Recording behaviour of indoor-housed farm animals automatically using machine vision technology: A systematic review

Abstract

Large-scale phenotyping of animal behaviour traits is time consuming and has led to increased demand for technologies that can automate these procedures. Automated tracking of animals has been successful in controlled laboratory settings, but recording from animals in large groups in highly variable farm settings presents challenges. The aim of this review is to provide a systematic overview of the advances that have occurred in automated, high throughput image detection of farm animal behavioural traits with welfare and production implications. Peer-reviewed publications written in English were reviewed systematically following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. After identification, screening, and assessment for eligibility, 108 publications met these specifications and were included for qualitative synthesis. Data collected from the papers included camera specifications, housing conditions, group size, algorithm details, procedures, and results. Most studies utilized standard digital colour video cameras for data collection, with increasing use of 3D cameras in papers published after 2013. Papers including pigs (across production stages) were the most common (n = 63). The most common behaviours recorded included activity level, area occupancy, aggression, gait scores, resource use, and posture. Our review revealed many overlaps in methods applied to analysing behaviour, and most studies started from scratch instead of building upon previous work. Training and validation sample sizes were generally small (mean±s.d. groups = 3.8±5.8) and in data collection and testing took place in relatively controlled environments. To advance our ability to automatically phenotype behaviour, future research should build upon existing knowledge and validate technology under commercial settings and publications should explicitly describe recording conditions in detail to allow studies to be reproduced.

Large-Scale Phenotyping of Livestock Welfare in Commercial Production Systems: A New Frontier in Animal Breeding

“…For instance, facial recognition to identify individual animals is currently being investigated (Hansen et al, 2018). Wurtz et al (2019) performed a comprehensive systematic review of studies that used machine vision technology to assess behavior of indoor-housed farm animals. The authors highlighted the need to build upon existing knowledge, instead of developing devices from scratch, and validate these devices under commercial settings (in large scale).…”

Digitalization of Animal Farming

“…To monitor and address concerns over aggression, researchers are investigating the use of automated video monitoring and depth imaging tracking. These technologies are generally able to monitor overall activity patterns but cannot yet track individual behavioral patterns (Wurtz et al, 2019). Infrared thermography has been used to identify illness in piglets (Benjamin and Yik, 2019).…”

“…Automated video detection software is largely nonfunctional within animal agriculture at the moment (Wurtz et al, 2019). Image analysis of aggression in pigs currently struggles to distinguish between different behaviors, such as play and aggression.…”

“…These technologies also cannot yet track individual animals, at least not for a long enough period of time to obtain meaningful information about behaviors of interest. Some technologies may be able to track individuals when they are up and moving but cannot track individuals when they lie in a pile and then get up again (Wurtz et al, 2019). There are also issues with distinguishing animals from the background of the environment; many video technologies were developed in specific test arenas where there was good contrast between the pen structures and the animals, so the technology can fail when applied in real-life farm situations (Wurtz et al, 2019).…”

“…Some technologies may be able to track individuals when they are up and moving but cannot track individuals when they lie in a pile and then get up again (Wurtz et al, 2019). There are also issues with distinguishing animals from the background of the environment; many video technologies were developed in specific test arenas where there was good contrast between the pen structures and the animals, so the technology can fail when applied in real-life farm situations (Wurtz et al, 2019). Additionally, many of the studies testing these technologies have been done on pigs, so more work is needed to assess the applicability on other species (Wurtz et al, 2019).…”