Efficient Aggressive Behavior Recognition of Pigs Based on Temporal Shift Module.

Aggressive behavior among pigs is a significant social issue that has severe repercussions on both the profitability and welfare of pig farms. Due to the complexity of aggression, recognizing it requires the consideration of both spatial and temporal features. To address this problem, we proposed an efficient method that utilizes the temporal shift module (TSM) for automatic recognition of pig aggression. In general, TSM is inserted into four 2D convolutional neural network models, including ResNet50, ResNeXt50, DenseNet201, and ConvNext-t, enabling the models to process both spatial and temporal features without increasing the model parameters and computational complexity. The proposed method was evaluated on the dataset established in this study, and the results indicate that the ResNeXt50-T (TSM inserted into ResNeXt50) model achieved the best balance between recognition accuracy and model parameters. On the test set, the ResNeXt50-T model achieved accuracy, recall, precision, F1 score, speed, and model parameters of 95.69%, 95.25%, 96.07%, 95.65%, 29 ms, and 22.98 M, respectively. These results show that the proposed method can effectively improve the accuracy of recognizing pig aggressive behavior and provide a reference for behavior recognition in actual scenarios of smart livestock farming.

Modeling of Heat Stress in Sows Part 2: Comparison of Various Thermal Comfort Indices.

Heat stress has an adverse effect on the production performance of sows, and causes a large economic loss every year. The thermal environment index is an important indicator for evaluating the level of heat stress in animals. Many thermal indices have been used to analyze the environment of the pig house, including temperature and humidity index (THI), effective temperature (ET), equivalent temperature index of sows (ETIS), and enthalpy (H), among others. Different heat indices have different characteristics, and it is necessary to analyze and compare the characteristics of heat indices to select a relatively suitable heat index for specific application. This article reviews the thermal environment indices used in the process of sow breeding, and compares various heat indices in four ways: (1) Holding the value of the thermal index constant and analyzing the equivalent temperature changes caused by the relative humidity. (2) Analyzing the variations of ET and ETIS caused by changes in air velocity. (3) Conducting a comparative analysis of a variety of isothermal lines fitted to the psychrometric chart. (4) Analyzing the distributions of various heat index values inside the sow barn and the correlation between various heat indices and sow heat dissipation with the use of computational fluid dynamics (CFD) technology. The results show that the ETIS performs better than other thermal indices in the analysis of sows' thermal environment, followed by THI2, THI4, and THI7. Different pigs have different heat transfer characteristics and different adaptability to the environment. Therefore, based on the above results, the following suggestions have been given: The thermal index thresholds need to be divided based on the adaptability of pigs to the environment at different growth stages and the different climates in different regions. An appropriate threshold for a thermal index can provide a theoretical basis for the environmental control of the pig house.

Modeling of Heat Stress in Sows-Part 1: Establishment of the Prediction Model for the Equivalent Temperature Index of the Sows.

Heat stress affects the estrus time and conception rate of sows. Compared with other life stages of pigs, sows are more susceptible to heat stress because of their increased heat production. Various indicators can be found in the literature assessing the level of heat stress in pigs. However, none of them is specific to assess the sows' thermal condition. Moreover, thermal indices are mainly developed by considering partial environment parameters, and there is no interaction between the index and the animal's physiological response. Therefore, this study aims to develop a thermal index specified for sows, called equivalent temperature index for sows (ETIS), which includes parameters of air temperature, relative humidity and air velocity. Based on the heat transfer characteristics of sows, multiple regression analysis is used to combine air temperature, relative humidity and air velocity. Environmental data are used as independent variables, and physiological parameters are used as dependent variables. In 1029 sets of data, 70% of the data is used as the training set, and 30% of the data is used as the test set to create and develop a new thermal index. According to the correlation equation between ETIS and temperature-humidity index (THI), combined with the threshold of THI, ETIS was divided into thresholds. The results show that the ETIS heat stress threshold is classified as follows: suitable temperature ETIS < 33.1 °C, mild temperature 33.1 °C ≤ ETIS < 34.5 °C, moderate stress temperature 34.5 °C ≤ ETIS < 35.9 °C, and severe temperature ETIS ≥ 35.9 °C. The ETIS model can predict the sows' physiological response in a good manner. The correlation coefficients R of skin temperature was 0.82. Compared to early developed thermal indices, ETIS has the best predictive effect on skin temperature. This index could be a useful tool for assessing the thermal environment to ensure thermal comfort for sows.

Pig Weight and Body Size Estimation Using a Multiple Output Regression Convolutional Neural Network: A Fast and Fully Automatic Method.

Pig weight and body size are important indicators for producers. Due to the increasing scale of pig farms, it is increasingly difficult for farmers to quickly and automatically obtain pig weight and body size. Due to this problem, we focused on a multiple output regression convolutional neural network (CNN) to estimate pig weight and body size. DenseNet201, ResNet152 V2, Xception and MobileNet V2 were modified into multiple output regression CNNs and trained on modeling data. By comparing the estimated performance of each model on test data, modified Xception was selected as the optimal estimation model. Based on pig height, body shape, and contour, the mean absolute error (MAE) of the model to estimate body weight (BW), shoulder width (SW), shoulder height (SH), hip width (HW), hip width (HH), and body length (BL) were 1.16 kg, 0.33 cm, 1.23 cm, 0.38 cm, 0.66 cm, and 0.75 cm, respectively. The coefficient of determination (R2) value between the estimated and measured results was in the range of 0.9879-0.9973. Combined with the LabVIEW software development platform, this method can estimate pig weight and body size accurately, quickly, and automatically. This work contributes to the automatic management of pig farms.

Oestrus Analysis of Sows Based on Bionic Boars and Machine Vision Technology.

This study proposes a method and device for the intelligent mobile monitoring of oestrus on a sow farm, applied in the field of sow production. A bionic boar model that imitates the sounds, smells, and touch of real boars was built to detect the oestrus of sows after weaning. Machine vision technology was used to identify the interactive behaviour between empty sows and bionic boars and to establish deep belief network (DBN), sparse autoencoder (SAE), and support vector machine (SVM) models, and the resulting recognition accuracy rates were 96.12%, 98.25%, and 90.00%, respectively. The interaction times and frequencies between the sow and the bionic boar and the static behaviours of both ears during heat were further analysed. The results show that there is a strong correlation between the duration of contact between the oestrus sow and the bionic boar and the static behaviours of both ears. The average contact duration between the sows in oestrus and the bionic boars was 29.7 s/3 min, and the average duration in which the ears of the oestrus sows remained static was 41.3 s/3 min. The interactions between the sow and the bionic boar were used as the basis for judging the sow's oestrus states. In contrast with the methods of other studies, the proposed innovative design for recyclable bionic boars can be used to check emotions, and machine vision technology can be used to quickly identify oestrus behaviours. This approach can more accurately obtain the oestrus duration of a sow and provide a scientific reference for a sow's conception time.

Assessment of Laying Hens' Thermal Comfort Using Sound Technology.

Heat stress is one of the most important environmental stressors facing poultry production and welfare worldwide. The detrimental effects of heat stress on poultry range from reduced growth and egg production to impaired health. Animal vocalisations are associated with different animal responses and can be used as useful indicators of the state of animal welfare. It is already known that specific chicken vocalisations such as alarm, squawk, and gakel calls are correlated with stressful events, and therefore, could be used as stress indicators in poultry monitoring systems. In this study, we focused on developing a hen vocalisation detection method based on machine learning to assess their thermal comfort condition. For extraction of the vocalisations, nine source-filter theory related temporal and spectral features were chosen, and a support vector machine (SVM) based classifier was developed. As a result, the classification performance of the optimal SVM model was 95.1 ± 4.3% (the sensitivity parameter) and 97.6 ± 1.9% (the precision parameter). Based on the developed algorithm, the study illustrated that a significant correlation existed between specific vocalisations (alarm and squawk call) and thermal comfort indices (temperature-humidity index, THI) (alarm-THI, R = -0.414, P = 0.01; squawk-THI, R = 0.594, P = 0.01). This work represents the first step towards the further development of technology to monitor flock vocalisations with the intent of providing producers an additional tool to help them actively manage the welfare of their flock.

Concentration and size distribution of particulate matter in a new aviary system for laying hens in China.

Particulate matter (PM) from poultry production facilities may strongly affect the health of animals and workers in the houses, and PM emitted to the ambient air is an important pollution source to the surrounding areas. Aviary system is considered as a welfare friendly production system for laying hens. However, its air quality is typically worse as compared with conventional cage systems, because of the higher PM concentration of indoor air and other airborne contaminants. Furthermore, PM's physical property, which has a direct impact on the penetration depth into the lungs of the birds and humans, is largely unknown for the aviary system. Therefore, a systematic method was utilized to investigate the characteristics of particles in the aviary house with large cage aviary unit system (LCAU) in Beijing, China. For the field measurements, three measuring locations were selected with two inside and one outside the house with LCAU to continuously monitor PM concentrations and collect the samples for particle size distribution (PSD) analysis. Results showed that PM2.5, PM10, and total suspended particulate (TSP) concentrations averaged at 0.037 ± 0.025 mg/m3, 0.42 ± 0.10 mg/m3, and 1.92 ± 1.91 mg/m3, respectively. Particle concentrations increased from October to December due to less ventilation as the weather got colder, and were generally affected by stocking density, ventilation rate, birds' activities, and housing system. Meanwhile, indoor PM2.5 concentration was easily impacted by the ambient air quality. Mass median diameter (MMD) and mass geometric standard deviation (MGSD) of the TSP during the measurement were 18.92 ± 7.08 µm and 3.11 ± 0.31, respectively. Count median diameter (CMD) and count geometric standard deviation (CGSD) were 1.94 ± 0.14 µm and 1.48 ± 0.08, respectively. Results indicated that the aviary system can attain a good indoor condition by suitable system design and environment control strategy.Implications: Indoor PM2.5 concentration of the layer house can be significantly affected by ambient air quality when the air quality index (AQI) was larger than 100. PM2.5 and PM10 concentrations of the layer house with a LCAU system were comparable to the cage system. TSP concentration was higher, and PM size was larger than most of the cage system. System design, larger space volume, and higher ventilation rate were the main influence factors. Good indoor environment of the aviary system can be achieved through the reasonable design of the production system and appropriate environment control strategy.

Influence of deployment time and surface wind speed on the accuracy of measurements of greenhouse gas fluxes using a closed chamber method under low surface wind speed.

As a convenient method, the closed chamber method has been applied to determine gaseous emission fluxes from fully open animal feeding operations despite the measured fluxes being theoretically affected by deployment time, wind speed over the emitting surface and detected gas mass. This laboratory study evaluated the effects of deployment time (0 to 120 min) and external surface wind speed (ESWS) (0.00, 0.25, 0.50, 0.75, 1.00, 1.50 and 2.00 m sec-1) on the measurement accuracy of a 300 mm (diameter) × 400 mm (height) (D300×H400) closed chamber using methane (CH4), nitrous oxide (N2O) and sulfur hexafluoride (SF6) as reference gases. The results showed that the overall deviation ratio between the measured and reference CH4 fluxes ranged from 9.99 % to -37.32 % and the flux was overestimated in the first 20 min. The measured N2O and SF6 emissions were smaller than the reference fluxes using the chamber. N2O measurement accuracy decreased from -14.47 to -35.09% with deployment time extended to 120 min, while SF6 accuracy sharply increased in the first 40 min, with the deviation stabilizing at approximately -5.00%. CH4, N2O and SF6 measurements were significantly affected by deployment time and ESWS (P<0.05), and the interaction of those two factors greatly influenced CH4 and SF6 measurements (P<0.05). With the D300×H400 closed chamber, deployment times of 20 to 30 min and 10 to 20 min are recommended to measure CH4 and N2O, respectively, from the open operations of dairy farms under wind speeds lower than 2 m sec-1. Implications: This study recommended the suitable deployment times and wind speeds for using a D300 × H400 closed chamber to measure CH4, N2O, and SF6 in an open system, such as a dairy open lot and manure stockpile, to help researchers and other related industry workers get accurate data for gas emission rate. Deployment times of 20 to 30 min and 10 to 20 min were recommended to measure CH4 and N2O emissions using the D300 × H400 closed chamber, respectively, from the open operations of dairy farms under wind speeds lower than 2 m sec-1. For the measurement of SF6, a typical tracer gas, a deployment of 70 to 90 min was suggested.

A Sound Source Localisation Analytical Method for Monitoring the Abnormal Night Vocalisations of Poultry.

Due to the increasing scale of farms, it is increasingly difficult for farmers to monitor their animals in an automated way. Because of this problem, we focused on a sound technique to monitor laying hens. Sound analysis has become an important tool for studying the behaviour, health and welfare of animals in recent years. A surveillance system using microphone arrays of Kinects was developed for automatically monitoring birds' abnormal vocalisations during the night. Based on the principle of time-difference of arrival (TDOA) of sound source localisation (SSL) method, Kinect sensor direction estimations were very accurate. The system had an accuracy of 74.7% in laboratory tests and 73.6% in small poultry group tests for different area sound recognition. Additionally, flocks produced an average of 40 sounds per bird during feeding time in small group tests. It was found that, on average, each normal chicken produced more than 53 sounds during the daytime (noon to 6:00 p.m.) and less than one sound at night (11:00 p.m.⁻3:00 a.m.). This system can be used to detect anomalous poultry status at night by monitoring the number of vocalisations and area distributions, which provides a practical and feasible method for the study of animal behaviour and welfare.