Automatic detection and recognition of pig wasting diseases using sound data in audio surveillance systems

DCNN for Pig Vocalization and Non-Vocalization Classification: Evaluate Model Robustness with New Data

Since pig vocalization is an important indicator of monitoring pig conditions, pig vocalization detection and recognition using deep learning play a crucial role in the management and welfare of modern pig livestock farming. However, collecting pig sound data for deep learning model training takes time and effort. Acknowledging the challenges of collecting pig sound data for model training, this study introduces a deep convolutional neural network (DCNN) architecture for pig vocalization and non-vocalization classification with a real pig farm dataset. Various audio feature extraction methods were evaluated individually to compare the performance differences, including Mel-frequency cepstral coefficients (MFCC), Mel-spectrogram, Chroma, and Tonnetz. This study proposes a novel feature extraction method called Mixed-MMCT to improve the classification accuracy by integrating MFCC, Mel-spectrogram, Chroma, and Tonnetz features. These feature extraction methods were applied to extract relevant features from the pig sound dataset for input into a deep learning network. For the experiment, three datasets were collected from three actual pig farms: Nias, Gimje, and Jeongeup. Each dataset consists of 4000 WAV files (2000 pig vocalization and 2000 pig non-vocalization) with a duration of three seconds. Various audio data augmentation techniques are utilized in the training set to improve the model performance and generalization, including pitch-shifting, time-shifting, time-stretching, and background-noising. In this study, the performance of the predictive deep learning model was assessed using the k-fold cross-validation (k = 5) technique on each dataset. By conducting rigorous experiments, Mixed-MMCT showed superior accuracy on Nias, Gimje, and Jeongeup, with rates of 99.50%, 99.56%, and 99.67%, respectively. Robustness experiments were performed to prove the effectiveness of the model by using two farm datasets as a training set and a farm as a testing set. The average performance of the Mixed-MMCT in terms of accuracy, precision, recall, and F1-score reached rates of 95.67%, 96.25%, 95.68%, and 95.96%, respectively. All results demonstrate that the proposed Mixed-MMCT feature extraction method outperforms other methods regarding pig vocalization and non-vocalization classification in real pig livestock farming.

Detecting tail biters by monitoring pig screams in weaning pigs

Early identification of tail biting and intervention are necessary to reduce tail lesions and their impact on animal health and welfare. Removal of biters has become an effective intervention strategy, but finding them can be difficult and time-consuming. The aim of this study was to investigate whether tail biting and, in particular, individual biters could be identified by detecting pig screams in audio recordings. The study included 288 undocked weaner pigs housed in six pens in two batches. Once a tail biter (n = 7) was identified by visual inspection in the stable and removed by the farm staff, the previous days of video and audio recordings were analyzed for pig screams (sudden increase in loudness with frequencies above 1 kHz) and tail biting events until no biting before the removal was observed anymore. In total, 2893 screams were detected in four pens where tail biting occurred. Of these screams, 52.9% were caused by tail biting in the observed pen, 25.6% originated from other pens, 8.8% were not assignable, and 12.7% occurred due to other reasons. In case of a tail biting event, screams were assigned individually to biter and victim pigs. Based on the audio analysis, biters were identified between one and nine days prior to their removal from the pen after visual inspection. Screams were detected earlier than the increase in hanging tails and could therefore be favored as an early warning indicator. Analyzing animal vocalization has potential for monitoring and early detection of tail biting events. In combination with individual marks and automatic analysis algorithms, biters could be identified and tail biting efficiently reduced. In this way, biters can be removed earlier to increase animal health and welfare.

"Smart markets": harnessing the potential of new technologies for endemic and emerging infectious disease surveillance in traditional food markets

Emerging and endemic zoonotic diseases continue to threaten human and animal health, our social fabric, and the global economy. Zoonoses frequently emerge from congregate interfaces where multiple animal species and humans coexist, including farms and markets. Traditional food markets are widespread across the globe and create an interface where domestic and wild animals interact among themselves and with humans, increasing the risk of pathogen spillover. Despite decades of evidence linking markets to disease outbreaks across the world, there remains a striking lack of pathogen surveillance programs that can relay timely, cost-effective, and actionable information to decision-makers to protect human and animal health. However, the strategic incorporation of environmental surveillance systems in markets coupled with novel pathogen detection strategies can create an early warning system capable of alerting us to the risk of outbreaks before they happen. Here, we explore the concept of "smart" markets that utilize continuous surveillance systems to monitor the emergence of zoonotic pathogens with spillover potential.IMPORTANCEFast detection and rapid intervention are crucial to mitigate risks of pathogen emergence, spillover and spread-every second counts. However, comprehensive, active, longitudinal surveillance systems at high-risk interfaces that provide real-time data for action remain lacking. This paper proposes "smart market" systems harnessing cutting-edge tools and a range of sampling techniques, including wastewater and air collection, multiplex assays, and metagenomic sequencing. Coupled with robust response pathways, these systems could better enable Early Warning and bolster prevention efforts.

Research on Pig Sound Recognition Based on Deep Neural Network and Hidden Markov Models

In order to solve the problem of low recognition accuracy of traditional pig sound recognition methods, deep neural network (DNN) and Hidden Markov Model (HMM) theory were used as the basis of pig sound signal recognition in this study. In this study, the sounds made by 10 landrace pigs during eating, estrus, howling, humming and panting were collected and preprocessed by Kalman filtering and an improved endpoint detection algorithm based on empirical mode decomposition-Teiger energy operator (EMD-TEO) cepstral distance. The extracted 39-dimensional mel-frequency cepstral coefficients (MFCCs) were then used as a dataset for network learning and recognition to build a DNN- and HMM-based sound recognition model for pig states. The results show that in the pig sound dataset, the recognition accuracy of DNN-HMM reaches 83%, which is 22% and 17% higher than that of the baseline models HMM and GMM-HMM, and possesses a better recognition effect. In a sub-dataset of the publicly available dataset AudioSet, DNN-HMM achieves a recognition accuracy of 79%, which is 8% and 4% higher than the classical models SVM and ResNet18, respectively, with better robustness.

Study on a Pig Vocalization Classification Method Based on Multi-Feature Fusion

To improve the classification of pig vocalization using vocal signals and improve recognition accuracy, a pig vocalization classification method based on multi-feature fusion is proposed in this study. With the typical vocalization of pigs in large-scale breeding houses as the research object, short-time energy, frequency centroid, formant frequency and first-order difference, and Mel frequency cepstral coefficient and first-order difference were extracted as the fusion features. These fusion features were improved using principal component analysis. A pig vocalization classification model with a BP neural network optimized based on the genetic algorithm was constructed. The results showed that using the improved features to recognize pig grunting, squealing, and coughing, the average recognition accuracy was 93.2%; the recognition precisions were 87.9%, 98.1%, and 92.7%, respectively, with an average of 92.9%; and the recognition recalls were 92.0%, 99.1%, and 87.4%, respectively, with an average of 92.8%, which indicated that the proposed pig vocalization classification method had good recognition precision and recall, and could provide a reference for pig vocalization information feedback and automatic recognition.

Artificial Intelligence for Automatic Monitoring of Respiratory Health Conditions in Smart Swine Farming

Porcine respiratory disease complex is an economically important disease in the swine industry. Early detection of the disease is crucial for immediate response to the disease at the farm level to prevent and minimize the potential damage that it may cause. In this paper, recent studies on the application of artificial intelligence (AI) in the early detection and monitoring of respiratory disease in swine have been reviewed. Most of the studies used coughing sounds as a feature of respiratory disease. The performance of different models and the methodologies used for cough recognition using AI were reviewed and compared. An AI technology available in the market was also reviewed. The device uses audio technology that can monitor and evaluate the herd's respiratory health status through cough-sound recognition and quantification. The device also has temperature and humidity sensors to monitor environmental conditions. It has an alarm system based on variations in coughing patterns and abrupt temperature changes. However, some limitations of the existing technology were identified. Substantial effort must be exerted to surmount the limitations to have a smarter AI technology for monitoring respiratory health status in swine.

Can We Reliably Detect Respiratory Diseases through Precision Farming? A Systematic Review

Respiratory diseases commonly affect livestock species, negatively impacting animal's productivity and welfare. The use of precision livestock farming (PLF) applied in respiratory disease detection has been developed for several species. The aim of this systematic review was to evaluate if PLF technologies can reliably monitor clinical signs or detect cases of respiratory diseases. A technology was considered reliable if high performance was achieved (sensitivity > 90% and specificity or precision > 90%) under field conditions and using a reliable reference test. Risk of bias was assessed, and only technologies tested in studies with low risk of bias were considered reliable. From 23 studies included-swine (13), poultry (6), and bovine (4) -only three complied with our reliability criteria; however, two of these were considered to have a high risk of bias. Thus, only one swine technology fully fit our criteria. Future studies should include field tests and use previously validated reference tests to assess technology's performance. In conclusion, relying completely on PLF for monitoring respiratory diseases is still a challenge, though several technologies are promising, having high performance in field tests.

Perspectives of digital agriculture in diverse types of livestock supply chain systems. Making sense of uses and benefits

Digital technology is being introduced to global agriculture in a wide variety of forms that are collectively known as digital agriculture. In this paper we provide opportunities and value propositions of how this is occurring in livestock production systems, with a consistent emphasis on technology relating to animal health, animal welfare, and product quality for value creation. This is achieved by organizing individual accounts of digital agriculture in livestock systems according to four broad types-commodity-based; value seeking; subsistence and nature-based. Each type presents contrasting modes of value creation in downstream processing; as well as from the perspective of One Health. The ideal result of digital technology adoption is an equitable and substantial diversification of supply chains, increased monetization of animal product quality, and more sensitive management to meet customer demands and environmental threats. Such changes have a significance beyond the immediate value generated because they indicate endogenous growth in livestock systems, and may concern externalities imposed by the pursuit of purely commercial ends.

Combined spectral and speech features for pig speech recognition

The sound of the pig is one of its important signs, which can reflect various states such as hunger, pain or emotional state, and directly indicates the growth and health status of the pig. Existing speech recognition methods usually start with spectral features. The use of spectrograms to achieve classification of different speech sounds, while working well, may not be the best approach for solving such tasks with single-dimensional feature input. Based on the above assumptions, in order to more accurately grasp the situation of pigs and take timely measures to ensure the health status of pigs, this paper proposes a pig sound classification method based on the dual role of signal spectrum and speech. Spectrograms can visualize information about the characteristics of the sound under different time periods. The audio data are introduced, and the spectrogram features of the model input as well as the audio time-domain features are complemented with each other and passed into a pre-designed parallel network structure. The network model with the best results and the classifier were selected for combination. An accuracy of 93.39% was achieved on the pig speech classification task, while the AUC also reached 0.99163, demonstrating the superiority of the method. This study contributes to the direction of computer vision and acoustics by recognizing the sound of pigs. In addition, a total of 4,000 pig sound datasets in four categories are established in this paper to provide a research basis for later research scholars.

The Research Progress of Vision-Based Artificial Intelligence in Smart Pig Farming

Pork accounts for an important proportion of livestock products. For pig farming, a lot of manpower, material resources and time are required to monitor pig health and welfare. As the number of pigs in farming increases, the continued use of traditional monitoring methods may cause stress and harm to pigs and farmers and affect pig health and welfare as well as farming economic output. In addition, the application of artificial intelligence has become a core part of smart pig farming. The precision pig farming system uses sensors such as cameras and radio frequency identification to monitor biometric information such as pig sound and pig behavior in real-time and convert them into key indicators of pig health and welfare. By analyzing the key indicators, problems in pig health and welfare can be detected early, and timely intervention and treatment can be provided, which helps to improve the production and economic efficiency of pig farming. This paper studies more than 150 papers on precision pig farming and summarizes and evaluates the application of artificial intelligence technologies to pig detection, tracking, behavior recognition and sound recognition. Finally, we summarize and discuss the opportunities and challenges of precision pig farming.

A Review of Monitoring Techniques for Livestock Respiration and Sounds

This article reviews the different techniques used to monitor the respiration and sounds of livestock. Livestock respiration is commonly assessed visually by observing abdomen fluctuation; however, the traditional methods are time consuming, subjective, being therefore impractical for large-scale operations and must rely on automation. Contact and non-contact technologies are used to automatically monitor respiration rate; contact technologies (e.g., accelerometers, pressure sensors, and thermistors) utilize sensors that are physically mounted on livestock while non-contact technologies (e.g., computer vision, thermography, and sound analysis) enable a non-invasive method of monitoring respiration. This work summarizes the advantages and disadvantages of contact and non-contact technologies and discusses the emerging role of non-contact sensors in automating monitoring for large-scale farming operations. This work is the first in-depth examination of automated monitoring technologies for livestock respiratory diseases; the findings and recommendations are important for livestock researchers and practitioners who can gain a better understanding of these different technologies, especially emerging non-contact sensing.

Applications of Smart Technology as a Sustainable Strategy in Modern Swine Farming

The size of the pork market is increasing globally to meet the demand for animal protein, resulting in greater farm size for swine and creating a great challenge to swine farmers and industry owners in monitoring the farm activities and the health and behavior of the herd of swine. In addition, the growth of swine production is resulting in a changing climate pattern along with the environment, animal welfare, and human health issues, such as antimicrobial resistance, zoonosis, etc. The profit of swine farms depends on the optimum growth and good health of swine, while modern farming practices can ensure healthy swine production. To solve these issues, a future strategy should be considered with information and communication technology (ICT)-based smart swine farming, considering auto-identification, remote monitoring, feeding behavior, animal rights/welfare, zoonotic diseases, nutrition and food quality, labor management, farm operations, etc., with a view to improving meat production from the swine industry. Presently, swine farming is not only focused on the development of infrastructure but is also occupied with the application of technological knowledge for designing feeding programs, monitoring health and welfare, and the reproduction of the herd. ICT-based smart technologies, including smart ear tags, smart sensors, the Internet of Things (IoT), deep learning, big data, and robotics systems, can take part directly in the operation of farm activities, and have been proven to be effective tools for collecting, processing, and analyzing data from farms. In this review, which considers the beneficial role of smart technologies in swine farming, we suggest that smart technologies should be applied in the swine industry. Thus, the future swine industry should be automated, considering sustainability and productivity.

Integrating Audio Signal Processing and Deep Learning Algorithms for Gait Pattern Classification in Brazilian Gaited Horses

This study focused on assessing the usefulness of using audio signal processing in the gaited horse industry. A total of 196 short-time audio files (4 s) were collected from video recordings of Brazilian gaited horses. These files were converted into waveform signals (196 samples by 80,000 columns) and divided into training (N = 164) and validation (N = 32) datasets. Twelve single-valued audio features were initially extracted to summarize the training data according to the gait patterns (Marcha Batida—MB and Marcha Picada—MP). After preliminary analyses, high-dimensional arrays of the Mel Frequency Cepstral Coefficients (MFCC), Onset Strength (OS), and Tempogram (TEMP) were extracted and used as input information in the classification algorithms. A principal component analysis (PCA) was performed using the 12 single-valued features set and each audio-feature dataset—AFD (MFCC, OS, and TEMP) for prior data visualization. Machine learning (random forest, RF; support vector machine, SVM) and deep learning (multilayer perceptron neural networks, MLP; convolution neural networks, CNN) algorithms were used to classify the gait types. A five-fold cross-validation scheme with 10 repetitions was employed for assessing the models' predictive performance. The classification performance across models and AFD was also validated with independent observations. The models and AFD were compared based on the classification accuracy (ACC), specificity (SPEC), sensitivity (SEN), and area under the curve (AUC). In the logistic regression analysis, five out of the 12 audio features extracted were significant (p < 0.05) between the gait types. ACC averages ranged from 0.806 to 0.932 for MFCC, from 0.758 to 0.948 for OS and, from 0.936 to 0.968 for TEMP. Overall, the TEMP dataset provided the best classification accuracies for all models. The most suitable method for audio-based horse gait pattern classification was CNN. Both cross and independent validation schemes confirmed that high values of ACC, SPEC, SEN, and AUC are expected for yet-to-be-observed labels, except for MFCC-based models, in which clear overfitting was observed. Using audio-generated data for describing gait phenotypes in Brazilian horses is a promising approach, as the two gait patterns were correctly distinguished. The highest classification performance was achieved by combining CNN and the rhythmic-descriptive AFD.

Welfare Health and Productivity in Commercial Pig Herds

In recent years, there have been very dynamic changes in both pork production and pig breeding technology around the world. The general trend of increasing the efficiency of pig production, with reduced employment, requires optimisation and a comprehensive approach to herd management. One of the most important elements on the way to achieving this goal is to maintain animal welfare and health. The health of the pigs on the farm is also a key aspect in production economics. The need to maintain a high health status of pig herds by eliminating the frequency of different disease units and reducing the need for antimicrobial substances is part of a broadly understood high potential herd management strategy. Thanks to the use of sensors (cameras, microphones, accelerometers, or radio-frequency identification transponders), the images, sounds, movements, and vital signs of animals are combined through algorithms and analysed for non-invasive monitoring of animals, which allows for early detection of diseases, improves their welfare, and increases the productivity of breeding. Automated, innovative early warning systems based on continuous monitoring of specific physiological (e.g., body temperature) and behavioural parameters can provide an alternative to direct diagnosis and visual assessment by the veterinarian or the herd keeper.

Characterizing Sounds of Different Sources in a Commercial Broiler House

Simple Summary

Acoustic signal in commercial broiler houses is a mixture of sounds from different sources. However, the characteristics of sounds from different sources have not been well understood. In this study, the sound frequency ranges of six common sounds, including bird vocalization, fan, feed system, heater, wing flapping and dustbathing, were determined; and their relations with bird age were investigated. The outcome of this research provides valuable information for using sound signal to monitor animal behavior and equipment operation.

Abstract

Audio data collected in commercial broiler houses are mixed sounds of different sources that contain useful information regarding bird health condition, bird behavior, and equipment operation. However, characterizations of the sounds of different sources in commercial broiler houses have not been well established. The objective of this study was, therefore, to determine the frequency ranges of six common sounds, including bird vocalization, fan, feed system, heater, wing flapping, and dustbathing, at bird ages of week 1 to 8 in a commercial Ross 708 broiler house. In addition, the frequencies of flapping (in wing flapping events, flaps/s) and scratching (during dustbathing, scratches/s) behaviors were examined through sound analysis. A microphone was installed in the middle of broiler house at the height of 40 cm above the back of birds to record audio data at a sampling frequency of 44,100 Hz. A top-view camera was installed to continuously monitor bird activities. Total of 85 min audio data were manually labeled and fed to MATLAB for analysis. The audio data were decomposed using Maximum Overlap Discrete Wavelet Transform (MODWT). Decompositions of the six concerned sound sources were then transformed with the Fast Fourier Transform (FFT) method to generate the single-sided amplitude spectrums. By fitting the amplitude spectrum of each sound source into a Gaussian regression model, its frequency range was determined as the span of the three standard deviations (99% CI) away from the mean. The behavioral frequencies were determined by examining the spectrograms of wing flapping and dustbathing sounds. They were calculated by dividing the number of movements by the time duration of complete behavioral events. The frequency ranges of bird vocalization changed from 2481 ± 191–4409 ± 136 Hz to 1058 ± 123–2501 ± 88 Hz as birds grew. For the sound of fan, the frequency range increased from 129 ± 36–1141 ± 50 Hz to 454 ± 86–1449 ± 75 Hz over the flock. The sound frequencies of feed system, heater, wing flapping and dustbathing varied from 0 Hz to over 18,000 Hz. The behavioral frequencies of wing flapping were continuously decreased from week 3 (17 ± 4 flaps/s) to week 8 (10 ± 1 flaps/s). For dustbathing, the behavioral frequencies decreased from 16 ± 2 scratches/s in week 3 to 11 ± 1 scratches/s in week 6. In conclusion, characterizing sounds of different sound sources in commercial broiler houses provides useful information for further advanced acoustic analysis that may assist farm management in continuous monitoring of animal health and behavior. It should be noted that this study was conducted with one flock in a commercial house. The generalization of the results remains to be explored.

On-farm detection of claw lesions in dairy cows based on acoustic analyses and machine learning

Claw lesions are a serious problem on dairy farms, affecting both the health and welfare of the cow. Automated detection of lameness with a practical, on-farm application would support the early detection and treatment of lame cows, potentially reducing the number and severity of claw lesions. Therefore, in this study, a method was proposed for the detection of claw lesions based on the acoustic analysis of a cow's gait. A panel was constructed to measure the impact sound of animals walking over it. The recorded impact sound was edited, and 640 sound files from 64 cows were analyzed. The classification of animal-lameness status was performed using a machine-learning process with a random forest algorithm. The gold standard was a 2-point scale of hoof-trimming results (healthy vs. affected), and 38 properties of the recorded sound files were used as influencing factors. A prediction model for classifying the cow lameness was built using a random forest algorithm. This was validated by comparing the reference output from hoof-trimming with the model output concerning the impact sound. Altering the likelihood settings and changing the cutoff value to predict lame animals improved the prediction model. At a cutoff at 0.4, a decreased false-negative rate was generated, and the false-positive rate only increased slightly. This model obtained a sensitivity of 0.81 and a specificity of 0.97. With this procedure, Cohen's Kappa value of 0.80 showed good agreement between model classification and diagnoses from hoof-trimming. In summary, the prediction model enabled the detection of cows with claw lesions. This study shows that lameness can be detected by machine learning from the impact sound of hoofs in dairy cows.

Field-Applicable Pig Anomaly Detection System Using Vocalization for Embedded Board Implementations

Failure to quickly and accurately detect abnormal situations, such as the occurrence of infectious diseases, in pig farms can cause significant damage to the pig farms and the pig farming industry of the country. In this study, we propose an economical and lightweight sound-based pig anomaly detection system that can be applicable even in small-scale farms. The system consists of a pipeline structure, starting from sound acquisition to abnormal situation detection, and can be installed and operated in an actual pig farm. It has the following structure that makes it executable on the embedded board TX-2: (1) A module that collects sound signals; (2) A noise-robust preprocessing module that detects sound regions from signals and converts them into spectrograms; and (3) A pig anomaly detection module based on MnasNet, a lightweight deep learning method, to which the 8-bit filter clustering method proposed in this study is applied, reducing its size by 76.3% while maintaining its identification performance. The proposed system recorded an F1-score of 0.947 as a stable pig’s abnormality identification performance, even in various noisy pigpen environments, and the system’s execution time allowed it to perform in real time.

EmbeddedPigDet—Fast and Accurate Pig Detection for Embedded Board Implementations

Automated pig monitoring is an important issue in the surveillance environment of a pig farm. For a large-scale pig farm in particular, practical issues such as monitoring cost should be considered but such consideration based on low-cost embedded boards has not yet been reported. Since low-cost embedded boards have more limited computing power than typical PCs and have tradeoffs between execution speed and accuracy, achieving fast and accurate detection of individual pigs for “on-device” pig monitoring applications is very challenging. Therefore, in this paper, we propose a method for the fast detection of individual pigs by reducing the computational workload of 3 × 3 convolution in widely-used, deep learning-based object detectors. Then, in order to recover the accuracy of the “light-weight” deep learning-based object detector, we generate a three-channel composite image as its input image, through “simple” image preprocessing techniques. Our experimental results on an NVIDIA Jetson Nano embedded board show that the proposed method can improve the integrated performance of both execution speed and accuracy of widely-used, deep learning-based object detectors, by a factor of up to 8.7.

Recording behaviour of indoor-housed farm animals automatically using machine vision technology: A systematic review

Large-scale phenotyping of animal behaviour traits is time consuming and has led to increased demand for technologies that can automate these procedures. Automated tracking of animals has been successful in controlled laboratory settings, but recording from animals in large groups in highly variable farm settings presents challenges. The aim of this review is to provide a systematic overview of the advances that have occurred in automated, high throughput image detection of farm animal behavioural traits with welfare and production implications. Peer-reviewed publications written in English were reviewed systematically following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. After identification, screening, and assessment for eligibility, 108 publications met these specifications and were included for qualitative synthesis. Data collected from the papers included camera specifications, housing conditions, group size, algorithm details, procedures, and results. Most studies utilized standard digital colour video cameras for data collection, with increasing use of 3D cameras in papers published after 2013. Papers including pigs (across production stages) were the most common (n = 63). The most common behaviours recorded included activity level, area occupancy, aggression, gait scores, resource use, and posture. Our review revealed many overlaps in methods applied to analysing behaviour, and most studies started from scratch instead of building upon previous work. Training and validation sample sizes were generally small (mean±s.d. groups = 3.8±5.8) and in data collection and testing took place in relatively controlled environments. To advance our ability to automatically phenotype behaviour, future research should build upon existing knowledge and validate technology under commercial settings and publications should explicitly describe recording conditions in detail to allow studies to be reproduced.

Noise-Robust Sound-Event Classification System with Texture Analysis

Sound-event classification has emerged as an important field of research in recent years. In particular, investigations using sound data are being conducted in various industrial fields. However, sound-event classification tasks have become more difficult and challenging with the increase in noise levels. In this study, we propose a noise-robust system for the classification of sound data. In this method, we first convert one-dimensional sound signals into two-dimensional gray-level images using normalization, and then extract the texture images by means of the dominant neighborhood structure (DNS) technique. Finally, we experimentally validate the noise-robust approach by using four classifiers (convolutional neural network (CNN), support vector machine (SVM), k-nearest neighbors(k-NN), and C4.5). The experimental results showed superior classification performance in noisy conditions compared with other methods. The F1 score exceeds 98.80% in railway data, and 96.57% in livestock data. Besides, the proposed method can be implemented in a cost-efficient manner (for instance, use of a low-cost microphone) while maintaining high level of accuracy in noisy environments. This approach can be used either as a standalone solution or as a supplement to the known methods to obtain a more accurate solution.

A Combined Deep Learning GRU-Autoencoder for the Early Detection of Respiratory Disease in Pigs Using Multiple Environmental Sensors

We designed and evaluated an assumption-free, deep learning-based methodology for animal health monitoring, specifically for the early detection of respiratory disease in growing pigs based on environmental sensor data. Two recurrent neural networks (RNNs), each comprising gated recurrent units (GRUs), were used to create an autoencoder (GRU-AE) into which environmental data, collected from a variety of sensors, was processed to detect anomalies. An autoencoder is a type of network trained to reconstruct the patterns it is fed as input. By training the GRU-AE using environmental data that did not lead to an occurrence of respiratory disease, data that did not fit the pattern of "healthy environmental data" had a greater reconstruction error. All reconstruction errors were labelled as either normal or anomalous using threshold-based anomaly detection optimised with particle swarm optimisation (PSO), from which alerts are raised. The results from the GRU-AE method outperformed state-of-the-art techniques, raising alerts when such predictions deviated from the actual observations. The results show that a change in the environment can result in occurrences of pigs showing symptoms of respiratory disease within 1⁻7 days, meaning that there is a period of time during which their keepers can act to mitigate the negative effect of respiratory diseases, such as porcine reproductive and respiratory syndrome (PRRS), a common and destructive disease endemic in pigs.

Early detection of health and welfare compromises through automated detection of behavioural changes in pigs

Early detection of health and welfare compromises in commercial piggeries is essential for timely intervention to enhance treatment success, reduce impact on welfare, and promote sustainable pig production. Behavioural changes that precede or accompany subclinical and clinical signs may have diagnostic value. Often referred to as sickness behaviour, this encompasses changes in feeding, drinking, and elimination behaviours, social behaviours, and locomotion and posture. Such subtle changes in behaviour are not easy to quantify and require lengthy observation input by staff, which is impractical on a commercial scale. Automated early-warning systems may provide an alternative by objectively measuring behaviour with sensors to automatically monitor and detect behavioural changes. This paper aims to: (1) review the quantifiable changes in behaviours with potential diagnostic value; (2) subsequently identify available sensors for measuring behaviours; and (3) describe the progress towards automating monitoring and detection, which may allow such behavioural changes to be captured, measured, and interpreted and thus lead to automation in commercial, housed piggeries. Multiple sensor modalities are available for automatic measurement and monitoring of behaviour, which require humans to actively identify behavioural changes. This has been demonstrated for the detection of small deviations in diurnal drinking, deviations in feeding behaviour, monitoring coughs and vocalisation, and monitoring thermal comfort, but not social behaviour. However, current progress is in the early stages of developing fully automated detection systems that do not require humans to identify behavioural changes; e.g., through automated alerts sent to mobile phones. Challenges for achieving automation are multifaceted and trade-offs are considered between health, welfare, and costs, between analysis of individuals and groups, and between generic and compromise-specific behaviours.

Early recognition of bovine respiratory disease in calves using automated continuous monitoring of cough sounds

Bovine respiratory disease (BRD) complex in calves impairs health and welfare and causes severe economic losses for the Stockperson. Early recognition of BRD should lead to earlier veterinary (antibiotic/anti-inflammatory) treatment interventions thereby reducing the severity of the disease and associated costs. Coughing is one of the clinical manifestations of BRD. It is believed that by automatically and continuously monitoring the sounds within calf houses, and analysing the coughing frequency, early recognition of BRD in calves is possible. Therefore, the objective of the present study was to develop an automated calf cough monitor and examine its potential as an early warning system for BRD in artificially reared dairy calves. The coughing sounds of 62 calves were continuously recorded by a microphone over a three-month period. A sound analysis algorithm was developed to distinguish calf coughs from other sounds (e.g. mechanical sounds). During the sound recording period the health of the calves was assessed and scored periodically per week by a trained human observer. Calves presenting with BRD received antibiotic and/or anti-inflammatory treatment and the dates of treatment were recorded. This treatment date reference served as a comparison for the investigation of whether an increase in coughing frequency could be related to calves developing BRD. The calf cough detection algorithm achieved 50.3% sensitivity, 99.2% specificity and 87.5% precision. Four out of five periods, where coughing frequency was observed to be increased, coincided with the development of BRD in more than one calf. This period of increased coughing frequency was always observed before the calves were treated. Therefore, the calf cough monitor has the potential to identify early onset of BRD in calves.

Evaluation of MPEG-7-Based Audio Descriptors for Animal Voice Recognition over Wireless Acoustic Sensor Networks

Environmental audio monitoring is a huge area of interest for biologists all over the world. This is why some audio monitoring system have been proposed in the literature, which can be classified into two different approaches: acquirement and compression of all audio patterns in order to send them as raw data to a main server; or specific recognition systems based on audio patterns. The first approach presents the drawback of a high amount of information to be stored in a main server. Moreover, this information requires a considerable amount of effort to be analyzed. The second approach has the drawback of its lack of scalability when new patterns need to be detected. To overcome these limitations, this paper proposes an environmental Wireless Acoustic Sensor Network architecture focused on use of generic descriptors based on an MPEG-7 standard. These descriptors demonstrate it to be suitable to be used in the recognition of different patterns, allowing a high scalability. The proposed parameters have been tested to recognize different behaviors of two anuran species that live in Spanish natural parks; the Epidalea calamita and the Alytes obstetricans toads, demonstrating to have a high classification performance.

Automatic Recognition of Aggressive Behavior in Pigs Using a Kinect Depth Sensor

Aggression among pigs adversely affects economic returns and animal welfare in intensive pigsties. In this study, we developed a non-invasive, inexpensive, automatic monitoring prototype system that uses a Kinect depth sensor to recognize aggressive behavior in a commercial pigpen. The method begins by extracting activity features from the Kinect depth information obtained in a pigsty. The detection and classification module, which employs two binary-classifier support vector machines in a hierarchical manner, detects aggressive activity, and classifies it into aggressive sub-types such as head-to-head (or body) knocking and chasing. Our experimental results showed that this method is effective for detecting aggressive pig behaviors in terms of both cost-effectiveness (using a low-cost Kinect depth sensor) and accuracy (detection and classification accuracies over 95.7% and 90.2%, respectively), either as a standalone solution or to complement existing methods.

Fault Detection and Diagnosis of Railway Point Machines by Sound Analysis

Railway point devices act as actuators that provide different routes to trains by driving switchblades from the current position to the opposite one. Point failure can significantly affect railway operations, with potentially disastrous consequences. Therefore, early detection of anomalies is critical for monitoring and managing the condition of rail infrastructure. We present a data mining solution that utilizes audio data to efficiently detect and diagnose faults in railway condition monitoring systems. The system enables extracting mel-frequency cepstrum coefficients (MFCCs) from audio data with reduced feature dimensions using attribute subset selection, and employs support vector machines (SVMs) for early detection and classification of anomalies. Experimental results show that the system enables cost-effective detection and diagnosis of faults using a cheap microphone, with accuracy exceeding 94.1% whether used alone or in combination with other known methods.

Stress detection and classification of laying hens by sound analysis

Stress adversely affects the wellbeing of commercial chickens, and comes with an economic cost to the industry that cannot be ignored. In this paper, we first develop an inexpensive and non-invasive, automatic online-monitoring prototype that uses sound data to notify producers of a stressful situation in a commercial poultry facility. The proposed system is structured hierarchically with three binary-classifier support vector machines. First, it selects an optimal acoustic feature subset from the sound emitted by the laying hens. The detection and classification module detects the stress from changes in the sound and classifies it into subsidiary sound types, such as physical stress from changes in temperature, and mental stress from fear. Finally, an experimental evaluation was performed using real sound data from an audio-surveillance system. The accuracy in detecting stress approached 96.2%, and the classification model was validated, confirming that the average classification accuracy was 96.7%, and that its recall and precision measures were satisfactory.