Borja García Pascual

Mr. Borja García Pascual

School of Forest Sciences, University of Eastern Finland, Joensuu, Pohjois-Karjala, Finland
+1
orcid

Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
24 April 2022 in Sensors
Verified
Journal Article
Open access
PDF file
Francisco Rodríguez-PuertaFrancisco Rodríguez-Puerta
Francisco Rodríguez-Puerta
CBCarlos Barrera
Carlos Barrera
BPBorja García Pascual
Borja García Pascual
Fernando Pérez-RodríguezFernando Pérez-Rodríguez
Fernando Pérez-Rodríguez
Angel Garcia-PedreroAngel Garcia-Pedrero
Angel Garcia-Pedrero
Resilient cities incorporate a social, ecological, and technological systems perspective through their trees, both in urban and peri-urban forests and linear street trees, and help promote and understand the concept of ecosystem resilience. Urban tree inventories usually involve the collection of field data on the location, genus, species, crown shape and volume, diameter, height, and health status of these trees. In this work, we have developed a multi-stage methodology to update urban tree inventories in a fully automatic way, and we have applied it in the city of Pamplona (Spain). We have compared and combined two of the most common data sources for updating urban tree inventories: Airborne Laser Scanning (ALS) point clouds combined with aerial orthophotographs, and street-level imagery from Google Street View (GSV). Depending on the data source, different methodologies were used to identify the trees. In the first stage, the use of individual tree detection techniques in ALS point clouds was compared with the detection of objects (trees) on street level images using computer vision (CV) techniques. In both cases, a high success rate or recall (number of true positive with respect to all detectable trees) was obtained, where between 85.07% and 86.42% of the trees were well-identified, although many false positives (FPs) or trees that did not exist or that had been confused with other objects were always identified. In order to reduce these errors or FPs, a second stage was designed, where FP debugging was performed through two methodologies: (a) based on the automatic checking of all possible trees with street level images, and (b) through a machine learning binary classification model trained with spectral data from orthophotographs. After this second stage, the recall decreased to about 75% (between 71.43 and 78.18 depending on the procedure used) but most of the false positives were eliminated. The results obtained with both data sources were robust and accurate. We can conclude that the results obtained with the different methodologies are very similar, where the main difference resides in the access to the starting information. While the use of street-level images only allows for the detection of trees growing in trafficable streets and is a source of information that is usually paid for, the use of ALS and aerial orthophotographs allows for the location of trees anywhere in the city, including public and private parks and gardens, and in many countries, these data are freely available.

22 December 2024 in Remote Sensing
Verified
Journal Article
Open access
PDF file
Francisco Rodríguez-PuertaFrancisco Rodríguez-Puerta
Francisco Rodríguez-Puerta
RPRyan L. Perroy
Ryan L. Perroy
CBCarlos Barrera
Carlos Barrera
JPJonathan P. Price
Jonathan P. Price
BPBorja García Pascual
Borja García Pascual
The generation of cloud-free satellite mosaics is essential for a range of remote sensing applications, including land use mapping, ecosystem monitoring, and resource management. This study focuses on remote sensing across the climatic diversity of Hawai’i Island, which encompasses ten Köppen climate zones from tropical to Arctic: periglacial. This diversity presents unique challenges for cloud-free image generation. We conducted a comparative analysis of three cloud-masking methods: two Google Earth Engine algorithms (CloudScore+ and s2cloudless) and a new proprietary deep learning-based algorithm (L3) applied to Sentinel-2 imagery. These methods were evaluated against the best monthly composite selected from high-frequency Planet imagery, which acquires daily images. All Sentinel-2 bands were enhanced to a 10 m resolution, and an advanced weather mask was applied to generate monthly mosaics from 2019 to 2023. We stratified the analysis by cloud cover frequency (low, moderate, high, and very high), applying one-way and two-way ANOVAs to assess cloud-free pixel success rates. Results indicate that CloudScore+ achieved the highest success rate at 89.4% cloud-free pixels, followed by L3 and s2cloudless at 79.3% and 80.8%, respectively. Cloud removal effectiveness decreased as cloud cover increased, with clear pixel success rates ranging from 94.6% under low cloud cover to 79.3% under very high cloud cover. Additionally, seasonality effects showed higher cloud removal rates in the wet season (88.6%), while no significant year-to-year differences were observed from 2019 to 2023. This study advances current methodologies for generating reliable cloud-free mosaics in tropical and subtropical regions, with potential applications for remote sensing in other cloud-dense environments.

26 September 2024 in Agronomy
Verified
Journal Article
Open access
PDF file
RPReetta Palva
Reetta Palva
EKEerikki Kaila
Eerikki Kaila
BPBorja García Pascual
Borja García Pascual
VBVictor Bloch
Victor Bloch
Field robots are an important tool when improving the efficiency and decreasing the climatic impact of food production. Although several commercial field robots are available, the advantages, limitations, and optimal utilization methods of this technology are still not well understood due to its novelty. This study aims to evaluate the performance of a commercial field robot for seeding and weeding tasks. The evaluation was carried out in a 2-hectare sugar beet field. The robot’s performance was assessed by counting plants and weeds using image processing. The YOLOv8 model was trained to detect sugar beets and weeds. The plant and weed densities were compared on a robotically weeded area of the field, a chemically weeded control area, and an untreated control area. The average weed density on the robotically treated area was about two times lower than that on the untreated area and about three times higher than on the chemically treated area. The testing robot in the specific testing environment and mode showed intermediate results, weeding a majority of the weeds between the rows; however, it left the most harmful weeds close to the plants. Software for robot performance assessment can be used for monitoring robot performance and plant conditions several times during plant growth according to the weeding frequency.

29 October 2025 in ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Verified
Journal Article
Open access
PDF file
BPBorja García Pascual
Borja García Pascual
CMCarlos Martín-Cortés
Carlos Martín-Cortés
XZXin Zhou
Xin Zhou
ELEvgeny Lopatin
Evgeny Lopatin
MAMauricio Acuna
Mauricio Acuna
KKKalle Kärhä
Kalle Kärhä
Accurate diameter estimation from point cloud data allows for characterizing stem volume and shape without resorting to destructive methods. Typically, circles are fitted at various stem heights using statistical techniques. However, these techniques are susceptible to noise and occlusion in the point cloud, often caused by obstacles or weather phenomena. This susceptibility reduces the feasibility of applying such methods to point clouds captured by low-cost sensors, which tend to be less precise and noisier. Photogrammetry, however, can be used together with consumer-grade cameras and inexpensive UAVs to generate high-quality point clouds from under-canopy data. This study presents MACiF (Morphology-Aware Circle Fit), a novel method to accurately estimate diameters at various heights from noisy point clouds. Our approach uses robust statistical methods and Monte Carlo simulation to filter the point cloud. We also leverage how stems vary gradually to iteratively correct erroneous estimates. This iterative correction enables estimating diameters with an error lower than -3.34 cm, even when data quality limits the use of other methods. These results support the use of undercanopy low-cost photogrammetry as a viable source of data for automatic stem characterization.

22 March 2026
Verified
Conference Paper
Open access
HKHeli Kymäläinen
Heli Kymäläinen
BPBorja García Pascual
Borja García Pascual
CMCarlos Martín-Cortés
Carlos Martín-Cortés
MRMartino Rogai
Martino Rogai
MSMari Selkimäki
Mari Selkimäki
Blas Mola-YudegoBlas Mola-Yudego
Blas Mola-Yudego
This study investigates a gamified real-time assistance concept for forest harvesters, integrating augmented reality (AR) and LiDAR-based data visualization through a head-up display (HUD). The approach draws from gamification principles—enhancing engagement, feedback, and situational awareness—to support operator decision-making and ergonomics. A field test was conducted in Ilomantsi, Finland, employing a John Deere 1170G harvester equipped with an external LiDAR scanner and an in-cabin holographic HUD. Real-time tree detection and diameter estimation were achieved through LiDAR data processing with a YOLO-based convolutional neural network. The system successfully projected tree positions and estimated diameters onto the HUD during the felling of 60 trees. Operator workload, assessed using the NASA-TLX method, indicated low cognitive strain and positive usability feedback. The results demonstrate the feasibility of real-time holographic visualization in dynamic forest environments but highlight calibration and information flow challenges. Further improvements are needed in system setup, eye-position alignment, and data visualization to optimize ergonomic benefits and decision support in mechanized harvesting operations.