The body weight of pigs can be determined automatically using 3D camera technology
Researchers from Aarhus University (AU) have validated a commercially available 3D camera system that can automatically monitor the body weight of slaughter pigs - and the results were satisfactory.
In pig production, it is crucial to be able to follow the pigs’ growth. Knowledge of growth can be used to predict the optimal point in time to deliver the pigs for slaughter. In addition, knowledge of the pigs' growth and deviations from normal growth can be helpful in relation to identifying problems, e.g., poor feeding management or presence of sick animals. Monitoring body weight can thus contribute to ensuring the pigs' welfare and health and optimisation of production efficiency.
Typically, in commercial farms, a smaller part of the pigs is weighed manually. This is regarded as the most precise method of determining their body weight. However, pig herds in the EU have grown in size in recent decades, and manual weighing is thus very demanding and potentially stressful for both humans and animals. Therefore, an automatic estimation of the body weight of pigs will be a useful tool for ongoing monitoring of the growth in commercial pig production.
Study of the 3D camera system in two pig herds
As part of a major EU-funded project (ClearFarm), researchers from the Department of Animal and Veterinary Sciences (ANIVET), AU, with researchers from Wageningen University and Research, in the Netherlands, have recently completed an external validation of a commercially available 3D camera (iDOL65, dol-sensors A/S, Denmark) that can estimate the weight of pigs based on their 3D images. The researchers have examined how accurate the 3D camera estimated bodyweight compared with manual weighing (gold standard). The study was carried out partly in AU’s experimental stables with slaughter pigs (20-114 kg), where the 144 pigs were weighed weekly, and in a commercial herd in Germany with slaughter pigs (16-130 kg), where the 107 pigs were weighed three times throughout the fattening period.
In the German herd, the cameras were placed over electronic feed stations, where the pigs were fed individually. In AU’s experimental herd, the cameras were placed over the feed dispensers allowing up to three pigs to eat simultaneously. In both places, pigs had electronic ear tags, which via a so-called RFID reader could recognise the pigs under the camera. "Every time a pig visited the feeding area, a 3D image was taken. All images were used to estimate daily body weight of the individual pig and simultaneously estimated an average weight for all pigs in the pen," explains researcher Guilherme A. Franchi from ANIVET, who has been responsible for the study.
Satisfactory results
In the study, the researchers found a very high agreement between the manual weight and the estimated body weight. The overall errors of body weight estimation on the median weight during the fattening period were low (≤ 3.6%) at individual pig and pen level, and the researchers conclude that the 3D camera used (iDOL65) can provide satisfactory body weight information at both pig and pen levels. The agreement was high across the two herds, where there were differences in feeding system, breed, and age of the pigs. "As the camera-based weighing system is relatively simple to install and does not require physical handling of pigs, we see great potential in the use of the iDOL65 in commercial fattening pig farms" says Guilherme A. Franchi.
Moving towards farmers' adoption of iDOL65
"An opportunity to stimulate farmers to take the new weighing technology into use could be to combine the camera system with equipment that makes it possible to remotely control the camera across the pen or even across sections to take images of pigs continuously and at different pen locations. This improvement would make it possible to estimate the body weight of individual pigs without the use of RFID reading of the electronic ear tags; a technology which is not yet commercially available" concludes Guilherme A. Franchi.
To further explore the potential of this system in the future, researchers will focus on developing algorithms that can detect deviations in growth allowing for timely interventions. It may help reduce welfare and health problems and increase productivity, thereby reducing the climate and environmental impact of pig production.
Additional information | |
We strive to ensure that all our articles live up to the Universities Denmark’s principles for good research communication (scroll down to find the English version on the website). Because of this, the article will be supplemented with the following information: | |
Funding | This study was part of the EU project ClearFarm and funded by the European Union's Horizon 2020 research and innovation program under grant agreement no. 862919. |
Collaborators | This study was performed by Guilherme Amorim Franchi, Margit Bak Jensen and Lene Juul Pedersen (all AU) in collaboration with Jacinta Bus, Iris Boumans and Eddie Bokkers (Wageningen University and Research, The Netherlands). We also thank dol-sensors a/s for making the weight estimates from both farms available. |
External commenting |
|
Conflicts of interest | None |
Read more | |
Contact | Postdoc Guilherme A. Franchi Dept. Animal and Veterinary Sciences |