledge of the intra‐field variability of soil properties and crop characteristics is indispensable for the establishment of sustainable precision agriculture. We present an approach that combines ground‐based agrogeophysical soil and aerial crop data to delineate field‐specific management zones that we interpret with soil attribute measurements of texture, bulk density, and soil moisture, as well as yield and nitrate residue in the soil after potato ( Solanum tuberosum L.) cultivation. To delineate the management zones, we use aerial drone‐based normalized difference vegetation index (NDVI), spatial electromagnetic induction (EMI) soil scanning, and the EMI–NDVI data combination as input in a machine learning clustering technique. We tested this approach in three successive years on six agricultural fields (two per year). The field‐scale EMI data included spatial soil information of the upper 0–50 cm, to approximately match the soil depth sampled for attribute measurements. The NDVI measurements over the growing season provide information on crop development. The management zones delineated from EMI data outperformed the management zones derived from NDVI in terms of spatial coherence and showed differences in properties relevant for agricultural management: texture, soil moisture deficit, yield, and nitrate residue. The combined EMI–NDVI analysis provided no extra benefit. This underpins the importance of including spatially distributed soil information in crop data interpretation, while emphasizing that high‐resolution soil information is essential for variable rate applications and agronomic modeling.

High quality fruit production requires the regulation of the crop load on fruit trees by reducing the number of flowers and fruitlets early in the growing season, if the bearing is too high. Several automated flower cluster quantification methods based on proximal and remote imagery methods have been proposed to estimate flower cluster numbers, but their overall performance is still far from satisfactory. For other methods, the performance of the method to estimate flower clusters within a tree is unknown since they were only tested on images from one perspective. One of the main reported bottlenecks is the presence of occluded flowers due to limitations of the top-view perspective of the platform-sensor combinations. In order to tackle this problem, the multi-view perspective from the Red–Green–Blue (RGB) colored dense point clouds retrieved from drone imagery are compared and evaluated against the field-based flower cluster number per tree. Experimental results obtained on a dataset of two pear tree orchards (N = 144) demonstrate that our 3D object-based method, a combination of pixel-based classification with the stochastic gradient boosting algorithm and density-based clustering (DBSCAN), significantly outperforms the state-of-the-art in flower cluster estimations from the 2D top-view (R² = 0.53), with R² > 0.7 and RRMSE < 15%.