Using 3D Imaging and Machine Learning to Predict Liveweight and Carcass Characteristics of Live Finishing Beef Cattle

ASAS-NANP SYMPOSIUM: Applications of machine learning for livestock body weight prediction from digital images

“…Nevertheless, only a handful of studies have applied ML and DL methods on data acquired with 3D imaging systems. Miller et al (2019) applied 3D imaging technology and ML algorithms to predict BW and carcass characteristics of live animals. Three-dimensional images and BWs were passively obtained from finishing steer and heifer beef cattle of a variety of preslaughter breeds either on-farm or after entering the abattoir.…”

“…While animal biometrics is an emerging field focused on quantification and detection of the phenotypic appearance of species, individuals, behaviors, and morphological traits ( Kühl and Burghardt, 2013 ), animal morphometrics ( Rohlf, 1990 ; Adams et al, 2004 ; Doyle et al, 2018 ) is almost exclusively focused on landmark-based methods (and less on outline-based methods) using quantitative analysis of form relying on measuring the size and shape of animals, and the relation between size and shape (allometry). Estimation of livestock BW using biometric and morphometric measurements has been studied in detail for various species, such as cattle ( Taşdemir et al, 2011a , b ; Miller et al, 2019 ; Tasdemir and Ozkan, 2019 ; Gjergji et al, 2020 ; de Moraes Weber et al, 2020 ; Rudenko, 2020 ), pigs ( Brandl and Jørgensen, 1996 ; O’Connell et al, 2007 ; Mutua et al, 2011 ; Sungirai et al, 2014 ; Al Ard Khanji et al, 2018 ), sheep ( Eyduran et al, 2015 ; Huma and Iqbal, 2019 ), goats ( Sebolai et al, 2012 ; Eyduran et al, 2017 ; Temoso et al, 2017 ), camels ( Fadlelmoula et al, 2020 ; de Moraes Weber et al, 2020 ), yaks ( Yan et al, 2019 ), poultry ( Mendeş and Akkartal, 2009 ), and fish ( Fernandes et al, 2020b ). This process is typically applied to avoid drawbacks associated with manually performed individual animal weighing such as: 1) the animal and manual laborer stress associated with animal relocation, 2) the costs associated with this labor-intensive process, and 3) the significant cost associated with acquiring and maintaining industrial scales.…”

“…However, since affordability of direct weighing methods may be an impediment for small producers ( Dickinson et al, 2013 ), researchers have developed indirect weighing methods represented by regression models that relate morphometric measurements and image features to BW in livestock. The direct acquisition of morphometric measurements can be accomplished with the aid of technologies with various degrees of complexity, from measuring tapes and tubes to specialized software or manual, semi-automatic or automatic measurements extrapolated from images obtained with electro-optical devices such as mono-2D ( Li et al, 2014 ), stereo-2D ( Rudenko, 2020 ), 3D ( Miller et al, 2019 ), ultrasound, and infrared sensors. A review and comparison of five different low-level indirect weighing techniques (Rondo tape, Weigh tape, Weighbridge equation, Schaeffer’s formula, and Agarwal’s formula) to estimate BW is provided in Wangchuk et al (2018) .…”

Advancements in sensor technology and decision support intelligent tools to assist smart livestock farming

“… Greenwood et al (2016) reviewed the use and application of various sensors, imaging, and other emerging technologies concerning extensive beef production, and González et al (2018) and Halachmi et al (2019) further discussed the attributes of these technologies for livestock production in general. The range of remote, near real-time monitoring technologies being developed or applied or with potential applications for free-ranging livestock and extensive grazing and foraging environments is increasing rapidly and include 1) in-field fixed and ground-, aerial-, and satellite-based measurement of pastures, invasive weeds, and soil, water, and greenhouse gas monitoring using sensors, photogrammetry ( Bloch et al, 2019 ), or other technologies including LiDAR ( Fernández-Quintanilla et al, 2018 ; Reinermann et al, 2020 ; Segarra et al, 2020 ; Weiss et al, 2020 ); 2) multi-channel, satellite-based spectrometry ( Segarra et al, 2020 ), such as WorldView-2 Satellite Sensor ( https://www.satimagingcorp.com/satellite-sensors/worldview-2/ ), which may be coupled with weather and soil grids to model and predict pasture biomass components and to guide grazing management decisions for sheep and cattle ( http://grazingapp.com.au/ ; Badgery et al, 2017 ); 3) body composition ( McPhee et al, 2017 ; Miller et al, 2019 ; Zhao et al, 2020 ) and physiological assessments ( Beiderman et al, 2014 ), including thermal imaging ( Halachmi et al, 2008 , 2013 ) to assess body temperature ( González et al, 2013 ) using devices at, or fixed to, handling facilities; 4) automated in-field liveweight measurement ( Nir et al, 2018 ) and drafting of livestock coupled with radio frequency identification ( RFID ) to determine individual or herd liveweight and growth of cattle ( Charmley et al, 2006 ; González et al, 2014 , 2018 ) and sheep ( Brown et al, 2015 ; González-García et al, 2018a , 2018b ); 5) virtual fencing using global positioning system ( GPS )-enabled collars and a mobile phone app ( https://www.agersens.com/ ) to remotely fence, move and monitor animals, and control herd or flock access to pastures and environmentally sensitive areas without the need for conventional fencing ( Campbell et al, 2019 , 2020 ); 6) on-animal devices to monitor location, activity, and behaviors in grazing and foraging environments ( Dobos et al, 2...…”

The use of 3-dimensional imaging of Holstein cows to estimate body weight and monitor the composition of body weight change throughout lactation

“…However, these approaches are seldom used on commercial farms (Le Cozler et al, 2012) because they are time-consuming and require the manipulation and restraint of animals, which can be risky and stressful for both the operator and the animal. To overcome these issues, recent studies have demonstrated the potential of estimating BW using 3-dimensional (3D) imaging (Kuzuhara et al, 2015;Gomes et al, 2016;Song et al, 2018;Le Cozler et al, 2019b;Miller et al, 2019;Martins et al, 2020). This technology also provides access to traits that have seldom been measured (e.g., surface areas or volumes of living animals) or that have not yet been studied (diagonal length; DL), which could be of interest for livestock breeders (Le Cozler et al, 2019b).…”