# An Automatic Head Surface Temperature Extraction Method for Top-View Thermal Image with Individual Broiler

^{1}

^{2}

^{3}

^{4}

^{*}

## Abstract

**:**

^{®}(R2016a) and the testing results indicated that the head region in 92.77% of the broiler thermal images could be located correctly. The maximum error of the extracted head surface temperatures was not greater than 0.1 °C. Different trend features were observed in the smoothed head surface temperature time series of the broilers in experimental and control groups. Head surface temperature extracted by the presented algorithm lays a foundation for the development of an automatic system for febrile broiler identification.

## 1. Introduction

## 2. Materials and Methods

#### 2.1. Image System

#### 2.2. Broilers and Thermal Images Acquisition

#### 2.3. Head Surface Temperature Extraction

#### 2.3.1. Thermal Image Pre-Processing

#### 2.3.2. Head Region Locating

_{1}and x

_{2}, y

_{1}, and y

_{2}were the horizontal and vertical coordinates of the two endpoints, respectively. The extracted endpoints are shown in Figure 4b with two star points.

_{p}was the intensity of the pixel in row (p mod 240), column (p/240 + 1) of the grayscale image. “/” and “mod” were modulus and remainder operators on integer values, respectively. c

_{q}was the intensity of the qth centroid. Operator |w| referred to the operation of extracting the absolute value of w.

**Case 1**: EP1 and EP2 located inside L1 and L2, one endpoint in one region. Then, the alternative head region with higher maximum temperature was extracted as the head region.

**Case 2**: Only one endpoint located inside L1 or L2. Then, the alternative head region, in which the endpoint was located, was extracted as the head region.

**Case 3**: No endpoint located inside L1 or L2. Then, the shorter distance of each endpoint to the center of L1 and L2, denoted by D_min1 and D_min2, respectively, were calculated and obtained. The alternative head region with a distance of min{D_min1, D_min2} s extracted as the head region. Where, min{D_min1, D_min2} was the smaller value of D_min1 and D_min2.

_{i}

_{,j,k}(i ∈ [1, 20]). Where i was the number of the broiler and j was the number of the thermal image capture interval. For a broiler in the experimental group, the maximum value of j depended on its dead time. For broilers in the control group, the maximum value of j was 30. k was the number of the thermal image in each interval. It had a maximum value of 5.

_{j}is the quantity of the thermal images in the jth interval.

#### 2.3.3. Construction of RHT Time Series

_{i}is the quantity of the intervals in the thermal images acquisition step for the ith broiler. Two broilers, one in the experimental group and another in the control group, were randomly selected, the corresponding TSRHT of which are shown in Figure 8.

_{i}(i ∈ [1, 20]) was obtained by calculating the mean of the first t (t ∈ [1, NIH

_{i}]) elements in TSRHT

_{i}by using Equation (10):

_{i}. NIH

_{i}and $RH{T}_{i}^{j}$ are the quantity of the elements and the jth element in TSRHT

_{i}, respectively. The smoothed TSRHT of the broilers in the experimental and control groups are shown in Figure 9a with red cross and blue star points, respectively. The corresponding under-wing temperature time series is shown in Figure 9b.

## 3. Results

#### 3.1. Testing of HSTE

#### 3.2. Overall Trend Analysis for the Smoothed TSRHT

_{i}], $i\in \left[1,20\right]$) elements in a smoothed TSRHT

_{i}, which was denoted by $Slope\_TSRH{T}_{i}^{{t}^{\prime}}$, was used to describe these two different overall trend features. For the ith broiler, all of its $Slope\_TSRH{T}_{i}^{{t}^{\prime}}$ formed a slope time series. This series was denoted by $Slope\_TSRH{T}_{i}$ and expressed in expression (11).

_{i}respectively was 15 and 21, quantity of the elements in each $Slope\_TSRH{T}_{i}$ was between 11 and 17. $Slope\_TSRH{T}_{i}$ of all the 20 broilers are plotted in Figure 13.

^{th}time interval on, almost all the $Slope\_TSRH{T}_{i}^{{t}^{\prime}}$ of the smoothed TSRHT at each t′ in control and experimental groups had non-positive and positive values, respectively. It could be concluded that the positive or non-positive of the $Slope\_TSRH{T}_{i}^{{t}^{\prime}}$ can be used to identify whether a broiler is febrile at a given time interval.

## 4. Discussion

## 5. Conclusions

## Author Contributions

## Funding

## Acknowledgments

## Conflicts of Interest

## References

- Giloh, M.; Shinder, D.; Yahav, S. Skin surface temperature of broiler chickens is correlated to body core temperature and is indicative of their thermoregulatory status. Poult. Sci.
**2012**, 91, 175–188. [Google Scholar] [CrossRef] [PubMed] - Moller, A.P. Body temperature and fever in a free-living bird. Comp. Biochem. Physiol. B Biochem. Mol. Biol.
**2010**, 156, 68–74. [Google Scholar] [CrossRef] [PubMed] - Cooper, M.A.; Washburn, K.W. The relationships of body temperature to weight gain, feed consumption, and feed utilization in broilers under heat stress. Poult. Sci.
**1998**, 77, 237–242. [Google Scholar] [CrossRef] [PubMed] - Iwasaki, W.; Ishida, S.; Kondo, D.; Ito, Y.; Tateno, J.; Tomioka, M. Monitoring of the core body temperature of cows using implantable wireless thermometers. Comp. Electron. Agricult.
**2019**, 163, 104849. [Google Scholar] [CrossRef] - Nogami, H.; Arai, S.; Okada, H.; Zhan, L.; Itoh, T. Minimized bolus-type wireless sensor node with a built-in three-axis acceleration meter for monitoring a Cow’s Rumen conditions. Sensors
**2017**, 17, 687. [Google Scholar] [CrossRef] - Andersson, L.M.; Okada, H.; Miura, R.; Zhang, Y.; Yoshioka, K.; Aso, H.; Itoh, T. Wearable wireless estrus detection sensor for cows. Comp. Electron. Agricult.
**2016**, 127, 101–108. [Google Scholar] [CrossRef] - Reid, E.; Fried, K.; Velasco, J.; Dahl, G. Correlation of rectal temperature and peripheral temperature from implantable radio-frequency microchips in Holstein steers challenged with lipopolysaccharide under thermoneutral and high ambient temperatures. J. Anim. Science
**2012**, 90, 4788–4794. [Google Scholar] [CrossRef] - Li, L.; Chen, H.; Yu, H.; Huang, R.; Huo, L. Dynamic monitoring device of hens temperature based on wireless transmission. Trans. Chin. Soc. Agricult. Mach.
**2013**, 6, 242–245. [Google Scholar] - Zaninelli, M.; Redaelli, V.; Luzi, F.; Bronzo, V.; Mitchell, M.; Dell’Orto, V.; Bontempo, V.; Cattaneo, D.; Savoini, G. First evaluation of infrared thermography as a tool for the monitoring of udder health status in farms of dairy cows. Sensors
**2018**, 18, 862. [Google Scholar] [CrossRef] - Alsaaod, M.; Schaefer, A.; Büscher, W.; Steiner, A. The role of infrared thermography as a non-invasive tool for the detection of lameness in cattle. Sensors
**2015**, 15, 14513–14525. [Google Scholar] [CrossRef] - Soerensen, D.D.; Pedersen, L.J. Infrared skin temperature measurements for monitoring health in pigs: A review. Acta Veterinaria Scandinavica
**2015**, 57, 5. [Google Scholar] [CrossRef] [PubMed] - LokeshBabu, D.S.; Jeyakumar, S.; Vasant, P.J.; Sathiyabarathi, M.; Manimaran, A.; Kumaresan, A.; Pushpadass, H.A.; Sivaram, M.; Ramesha, K.P.; Kataktalware, M.A. Siddaramanna, Monitoring foot surface temperature using infrared thermal imaging for assessment of hoof health status in cattle: A review. J. Therm. Biol.
**2018**, 78, 10–21. [Google Scholar] - Schaefer, A.L.; Cook, N.; Tessaro, S.V.; Deregt, D.; Desroches, G.; Dubeski, P.L.; Tong, A.K.W.; Godson, D.L. Early detection and prediction of infection using infrared thermography. Can. J. Anim. Sci.
**2004**, 84, 73–80. [Google Scholar] [CrossRef] - Liu, X.; Liu, Y.; Liu, X.; Cao, C.; Lu, H.; Wang, F.; Ge, Y.; Lu, Y. Feasibility study for the use of surface-temperature discriminant poultry health. In 2017 ASABE Annual International Meeting; ASABE: St. Joseph, MI, USA, 2017; p. 1. [Google Scholar]
- Naeaes, I.d.A.; Bites Romanini, C.E.; Neves, D.P.; do Nascimento, G.R.; Vercellino, R.d.A. Broiler surface temperature distribution of 42 day old chickens. Scientia Agricola
**2010**, 67, 497–502. [Google Scholar] [CrossRef] - McManus, C.; Tanure, C.B.; Peripolli, V.; Seixas, L.; Fischer, V.; Gabbi, A.M.; Menegassi, S.R.; Stumpf, M.T.; Kolling, G.J.; Dias, E. Infrared thermography in animal production: An overview. Comp. Electron. Agricult.
**2016**, 123, 10–16. [Google Scholar] [CrossRef] - Otsu, N. A threshold selection method from gray-level histograms. IEEE Trans. Syst. Man Cybernet.
**1979**, 9, 62–66. [Google Scholar] [CrossRef] - Vincent, L. Morphological area openings and closings for grey-scale images. In Shape in Picture; Springer: Cham, Switzerland, 1994; pp. 197–208. [Google Scholar]
- Barber, C.B.; Dobkin, D.P.; Huhdanpaa, H. The Quickhull algorithm for convex hulls. ACM Trans. Math. Software
**1996**, 22, 469–483. [Google Scholar] [CrossRef] - Halir, R.; Flusser, J. Numerically Stable Direct Least Squares Fitting of Ellipses; University of West Bohemia Press: Plzen, Czech Republic, 1998; pp. 125–132. [Google Scholar]
- Lu, M.; Xiong, Y.; Li, K.; Liu, L.; Yan, L.; Ding, Y.; Lin, X.; Yang, X.; Shen, M. An automatic splitting method for the adhesive piglets’ gray scale image based on the ellipse shape feature. Comp. Electron. Agricult.
**2016**, 120, 53–62. [Google Scholar] [CrossRef] - Canny, J. A computational approach to edge detection. IEEE Trans. Pattern Anal. Mach. Intell.
**1986**, 8, 679–698. [Google Scholar] [CrossRef] - Arthur, D.; Vassilvitskii, S. Siam/Acm, k-Means Plus Plus: The Advantages of Careful Seeding; IEEE: Piscataway, NJ, USA, 2007; pp. 1027–1035. [Google Scholar]
- Leisch, F. A toolbox for -centroids cluster analysis. Comput. Stat. Data Anal.
**2006**, 51, 526–544. [Google Scholar] [CrossRef] - Meltzer, A. The effect of body temperature on the growth rate of broilers. Br. Poult. Sci.
**1983**, 24, 489–495. [Google Scholar] [CrossRef] [PubMed] - Donkoh, A. Ambient-temperature—A factor affecting performance and physiological-response of broiler-chickens. Int. J. Biometeorol.
**1989**, 33, 259–265. [Google Scholar] [CrossRef] [PubMed] - Nasirahmadi, A.; Richter, U.; Hensel, O.; Edwards, S.; Sturm, B. Using machine vision for investigation of changes in pig group lying patterns. Comp. Electron. Agricult.
**2015**, 119, 184–190. [Google Scholar] [CrossRef] - Psota, E.T.; Mittek, M.; Pérez, L.C.; Schmidt, T.; Mote, B. Multi-pig part detection and association with a fully-convolutional network. Sensors
**2019**, 19, 852. [Google Scholar] [CrossRef] [Green Version]

**Figure 2.**Temperature matrix and thermal image of an IS2 file: (

**a**) Temperature matrix; (

**b**) The corresponding thermal image.

**Figure 3.**Image pre-processing: (

**a**) Grayscale image; (

**b**) Binary image; (

**c**) Image after morphological processing; (

**d**) Convex hull image.

**Figure 4.**Ellipse fitting for individual broiler body contour: (

**a**) The contour of the convex hull image shown in Figure 3d; (

**b**) The fitted ellipse.

**Figure 6.**Head region locating: (

**a**) The final high temperature regions extracted from the image in Figure 2a; (

**b**) The candidate head regions; (

**c**) The alternative head regions; (

**d**) The extracted head region; (

**e**) Relationship between the extracted head region and the gray-scale thermal image.

**Figure 7.**Examples of the head region locating: (

**a**–

**c**) Head region locating by using case 1; (

**d**–

**f**) Head region locating by using case 3.

**Figure 9.**The smoothed representative head surface temperature time series (TSRHT) and under-wing temperature time series: (

**a**) The smoothed TSRHT time series; (

**b**) Under-wing temperature time series.

**Figure 11.**Head temperature extraction: (

**a**) Head temperature extracted manually by using Smartview; (

**b**) Head temperature extracted automatically by HSTE.

**Figure 12.**The errors between head temperatures extracted automatically by HSTE and by using Smartview.

Category | (i) | (ii) | (iii) | (iv) | (v) | (vi) |
---|---|---|---|---|---|---|

Number of images | 41 | 785 | 849 | 295 | 139 | 76 |

Number of correct locating | 32 | 745 | 806 | 270 | 118 | 56 |

Ratio of correct locating | 78.05% | 94.90% | 94.94% | 91.53% | 84.89% | 73.68% |

t′ | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|

E(%) | 73 | 80 | 80 | 80 | 86.7 | 86.7 | 86.7 | 86.7 | 100 | 100 | 100 | 92.9 | 92.3 | 100 | 100 | 100 | 100 |

C(%) | 80 | 100 | 100 | 100 | 100 | 80 | 80 | 80 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 |

**E**: experimental group;

**C**: control group.

© 2019 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 (http://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Xiong, X.; Lu, M.; Yang, W.; Duan, G.; Yuan, Q.; Shen, M.; Norton, T.; Berckmans, D.
An Automatic Head Surface Temperature Extraction Method for Top-View Thermal Image with Individual Broiler. *Sensors* **2019**, *19*, 5286.
https://doi.org/10.3390/s19235286

**AMA Style**

Xiong X, Lu M, Yang W, Duan G, Yuan Q, Shen M, Norton T, Berckmans D.
An Automatic Head Surface Temperature Extraction Method for Top-View Thermal Image with Individual Broiler. *Sensors*. 2019; 19(23):5286.
https://doi.org/10.3390/s19235286

**Chicago/Turabian Style**

Xiong, Xingguo, Mingzhou Lu, Weizhong Yang, Guanghui Duan, Qingyan Yuan, Mingxia Shen, Tomas Norton, and Daniel Berckmans.
2019. "An Automatic Head Surface Temperature Extraction Method for Top-View Thermal Image with Individual Broiler" *Sensors* 19, no. 23: 5286.
https://doi.org/10.3390/s19235286