TY - JOUR
T1 - Evaluation of a depth sensor for mass estimation of growing and finishing pigs
AU - Condotta, Isabella C.F.S.
AU - Brown-Brandl, Tami M.
AU - Silva-Miranda, Késia O.
AU - Stinn, John P.
N1 - Funding Information:
Mention of trade names or commercial products in this article is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the U.S. Department of Agriculture. The USDA is an equal opportunity provider and employer. The authors would like to thank John Holman, Dale Janssen, and Hannah Speer for their help in collecting data, and Donna Griess for her help in preparing this document for publication. This research was funded in part by USDA, Agricultural Research Service; by Fundação de Amparo à Pesquisa do Estado de São Paulo - FAPESP (Grant No. 2015/07254-3 ) and by Conselho Nacional de Desenvolvimento Científico e Tecnológico CNPq (Grant No. 131182/2016-1 ), Brazil.
Publisher Copyright:
© 2018
PY - 2018/9
Y1 - 2018/9
N2 - A method of continuously monitoring animal mass would aid producers by ensuring all pigs are gaining mass and would increase the precision of marketing pigs. Therefore, the development of methods for monitoring the physical conditions of animals would improve animal well-being and maximise the profitability of swine production. The objective of this research was to validate the use of depth images in predicting live animal mass. Seven hundred and seventy-two depth images and mass measurements were collected from a population of grow–finish pigs (equally divided between barrows and gilts). Three commercial sire lines (Landrace, Duroc, and Yorkshire) were equally represented. The pigs’ volumes were calculated from the depth image. Linear equations were developed to predict mass from volume. Independent equations were developed for both gilts and barrows, each of the three commercial sire lines used, and a global equation for all combined data. Efroymson's algorithm was used to test for differences between the global equation and the two equations for the gilts and barrows and between the three commercial sire lines. The results showed that there was no significant difference between the global equation and the individual equations for barrows and gilts (p < 0.05), and the global equation was also no different from individual equations for each of the three sire lines (p < 0.05). The global equation was developed to predict mass from a depth sensor with an R2 of 0.9905. In conclusion, it appears that the depth sensor would be a reasonable approach to continuously monitor pig mass.
AB - A method of continuously monitoring animal mass would aid producers by ensuring all pigs are gaining mass and would increase the precision of marketing pigs. Therefore, the development of methods for monitoring the physical conditions of animals would improve animal well-being and maximise the profitability of swine production. The objective of this research was to validate the use of depth images in predicting live animal mass. Seven hundred and seventy-two depth images and mass measurements were collected from a population of grow–finish pigs (equally divided between barrows and gilts). Three commercial sire lines (Landrace, Duroc, and Yorkshire) were equally represented. The pigs’ volumes were calculated from the depth image. Linear equations were developed to predict mass from volume. Independent equations were developed for both gilts and barrows, each of the three commercial sire lines used, and a global equation for all combined data. Efroymson's algorithm was used to test for differences between the global equation and the two equations for the gilts and barrows and between the three commercial sire lines. The results showed that there was no significant difference between the global equation and the individual equations for barrows and gilts (p < 0.05), and the global equation was also no different from individual equations for each of the three sire lines (p < 0.05). The global equation was developed to predict mass from a depth sensor with an R2 of 0.9905. In conclusion, it appears that the depth sensor would be a reasonable approach to continuously monitor pig mass.
KW - Image analysis
KW - Kinect® sensor
KW - Precision livestock farming
KW - Swine
KW - Weight prediction
UR - http://www.scopus.com/inward/record.url?scp=85054084944&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85054084944&partnerID=8YFLogxK
U2 - 10.1016/j.biosystemseng.2018.03.002
DO - 10.1016/j.biosystemseng.2018.03.002
M3 - Article
AN - SCOPUS:85054084944
SN - 1537-5110
VL - 173
SP - 11
EP - 18
JO - Biosystems Engineering
JF - Biosystems Engineering
ER -