Validation of body weight monitoring by a 3D camera in finisher pigs

TY - GEN

T1 - Validation of body weight monitoring by a 3D camera in finisher pigs

AU - Amorim Franchi, Guilherme

AU - Bus, Jacinta Daniëlle

AU - Boumans, Iris

AU - Webb, Laura

AU - Bokkers, Eddie

AU - Jensen, Margit Bak

AU - Pedersen, Lene Juul

N1 - Conference code: 10

PY - 2022/9

Y1 - 2022/9

N2 - Automated recording of body weight (BW) can be a useful tool for continuous health, performance, and welfare monitoring in commercial pigs. Hence, we aimed to validate a three-dimensional (3D) camera (iDOL65, DOL-sensors A/S, Aarhus, Denmark) for monitoring BW in conventional finisher pigs on two farms (F1:131 Yorkshire×Landrace pigs, 10-18 pigs/pen; F2:107 Landrace×Large White pigs, 9-11 pigs/pen). On each farm, BW data was recorded on one day. The camera was placed above the multi-partitioned feeder (F1) or individual feeding station (F2) and combined with a radio frequency identification system. Whenever a pig visited the feeding site, 3D images were taken and used to estimate BW. The individual estimated BW was calculated as the median of the outcome of all daily images. For each farm separately, the individual estimated BW was compared against the individual scale-based BW (F1 (average±SD):36±5 kg; F2:74±8 kg) in R. On Farm 1, concordance correlation coefficient (CCC) (0.85) and coefficient of determination (R 2=0.95) were high. The root mean square error (RMSE) was 1 kg. On Farm 2, CCC (0.92) and R 2 (0.94) were very high, and the RMSE was 1.9 kg. The camera’s BW estimation performance was high for both farms, especially in F2. Differences in camera set-up potentially influencing body boundary detection, and differences in body shape arising from different breeds may explain the better camera performance on F2 over F1. A further validation including a larger sample size and pigs at various development stages is necessary to confirm the use of this system in commercial farms.

AB - Automated recording of body weight (BW) can be a useful tool for continuous health, performance, and welfare monitoring in commercial pigs. Hence, we aimed to validate a three-dimensional (3D) camera (iDOL65, DOL-sensors A/S, Aarhus, Denmark) for monitoring BW in conventional finisher pigs on two farms (F1:131 Yorkshire×Landrace pigs, 10-18 pigs/pen; F2:107 Landrace×Large White pigs, 9-11 pigs/pen). On each farm, BW data was recorded on one day. The camera was placed above the multi-partitioned feeder (F1) or individual feeding station (F2) and combined with a radio frequency identification system. Whenever a pig visited the feeding site, 3D images were taken and used to estimate BW. The individual estimated BW was calculated as the median of the outcome of all daily images. For each farm separately, the individual estimated BW was compared against the individual scale-based BW (F1 (average±SD):36±5 kg; F2:74±8 kg) in R. On Farm 1, concordance correlation coefficient (CCC) (0.85) and coefficient of determination (R 2=0.95) were high. The root mean square error (RMSE) was 1 kg. On Farm 2, CCC (0.92) and R 2 (0.94) were very high, and the RMSE was 1.9 kg. The camera’s BW estimation performance was high for both farms, especially in F2. Differences in camera set-up potentially influencing body boundary detection, and differences in body shape arising from different breeds may explain the better camera performance on F2 over F1. A further validation including a larger sample size and pigs at various development stages is necessary to confirm the use of this system in commercial farms.

KW - Sus scrofa

KW - animal welfare

KW - depth sensor

KW - image analysis

KW - precision livestock farming

UR - http://www.scopus.com/inward/record.url?scp=85172381081&partnerID=8YFLogxK

M3 - Article in proceedings

SN - 978-83-965360-0-6

SP - 551

EP - 557

BT - Precision Livestock Farming 2022 - Papers Presented at the 10th European Conference on Precision Livestock Farming, ECPLF 2022

A2 - Berckmans, Daniel

A2 - Oczak, Maciej

A2 - Iwersen, Michael

A2 - Wagener, Karen

PB - Vetmeduni

CY - Vienna, Austria

T2 - 10th European Conference on Precision Livestock Farming

Y2 - 29 August 2022 through 2 September 2022

ER -