Use of Biometric Images to Predict Body Weight and Hot Carcass Weight of Nellore Cattle
Abstract
:Simple Summary
Abstract
1. Introduction
2. Materials and Methods
2.1. Dataset and Obtaining Biometric Measurements by Images
2.1.1. Samples Used
2.1.2. Collection of Weights and Biometric Measurements by Images
2.2. Statistical Methods for the Prediciton of Body Weight and Hot Carcass
2.2.1. Correlation and Exclusion of Variables
2.2.2. Formation of Training and Test Sets
2.2.3. Models Used for Prediction
2.3. Evaluation of Weight Predictions
3. Results and Discussion
4. Conclusions
Supplementary Materials
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Conflicts of Interest
References
- Polgar, P.; Szücs, E.; Szabo, F. Effect of selection for improvement of dairy performance on growth and body structure in Holstein Friesian and Hungarian Fleckvieh young bulls. Arch. Tierz. 1997, 40, 505–510. (In German) [Google Scholar]
- Ulutas, Z.; Saatci, M.; Ozluturk, A. Prediction of body weights from body measurements in East Anatolian Red calves. J. Atatrk Univ. Agric. Fac. 2001, 32, 61–65. [Google Scholar]
- Schofield, C.P.; Marchant, J.A.; White, R.P.; Brandl, N.; Wilson, M. Monitoring pig growth using a prototype imaging system. J. Agric. Eng. Res. 1999, 72, 205–210. [Google Scholar] [CrossRef]
- Kuzuhara, Y.; Kawamura, K.; Yoshitoshi, R.; Tamaki, T.; Sugai, S.; Ikegami, M.; Yasuda, T. A preliminarily study for predicting body weight and milk properties in lactating Holstein cows using a three-dimensional camera system. Comput. Electron. Agric. 2015, 111, 186–193. [Google Scholar] [CrossRef]
- Wang, Z.; Shadpour, S.; Chan, E.; Rotondo, V.; Wood, K.M.; Tulpan, D. ASASNANP Symposium: Applications of machine learning for livestock body weight prediction from digital images. J. Anim. Sci. 2021, 99, 22. [Google Scholar] [CrossRef] [PubMed]
- Gomes, R.A.; Monteiro, G.R.; Assis, G.J.F.; Busato, K.C.; Ladeira, M.M.; Chizzotti, M.L. Technical note: Estimating body weight and body composition of beef cattle trough digital image analysis. J. Anim. Sci. 2016, 94, 5414–5422. [Google Scholar] [CrossRef] [PubMed]
- Liu, D.; He, D.; Norton, T. Automatic estimation of dairy cattle body condition score from depth image using ensemble model. Biosyst. Eng. 2020, 194, 16–27. [Google Scholar] [CrossRef]
- Fernandes, A.F.; Dórea, J.R.; Fitzgerald, R.; Herring, W.; Rosa, G.J. A novel automated system to acquire biometric and morphological measurements and predict body weight of pigs via 3D computer vision. J. Anim. Sci. 2019, 97, 496–508. [Google Scholar] [CrossRef] [PubMed]
- Cominotte, A.; Fernandes, A.F.A.; Dorea, J.R.R.; Rosa, G.J.M.; Ladeira, M.M.; van Cleef, E.H.C.B.; Neto, O.M. Automated computer vision system to predict body weight and average daily gain in beef cattle during growing and finishing phases. Livest. Sci. 2020, 232, 103904. [Google Scholar] [CrossRef]
- Carrascal, L.M.; Galván, I.; Gordo, O. Partial least squares regression as an alternative to current regression methods used in ecology. Oikos 2009, 5, 681–690. [Google Scholar] [CrossRef]
- Gianola, D.; Okut, H.; Weigel, K.A.; Rosa, G.J.M. Predicting complex quantitative traits with Bayesian neural networks: A case study with Jersey cows and wheat. BMC Genet. 2011, 12, 87. [Google Scholar] [CrossRef] [PubMed]
- Hough, P.V.C. Method and Means for Recognizing Complex Patterns. U.S. Patent 4746348, 18 December 1962. [Google Scholar]
- Atherton, T.J.; Kerbyson, D.J. Size in variant circle detection. Image Vis. Comput. 1999, 17, 795–803. [Google Scholar] [CrossRef]
- Bowman, E.T.; Soga, K.; Drummond, T.W. Particleshape Characterisation Using Fourier Analysis; CUED/DSoils/TR315; University of Cambridge, Department of Engineering: Cambridge, UK, 2000; Volume 1, p. 20. [Google Scholar]
- Zhang, D.; Lu, G. Shape-based image retrieval using generic Fourier descriptor. Signal process. Image Commun. 2002, 17, 825–848. [Google Scholar]
- Burger, W.; Burge, M.J. Digital Image Processing: An Algorithmic Introduction Using Java, 2nd ed.; Gries, D., Schneider, F.B., Eds.; Springer: London, UK, 2016. [Google Scholar] [CrossRef]
- Friedman, J.; Hastie, T.; Tibshirani, R. Regularization paths for generalized linear models via coordinate descent. J. Stat. Softw. 2010, 33, 1–22. [Google Scholar] [CrossRef] [PubMed]
- Dórea, J.R.R.; Rosa, G.J.M.; Weld, K.A.; Armentano, L.E. Mining data from milk infrared spectroscopy to improve feed intake predictions in lactating dairy cows. J. Dairy Sci. 2018, 101, 5878–5889. [Google Scholar] [CrossRef] [PubMed]
- Zeiler, M.D. ADADELTA: An adaptive learning rate method. arXiv 2012, arXiv:1212.5701:1:6. [Google Scholar]
- Tedeschi, L.O. Assessment of the adequacy of mathematical models. Agric. Syst. 2006, 89, 225–247. [Google Scholar] [CrossRef]
- Cochran, W.G.; Cox, G.M. Experimental Design, 2nd ed.; John Wiley and Sons: New York, NY, USA, 1957. [Google Scholar]
- Lin, L.I.K. A concordance correlation coefficient to evaluate reproducibility. Biometrics 1898, 45, 255–268. [Google Scholar] [CrossRef]
- Bibby, J.; Toutenburg, D.H. Prediction and Improved Estimation in Linear Models; John Wiley & Sons: Berlin, Germany, 1977. [Google Scholar]
- Theil, H. Economic Forecasts and Policy; Springer: Berlin/Heidelberg, Germany, 1961. [Google Scholar]
- Shetty, N.; Lovenhahl, P.; Lund, M.; Buitenhuis, A. Prediction and validation of residual feed intake and dry matter intake in Danish lactating dairy cows using mid infrared spectroscopy of milk. J. Dairy Sci. 2017, 100, 253–264. [Google Scholar] [CrossRef]
Training | Validation (Test) | ||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
n | Mean (kg) | SD 1 (kg) | CV 2 (%) | Min (kg) | Max (kg) | n | Age 3 | Mean (kg) | SD 1 (kg) | CV 2 (%) | Min (kg) | Max (kg) | |||
BW | Set 1 | 359 | 553.30 | 63.08 | 11.40 | 359.00 | 665.00 | Exp. 4 | 91 | 24 ± 2 | 589.94 | 30.26 | 5.12 | 505.00 | 653.00 |
Set 2 | 222 | 532.18 | 65.90 | 12.38 | 359.00 | 653.00 | Exp. 3 | 228 | 22 ± 2 | 558.49 | 35.61 | 6.05 | 485.00 | 665.00 | |
Set 3 | 367 | 570.67 | 59.30 | 10.39 | 359.00 | 665.00 | Exp. 2 | 83 | 22 ± 2 | 516.69 | 38.22 | 7.39 | 450.00 | 612.00 | |
Set 4 | 402 | 573.99 | 45.59 | 7.94 | 450.00 | 665.00 | Exp. 1 | 48 | 22 ± 2 | 449.48 | 47.46 | 10.55 | 359.00 | 554.00 | |
HCW | Set 1 | 359 | 315.34 | 37.32 | 11.83 | 205.00 | 394.00 | Exp. 4 | 91 | 24 ± 2 | 334.22 | 16.79 | 5.02 | 277.60 | 376.80 |
Set 2 | 222 | 301.70 | 36.42 | 12.07 | 205.00 | 376.80 | Exp. 3 | 228 | 22 ± 2 | 336.17 | 23.17 | 6.92 | 259.20 | 394.40 | |
Set 3 | 367 | 325.74 | 34.02 | 10.44 | 204.80 | 394.40 | Exp. 2 | 83 | 22 ± 2 | 290.08 | 22.19 | 7.65 | 248.80 | 343.60 | |
Set 4 | 402 | 326.21 | 28.49 | 8.73 | 248.80 | 394.40 | Exp. 1 | 48 | 22 ± 2 | 260.12 | 28.10 | 10.80 | 205.00 | 317.00 |
Training Dataset/Test | Models | Slope (±SE) 1 | H0: b = 1 | r2 | Mean Bias (kg) | CCC 2 | RMSEP (kg) 3 | RMSEP/Mean (%) | Decomposition of MSEP (%) | ||
---|---|---|---|---|---|---|---|---|---|---|---|
Mean Bias | Slope Bias | Random Error | |||||||||
Set 1/ Exp. 4 | ANN | 0.59 (±0.05) | 0.01 | 0.58 | 3.22 | 0.73 (±0.04) | 19.68 | 3.34 | 2.68 | 37.90 | 59.42 |
PLS | 0.69 (±0.06) | 0.01 | 0.55 | 5.33 | 0.72 (±0.04) | 21.60 | 3.66 | 6.09 | 18.11 | 75.80 | |
LASSO | 0.70 (±0.06) | 0.02 | 0.58 | 6.52 | 0.74 (±0.04) | 20.95 | 3.55 | 9.71 | 18.55 | 71.74 | |
MLR | 0.69 (±0.06) | 0.01 | 0.58 | 6.62 | 0.74 (±0.04) | 20.95 | 3.55 | 10.00 | 18.95 | 71.05 | |
Set 2/ Exp. 3 | ANN | 0.57 (±0.03) | 0.01 | 0.53 | −12.19 | 0.66 (±0.03) | 27.22 | 5.11 | 20.05 | 30.33 | 49.62 |
PLS | 0.59 (±0.03) | 0.01 | 0.53 | −13.80 | 0.65 (±0.03) | 28.07 | 5.03 | 24.18 | 26.17 | 49.65 | |
LASSO | 0.60 (±0.03) | 0.01 | 0.53 | −14.12 | 0.65 (±0.03) | 28.31 | 5.07 | 24.90 | 24.70 | 50.40 | |
MLR | 0.59 (±0.03) | 0.01 | 0.53 | −13.55 | 0.66 (±0.03) | 27.95 | 5.00 | 23.53 | 25.86 | 50.61 | |
Set 3/ Exp. 2 | ANN | 0.62 (±0.06) | 0.01 | 0.53 | 6.04 | 0.70 (±0.05) | 27.23 | 5.27 | 4.91 | 27.11 | 67.98 |
PLS | 0.56 (±0.09) | 0.02 | 0.30 | −0.65 | 0.54 (±0.07) | 36.87 | 7.14 | 0.03 | 20.22 | 79.75 | |
LASSO | 0.58 (±0.08) | 0.01 | 0.39 | 5.81 | 0.61 (±0.06) | 32.34 | 6.26 | 3.23 | 24.05 | 72.72 | |
MLR | 0.56 (±0.07) | 0.01 | 0.39 | 1.96 | 0.62 (±0.06) | 31.36 | 6.07 | 0.40 | 27.65 | 71.95 | |
Set 4/ Exp. 1 | ANN | 0.74 (±0.09) | 0.08 | 0.59 | 12.40 | 0.74 (±0.06) | 33.75 | 5.87 | 12.51 | 12.14 | 74.35 |
PLS | 0.78 (±0.11) | 0.06 | 0.51 | 11.88 | 0.69 (±0.07) | 39.34 | 8.75 | 9.12 | 6.35 | 84.53 | |
LASSO | 0.76 (±0.11) | 0.04 | 0.50 | 16.77 | 0.67 (±0.07) | 40.63 | 9.04 | 17.04 | 7.33 | 75.63 | |
MLR | 0.72 (±0.06) | 0.04 | 0.51 | 13.77 | 0.67 (±0.07) | 42.65 | 9.49 | 7.35 | 17.06 | 75.59 |
Training Dataset/Test | Models | Slope (±SE) 1 | H0:b = 1 | r2 | Mean Bias (kg) | CCC 2 | RMSEP (kg) 3 | RMSEP/ Mean (%) | Decomposition of MSEP (%) | ||
---|---|---|---|---|---|---|---|---|---|---|---|
Mean Bias | Slope Bias | Random Error | |||||||||
Set 1/ Exp. 4 | ANN | 0.56 (±0.08) | 0.01 | 0.45 | 3.16 | 0.65 (±0.05) | 13.06 | 3.90 | 5.87 | 30.78 | 63.35 |
PLS | 0.59 (±0.06) | 0.01 | 0.44 | 6.89 | 0.60 (±0.06) | 14.62 | 4.37 | 22.23 | 21.77 | 56.00 | |
LASSO | 0.63 (±0.06) | 0.01 | 0.49 | 4.43 | 0.67 (±0.05) | 13.17 | 3.94 | 11.31 | 21.37 | 67.32 | |
MLR | 0.08 (±0.03) | 0.01 | 0.05 | 0.01 | 0.14 (±0.06) | 8.75 | 2.62 | 0.00 | 86.95 | 13.05 | |
Set 2/ Exp. 3 | ANN | 0.19 (±0.03) | 0.01 | 0.17 | −21.27 | 0.19 (±0.03) | 30.00 | 8.92 | 50.25 | 38.55 | 11.20 |
PLS | 0.44 (±0.03) | 0.01 | 0.42 | −10.06 | 0.54 (±0.03) | 20.33 | 6.05 | 24.49 | 39.51 | 36.00 | |
LASSO | 0.43 (±0.03) | 0.01 | 0.40 | −9.11 | 0.53 (±0.03) | 20.13 | 5.99 | 20.50 | 42.12 | 37.38 | |
MLR | 0.43 (±0.03) | 0.01 | 0.40 | −9.51 | 0.53 (±0.03) | 20.29 | 6.04 | 21.96 | 41.55 | 36.49 | |
Set 3/ Exp. 2 | ANN | 0.52 (±0.06) | 0.01 | 0.44 | 11.33 | 0.56 (±0.06) | 20.06 | 6.91 | 31.87 | 26.99 | 41.14 |
PLS | 0.57 (±0.09) | 0.02 | 0.32 | 10.49 | 0.51 (±0.07) | 23.01 | 7.93 | 20.77 | 16.57 | 62.66 | |
LASSO | 0.55 (±0.06) | 0.01 | 0.46 | 12.89 | 0.55 (±0.06) | 20.89 | 7.20 | 38.10 | 22.31 | 39.59 | |
MLR | 0.42 (±0.09) | 0.01 | 0.20 | 9.82 | 0.40 (±0.08) | 24.61 | 8.48 | 16.25 | 26.83 | 56.92 | |
Set 4/ Exp. 1 | ANN | 0.61 (±0.08) | 0.05 | 0.52 | 12.67 | 0.64 (±0.07) | 23.20 | 8.91 | 29.83 | 21.14 | 49.03 |
PLS | 0.78 (±0.12) | 0.09 | 0.45 | 5.18 | 0.65 (±0.08) | 25.35 | 9.75 | 4.19 | 5.55 | 90.26 | |
LASSO | 0.75 (±0.11) | 0.03 | 0.48 | 8.53 | 0.66 (±0.07) | 24.16 | 9.29 | 12.48 | 8.13 | 79.39 | |
MLR | 0.75 (±0.10) | 0.02 | 0.52 | 8.49 | 0.69 (±0.07) | 22.76 | 8.75 | 13.93 | 8.91 | 77.16 |
Root Mean Square Error Prediction (RMSEP) | |||||||||
---|---|---|---|---|---|---|---|---|---|
Body Weight (kg) | Hot Carcass Weight (kg) | ||||||||
Set 1 | Set 2 | Set 3 | Set 4 | Set 1 | Set 2 | Set 3 | Set 4 | ||
Training set 10-kfold cross validation | ANN | 20.16 | 20.91 | 18.69 | 20.27 | 15.71 | 13.42 | 15.25 | 14.82 |
PLS | 21.28 | 20.72 | 20.53 | 18.23 | 16.77 | 13.77 | 15.98 | 14.70 | |
LASSO | 21.54 | 20.33 | 20.43 | 18.06 | 16.67 | 13.94 | 15.67 | 14.82 | |
RLM | 23.62 | 21.83 | 22.21 | 21.00 | 7.33 | 13.95 | 15.04 | 15.08 | |
Independent validation | ANN | 19.68 | 27.22 | 27.23 | 33.75 | 13.06 | 30.00 | 20.06 | 23.20 |
PLS | 21.60 | 28.07 | 36.87 | 39.34 | 14.62 | 20.33 | 23.01 | 25.35 | |
LASSO | 20.95 | 28.31 | 32.34 | 40.63 | 13.17 | 20.13 | 20.89 | 24.16 | |
RLM | 20.95 | 27.95 | 31.36 | 40.63 | 8.75 | 20.29 | 24.61 | 22.76 |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Cominotte, A.; Fernandes, A.; Dórea, J.; Rosa, G.; Torres, R.; Pereira, G.; Baldassini, W.; Machado Neto, O. Use of Biometric Images to Predict Body Weight and Hot Carcass Weight of Nellore Cattle. Animals 2023, 13, 1679. https://doi.org/10.3390/ani13101679
Cominotte A, Fernandes A, Dórea J, Rosa G, Torres R, Pereira G, Baldassini W, Machado Neto O. Use of Biometric Images to Predict Body Weight and Hot Carcass Weight of Nellore Cattle. Animals. 2023; 13(10):1679. https://doi.org/10.3390/ani13101679
Chicago/Turabian StyleCominotte, Alexandre, Arthur Fernandes, João Dórea, Guilherme Rosa, Rodrigo Torres, Guilherme Pereira, Welder Baldassini, and Otávio Machado Neto. 2023. "Use of Biometric Images to Predict Body Weight and Hot Carcass Weight of Nellore Cattle" Animals 13, no. 10: 1679. https://doi.org/10.3390/ani13101679
APA StyleCominotte, A., Fernandes, A., Dórea, J., Rosa, G., Torres, R., Pereira, G., Baldassini, W., & Machado Neto, O. (2023). Use of Biometric Images to Predict Body Weight and Hot Carcass Weight of Nellore Cattle. Animals, 13(10), 1679. https://doi.org/10.3390/ani13101679