Automatic Weight Prediction System for Korean Cattle Using Bayesian Ridge Algorithm on RGB-D Image
Abstract
:1. Introduction
2. Related Works
3. Materials and Methods
3.1. Korean Cattle Image Collection
3.2. Proposed Method
3.2.1. Cattle Segmentation
- Case 1: If the two segmentation results match and they are assigned to the background, then the corresponding pixel is assigned to the background.
- Case 2: If the two segmentation results match and they are assigned to the foreground, then the corresponding pixel is assigned to the foreground.
- Case 3: If the two segmentation results do not match, the pixel is assigned as an object or background according to the segmentation result given by using the depth image.
3.2.2. Position Correction
3.2.3. Feature Extraction
3.2.4. Weight Prediction Models
- Step 1: Pick at random k data points from the training set.
- Step 2: Build a decision tree associated to these k data points.
- Step 3: Choose the number N of trees you want to build and repeat steps 1 and 2.
- Step 4: For a new data point, make each one of your N-tree trees predict the value of a response variable for the data point in question and assign the new data point to the average across all of the predicted value of response variable.
- Step 1: Each data input from the input layer is weighted through each node and the weight that connects the nodes.
- Step 2: It is transformed through the activation function and transferred to the input value (z) of the hidden layer. At this time, ‘Logistic sigmoid unit’ or ‘tanh unit’ is mainly used as the activation function existing in the hidden layer.
- Step 3: In the same way, it is weighted through the weight connected to the output layer, and finally converted and output through the activation function. At this time, the activation function is output as it is in the case of regression, and in the case of binary classification with two-classes of 0 and 1, sigmoid units and the softmax function are met in the case of multi-class.
3.2.5. Pseudocode for Our Proposed Method
Algorithm 1:Procedure Korean Cattle Weight Prediction: |
InputRGB-D image and weights. |
Fork = 1 to images do 1. Segmentation % Applying a threshold value to the depth image considering the height of Korean cattle. then ; ; % Korean cattle image segmentation using HSI color. then ; ; % Create a binarized image by AND-combining the segmented image using depth and color. ; 2. Feature Extraction Find contour for Calculate the central moment about the contour to find the center and rotation angle of Korean cattle. Apply Affine transformation to adjust the posture of Korean cattle. Crop cattle area. % Compute vertical projection profile to remove Korean cattle head. for j = 1 to width do ; for i = 0 to height do if C(i, j) > 0 then P(j) += 1; end if end for end for Calculate Body measurement parameters for cropped region. Calculate Size, Shape, Gradient-based descriptors. Creating feature vectors using body measures and descriptors. End for 3. Weight Prediction Split feature vectors and Korean cattle weights for training and testing. Training several regression models. |
OutputPredicted weight of Korean cattle using testing data. |
4. Experiments and Results
4.1. Materials and Interpretation
4.2. Experimental Results
5. Conclusions
Author Contributions
Funding
Institutional Review Board Statement
Conflicts of Interest
References
- Silvia, L.-C.; Rodolfo, R.V.A.; Gilberto, A.-O.A.; María, E.O.-C.B.; Juan, C.G.-O. Stress indicators in cattle in response to loading, transport and unloading practices. Rev. Mex. Cienc. Pecu. 2019, 10, 885–902. [Google Scholar] [CrossRef] [Green Version]
- Insaf, A.; Abdeldjalil, O.; Amir, B.; Abdelmalik, T.A. Past, Present, and Future of Face Recognition: A Review. Electronics 2020, 9, 1188. [Google Scholar] [CrossRef]
- Yacine, K.; Amir, B.; Abdeldjalil, O.; Sébastien, J.; Abdelmalik, T.-A. Ear Recognition Based on Deep Unsupervised Active Learning. IEEE Sens. J. 2021, 21, 20704–20713. [Google Scholar] [CrossRef]
- Bultakov, K.; Kim, Y.C.; Lee, H.M.; Na, I.S. A BPR-CNN Based Hand Motion Classifier Using Electric Field Sensors. CMC-Comput. Mater. Contin. 2022, 71, 5413–5425. [Google Scholar]
- Do, L.N.; Yang, H.J.; Nguyen, H.D.; Kim, S.H.; Lee, G.S.; Na, I.S. Deep neural network-based fusion model for emotion recognition using visual data. J. Supercomput. 2021, 77, 10773–10790. [Google Scholar] [CrossRef]
- Cho, W.H.; Kim, S.K.; Na, M.H.; Na, I.S. Fruit Ripeness Prediction Based on DNN Feature Induction from Sparse Dataset. CMC-Comput. Mater. Contin. 2021, 69, 4003–4024. [Google Scholar]
- Lee, K.O.; Lee, M.K.; Na, I.S. Predicting Regional Outbreaks of Hepatitis A Using 3D LSTM and Open Data in Korea. Electronics 2021, 10, 2668. [Google Scholar] [CrossRef]
- Cho, W.H.; Kim, S.K.; Na, M.H.; Na, I.S. Forecasting of Tomato Yields Using Attention-Based LSTM Network and ARMA Model. Electronics 2021, 10, 1576. [Google Scholar] [CrossRef]
- Chang, Y.; He, H.; Qiao, Y. An Intelligent Pig Weights Estimate Method Based on Deep Learning in Sow Stall Environments. IEEE Access 2019, 7, 164867–164875. [Google Scholar] [CrossRef]
- Femandes, A.F.A.; Dorea, J.R.R.; Rosa, G.J.M. Image Analysis and Computer Vision Applications in Animal Sciences: An Overview. Front. Vet. Sci. 2020, 7, 551269. [Google Scholar] [CrossRef]
- Nasirahmadi, A.; Edwards, S.A.; Sturm, B. Implementation of machine vision for detecting behavior of cattle and pigs. Livest. Sci. 2017, 202, 25–38. [Google Scholar] [CrossRef] [Green Version]
- Wang, Z.; Shadpour, S.; Chan, E.; Rotobdo, V.; Wood, K.M.; Tulpan, D. ASAS-NANP SYMPOSIUM: Application of machine learning for livestock body weight prediction from digital images. J. Anim. Sci. 2021, 99, skab022. [Google Scholar] [CrossRef] [PubMed]
- Seo, K.-W.; Lee, D.-W.; Choi, E.-G.; Kim, C.-H.; Kim, H.-T. Algorithm for Measurement of the Dairy Cow’s Body Parameters by Using Image Processing. J. Biosyst. Eng. 2012, 37, 122–129. [Google Scholar] [CrossRef] [Green Version]
- Tasdemir, S.; Ozkan, İ.A. Ann Approach for Estimation of Cow Weight Depending on Photogrammetric Body Dimensions. Int. J. Eng. Geosci. 2019, 4, 36–44. [Google Scholar] [CrossRef]
- Huang, L.; Guo, H.; Rao, Q.; Hou, Z.; Li, S.; Qiu, S.; Fan, X.; Wang, H. Body Dimension Measurements of Qinchuan Cattle with Transfer Learning from LiDAR Sensing. Sensors 2019, 19, 5046. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Rudenko, O.; Megel, Y.; Bezsonov, O.; Rybalka, A. Cattle breed identification and live weight evaluation on the basis of machine learning and computer vision. In Proceedings of the Third International Workshop on Computer Modeling and Intelligent Systems (CMIS-2020), CMIS-2020 Computer Modeling and Intelligent Systems, Zaporizhzhia, Ukraine, 27 April–1 May 2020; pp. 939–954. [Google Scholar]
- Pradana, Z.H.; Hidayat, B.; Darana, S. Beef cattle weight determine by using digital image processing. In Proceedings of the 2016 International Conference on Control, Electronics, Renewable Energy and Communications (ICCEREC), Bandung, Indonesia, 13–15 September 2016; pp. 179–184. [Google Scholar] [CrossRef]
- Weber, V.A.M.; Weber, F.L.; Gomes, R.C.; Oliveira Junor, A.S.; Menezes, G.V.; Abreu, U.G.P.; Belete, N.A.S.; Pistori, H. Prediction of Girolando cattle weight by means of body measurements extracted from images. Rev. Bras. Zootec. 2020, 49, e20190110. [Google Scholar] [CrossRef] [Green Version]
- Gjergji, M.; de Weber, V.M.; Silva, L.O.C.; da Gomes, R.C.; de Araujo, T.L.A.C.; Pistori, H.; Alvarez, M. Deep Learning Techniques for Beef Cattle Body Weight Prediction. In Proceedings of the International Joint Conference on Neural Networks (IJCNN), Glasgow, UK, 19–24 July 2020; pp. 1–8. [Google Scholar]
- Rodríguez, A.J.; Arroqui, M.; Mangudo, P.; Toloza, J.; Jatip, D.; Rodriguez, J.M.; Teyseyre, A.; Sanz, C.; Zunino, A.; Machado, C.; et al. Estimating Body Condition Score in Dairy Cows from Depth Images Using Convolutional Neural Networks, Transfer Learning and Model Ensembling Techniques. Agronomy 2019, 9, 90. [Google Scholar] [CrossRef] [Green Version]
- Lee, D.H.; Lee, S.H.; Cho, B.K.; Wakholi, C.; Seo, Y.W.; Cho, S.H.; Kang, T.H.; Lee, W.H. Estimation of carcass weight of Hanwoo (Korean native cattle) as a function of body measurements using statistical models and a neural network. Asian-Australas. J. Anim. Sci. 2020, 33, 1633–1641. [Google Scholar] [CrossRef] [Green Version]
- Jang, D.H.; Kim, C.; Ko, Y.-G.; Kim, Y.H. Estimation of Body Weight for Korean Cattle Using Three-Dimensional Image. J. Biosyst. Eng. 2020, 45, 325–332. [Google Scholar] [CrossRef]
- Lee, S.U.; Chung, S.Y.; Park, R.H. A comparative performance study of several global thresholding techniques for segmentation. Comput. Vis. Graph. Image Process. 1990, 52, 171–190. [Google Scholar] [CrossRef]
- Liu, Z.; Zhao, C.C.; Wu, X.M.; Chen, W.H. An Effective 3D Shape Descriptor for Object Recognition with RGB-D Sensors. Sensors 2017, 17, 451. [Google Scholar] [CrossRef] [PubMed] [Green Version]
Method | Coefficient of Determination (R2) | ||
---|---|---|---|
Total (Male + Female) | Male | Female | |
Linear Regression | 0.955500 | 0.718056 | 0.658056 |
SVM Regression | 0.947167 | 0.655915 | 0.735438 |
Ridge | 0.924378 | 0.559549 | 0.402620 |
LASSO | 0.905019 | 0.485137 | 0.070050 |
Bayesian Ridge | 0.955736 | 0.697338 | 0.717989 |
MLP | 0.941887 | 0.667934 | 0.705792 |
Decision Tree | 0.936354 | 0.464219 | 0.279362 |
Random Forest | 0.968880 | 0.755998 | 0.732490 |
Method | Mean Square Error | ||
---|---|---|---|
Total (Male + Female) | Male | Female | |
Linear Regression | 1051.7 | 1108.5 | 300.6 |
SVM Regression | 1248.5 | 1352.8 | 232.6 |
Ridge | 1787.1 | 1731.7 | 525.2 |
LASSO | 2244.5 | 2024.3 | 817.6 |
Bayesian Ridge | 1046.0 | 1190.0 | 247.9 |
MLP | 1373.3 | 1305.6 | 258.6 |
Decision Tree | 1504.0 | 1973.9 | 663.4 |
Random Forest | 735.4 | 959.3 | 235.2 |
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |
© 2022 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
Na, M.H.; Cho, W.H.; Kim, S.K.; Na, I.S. Automatic Weight Prediction System for Korean Cattle Using Bayesian Ridge Algorithm on RGB-D Image. Electronics 2022, 11, 1663. https://doi.org/10.3390/electronics11101663
Na MH, Cho WH, Kim SK, Na IS. Automatic Weight Prediction System for Korean Cattle Using Bayesian Ridge Algorithm on RGB-D Image. Electronics. 2022; 11(10):1663. https://doi.org/10.3390/electronics11101663
Chicago/Turabian StyleNa, Myung Hwan, Wan Hyun Cho, Sang Kyoon Kim, and In Seop Na. 2022. "Automatic Weight Prediction System for Korean Cattle Using Bayesian Ridge Algorithm on RGB-D Image" Electronics 11, no. 10: 1663. https://doi.org/10.3390/electronics11101663
APA StyleNa, M. H., Cho, W. H., Kim, S. K., & Na, I. S. (2022). Automatic Weight Prediction System for Korean Cattle Using Bayesian Ridge Algorithm on RGB-D Image. Electronics, 11(10), 1663. https://doi.org/10.3390/electronics11101663