Thorsten Cardoen

Dr. Thorsten Cardoen

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
15 May 2025 in Computers and Electronics in Agriculture
Verified
Journal Article
Closed access
TCThorsten Cardoen
Thorsten Cardoen
PCPatricia Soster de Carvalho
Patricia Soster de Carvalho
GAGunther Antonissen
Gunther Antonissen
FTFrank A.M. Tuyttens
Frank A.M. Tuyttens
SLSam Leroux
Sam Leroux
PSPieter Simoens
Pieter Simoens

13 December 2022 in Sensors
Verified
Journal Article
Open access
PDF file
TCThorsten Cardoen
Thorsten Cardoen
SLSam Leroux
Sam Leroux
PSPieter Simoens
Pieter Simoens
3D reconstruction is the computer vision task of reconstructing the 3D shape of an object from multiple 2D images. Most existing algorithms for this task are designed for offline settings, producing a single reconstruction from a batch of images taken from diverse viewpoints. Alongside reconstruction accuracy, additional considerations arise when 3D reconstructions are used in real-time processing pipelines for applications such as robot navigation or manipulation. In these cases, an accurate 3D reconstruction is already required while the data gathering is still in progress. In this paper, we demonstrate how existing batch-based reconstruction algorithms lead to suboptimal reconstruction quality when used for online, iterative 3D reconstruction and propose appropriate modifications to the existing Pix2Vox++ architecture. When additional viewpoints become available at a high rate, e.g., from a camera mounted on a drone, selecting the most informative viewpoints is important in order to mitigate long term memory loss and to reduce the computational footprint. We present qualitative and quantitative results on the optimal selection of viewpoints and show that state-of-the-art reconstruction quality is already obtained with elementary selection algorithms.

14 March 2026 in Poultry Science
Verified
Journal Article
Open access
PSPatricia Soster Soster
Patricia Soster Soster
TCThorsten Cardoen
Thorsten Cardoen
Camila Lopes CarvalhoCamila Lopes Carvalho
Camila Lopes Carvalho
IKImad Khan
Imad Khan
BKBassem Khalfi
Bassem Khalfi
GAGunther Antonissen
Gunther Antonissen
Understanding behavioral changes across the production cycle is essential for broiler welfare assessment. This study monitored three key broiler chicken behaviors over time: activity level (based on net displacement per second) and time spent at the feeder and drinker zones. It includes the analysis of activity dynamics (high, medium, and low activity level), distance traveled over time, the effect of time of day (morning: 9:00-12:00, midday: 13:00-16:00, and evening: 18:00-21:00) on behaviors, and behavior during periods of experimentally induced high environmental temperature (29,4 ± 1°C; from days 29 to 33 and 36 to 40). This study involved four production rounds, each with 280 as hatched Ross 308 broilers (1,120 total), housed in two climate-controlled pens (9×4 m) per round (140 birds/pen; 14 kg/m²). A multi-camera system tracked broilers using a YOLOv11 model and the SORT algorithm. Tracklets (short segments of a moving object's trajectory) were projected onto ground-plane coordinates and merged across views to form unified movement paths. The system achieved high tracking accuracy (MOTA: 0.81; IDF1: 0.89). Video segments (n=1,641) of 3 minutes each were recorded between days 8 to 40. Broiler behaviors from 8 to 40 days of age were significantly influenced by high environmental temperature and age. Heat exposure increased time spent at the drinker and reduced time spent at the feeder zone, while overall activity levels were unaffected. Activity dynamics shifted across the production cycle, with high-intensity activity decreasing earlier and more rapidly than medium and low activity, leading to a progressive dominance of lower activity levels as broilers aged. Behavioral patterns also varied by time of day, with time at feeder zone and activity partially rebounding in the evening after the heat-challenge period. While the system offers valuable monitoring insights, it has only been tested under controlled conditions with low density, and validation under commercial conditions is still needed.

24 June 2026 in Avian Pathology
Verified
Journal Article
Closed access
PSPatricia Soster Soster
Patricia Soster Soster
TCThorsten Cardoen
Thorsten Cardoen
Camila Lopes CarvalhoCamila Lopes Carvalho
Camila Lopes Carvalho
IKImad Khan
Imad Khan
SRSmets Rutger
Smets Rutger
GAGunther Antonissen
Gunther Antonissen
Automated behavioral monitoring may offer opportunities to assess health in broiler chickens. This study aimed to (i) evaluate behavioral changes following challenge with Eimeria spp. and infectious bronchitis virus (IBV) combined with avian pathogenic Escherichia coli (APEC), and (ii) characterize patterns across age and time of the day. Across four rounds, 132 Ross 308 broilers were assigned to three groups (control, Eimeria, and IBV+APEC), with four replicate pens of 11 birds each. Challenges were applied on day 14 (Eimeria), day 21 (IBV), and day 24 (APEC). A total of 12 behaviors were quantified from 10-min video recordings collected every 2 h between 07:00 and 21:00 from days 14 to 41, divided into four periods: Period 1 (d14-20), Period 2 (d21-27), Period 3 (d28-34), and Period 4 (d35-41). Behavior was primarily driven by age. In control birds, locomotion and exploratory/foraging decreased progressively, while sleeping increased; inactivity remained stable. Time-of-day analysis revealed diurnal patterns with higher afternoon/evening activity and lower sleeping in these periods compared with morning. Relative to controls, Eimeria-infected birds showed variable drinking and sustained increased locomotion and exploratory behavior (periods 1-4), representing a non-classical response to intestinal challenge. IBV+APEC-challenged birds exhibited modest changes: decreased inactivity (period 2) and increased panting (period 4), with largely unchanged activity patterns. Disease challenge altered diurnal patterns, with disease-specific effects on circadian-dependent behaviors. These findings highlight the importance of age-specific baselines and the potential of time-of-day-dependent behavioral analyses and circadian markers for detecting disease-related deviations and enabling early disease detection through automated behavioral monitoring.