Oestrus Detection In Dairy Cows Research Articles

420 Automated identification and behavioral estrus detection in freestall-housed dairy cows using computer vision and machine learning: A two-layered algorithm approach

Abstract A critical aspect of dairy management involves maintaining accurate records of individual animals and promptly detecting estrus behavior, which is vital for successful breeding and reproductive outcomes. Recent advancements in precision livestock management, utilizing computer vision and machine learning technologies, have paved the way for automated and precise solutions in these areas. The objective of this study is to investigate the application of computer vision and machine learning for animal identification and behavioral estrus detection in dairy cows housed in freestall settings, employing a two-layered algorithmic approach. The research was conducted at the Joe Bearden Dairy Research Center, where a total of 10 lactating Holstein-Friesian cows were assessed across a 7-d trial period. For synchronization efforts, cows were provided a single injection of Prostaglandin F2α at trial initiation with expected estrus behaviors to be exhibited within the following 56 h. Continuous video was captured in 4k resolution using a five-camera closed caption media recording system strategically placed in the pen corners and over the feed bunk alleyway to capture multiple perspectives. Video data extrapolated into frames and manual annotation was utilized to generate the machine learning algorithm’s training data. Frames were randomly selected from a scattering of 56 h of video across all five camera viewpoints and down sampled to 384 x 640 p for both layers. A two-layer approach was employed utilizing an algorithm for each aspect of the study with the first model detecting the behavior, and the second model using the detected region of interest to identify specific cows. The first layer was trained on 682 frames and validated against 123 manually annotated frames. The second layer was trained on 102 frames and validated against 16 frames manually annotated for individual animals. Results demonstrated that the model used for mounting behavior identification has high precision (0.819) and recall (0.89), achieving a 0.926 mAP at an IoU of 0.5. This model achieved an F1 score of 0.86 at a confidence threshold of 0.268. The results demonstrated that the model used for individual animal identification across different classes achieved high precision (0.861) and recall (0.842), with a mAP of 0.897 at an IoU of 0.5. The F1 score for all classes was 0.89 at a confidence threshold of 0.788. In conclusion, these promising findings underscore the potential of computer vision and machine learning technologies to revolutionize dairy cow management practices. This innovative approach not only has potential to enhance operational efficiency, but also contributes significantly to improved reproductive management, thereby advancing precision and productivity in dairy operations.

Performance of a UHF RFID Detection System to Assess Activity Levels and Lying Behaviour in Fattening Bulls

Animal welfare strongly influences the health and performance of cattle and is an important factor for consumer acceptance. One parameter for the quantification of health status is the lying duration, which can be deployed for the early detection of possible production-related illnesses. Usually, 3D-accelerometers are the tool to detect lying duration in cattle, but the handling of bulls sometimes has special requirements because frequent manipulation in daily farming routines is often not possible. An ultrahigh-frequency (UHF) radio-frequency identification (RFID) system was installed in a beef cattle barn in Germany to measure the activity and lying time of bulls. Such UHF RFID systems are typically used for estrus detection in dairy cows via activity level, but can also be considered, for instance, as an early detection for lameness or other diseases. The aim of the study was to determine whether the estimations of activity level and lying duration can also be traced in husbandry systems for fattening bulls. Two groups of bulls (Uckermärker cattle, n = 10 and n = 13) of the same age were equipped with passive UHF RFID ear transponders. Three cameras were installed to proof the system and to observe the behaviour of the animals (standing, lying, and moving). Furthermore, accelerometers were attached to the hind legs of the bulls to validate their activity and lying durations measured by the RFID system in the recorded area. Over a period of 20 days, position (UHF RFID) and accelerometer data were recorded. Videos were recorded over a period of five days. The UHF RFID system showed an overall specificity of 95.9%, a sensitivity of 97.05%, and an accuracy of 98.45%. However, the comparison of the RFID and accelerometer data revealed residuals (ԑ) of median lying time (in minutes per day) for each group of ԑGroup1 = 51.78 min/d (p < 0.001), ԑGroup2 = −120.63 min/d (p < 0.001), and ԑGroup1+2 = −34.43 min/d (p < 0.001). In conclusion, UHF RFID systems can provide reliable activity and lying durations in 60 min intervals, but accelerometer data are more accurate.

Discrimination strategy using machine learning technique for oestrus detection in dairy cows by a dual-channel-based acoustic tag

Timely identification of cows in heat is a fundamental issue in dairy farming. Although current studies have confirmed that several cow vocalisation characteristics can be used for oestrus identification, the reliability and practicability of oestrus detection methods are limited due to background noise, a lack of data support for sound feature selection, and the unsatisfactory effectiveness of vocalisation recognition algorithms.To overcome these predicaments, a dual-channel recording device and a sound event extraction method were developed in this study. The experimental results in respect of 62 Holstein dairy cows showed that the sound sample recording accuracy was 80.4%, and the noise filtering accuracy was 94.3%, indicating the effectiveness of the vocalisation extraction algorithm in a noisy farm environment. Moreover, two unique sound features with remarkable discrimination ability, i.e., the number of consecutive vocalisations and the maximum consecutive times, were used to improve the effectiveness of oestrus detection. The Friedman test, the Spearman rank order correlation coefficient, and the Kruskal-Wallis test were used to obtain the combination of features with the most optimal discrimination. Subsequently, a dual-LSTM (Long Short-Term Memory) joint discriminant strategy based on optimal combinations was proposed to promote oestrus detection performance. The results of the blind test on the vocalisation data of 20 oestrus cows and 11 non-oestrus cows demonstrated that the oestrus detection rate of the discrimination strategy reached 100%, and had a temporal advantage over the activity-index-based method in determining the onset of oestrus.The improvements of the selection and-based oestrus detection method mainly included accurately extracting individual cow vocalisations, supplementing the explicability of feature selection, and enhancing the refinement of oestrus recognition. Meanwhile, it is of positive significance for applying sound acquisition devices in oestrus detection. Furthermore, combining sound identification with other automated detection techniques appears to be a promising method for promoting the oestrus detection rate.

Identification of oestrus cows based on vocalisation characteristics and machine learning technique using a dual-channel-equipped acoustic tag

Timely and accurate detection of oestrus in cows is an essential element of the good management of dairy farms. At present, the detection of cows in oestrus by acoustic means is impeded by the problems of filtering, incomplete feature selection, and poor recognition accuracy. To overcome these difficulties, this study proposes a sound detection method for cows in oestrus based on machine learning technology using an optimal feature combination and an optimal time window. Firstly, a dual-channel sound detection tag consisting of a unidirectional microphone and an omnidirectional microphone (OM) was developed. The Least Mean Squares adaptive algorithm based on wavelet thresholds was used to filter the signals from the OM, and the dual-channel endpoint detection algorithm was used to identify the lowing of individual cows. The Friedman analysis was then used to select the sound features with significant differences before and after oestrus in terms of time, frequency, and cepstrum, and these were used to determine the most acceptable feature combination. We then analysed the effects of Back Propagation Neural Network (BPNN), Cartesian Regression Tree, Support Vector Machine, and Random Forest classification on the accuracy, precision, sensitivity, specificity, and F1 score of oestrus discrimination. Different time windows were used, and the discrimination performance of these algorithms was evaluated using the Area under Receiver Operating Characteristic Curve to find the most satisfactory match between the time window and the recognition algorithm. The dual-channel acoustic tag’s accuracy, precision, sensitivity, and specificity results were 91.25, 98.83, 91.75, and 83.68%, respectively. BPNN with the 70 ms time window and the feature combination (spectral roll-off + spectral flatness + Mel-Frequency Cepstrum Coefficients) was confirmed as the most suitable oestrus recognition method. The average accuracy, precision, sensitivity, specificity, and F1 score of this method were 97.62, 98.07, 97.17, 97.19, and 97.63%, respectively. Based on these results, the approach was shown to be a feasible means of oestrus detection in dairy cows. Based on its ability to differentiate cows and its consistency, it was demonstrated that sound has the potential to replace accelerometers as an early indicator of oestrus in dairy cows.

Oestrus detection in dairy cows by using atrous spatial pyramid and attention mechanism

The mounting behaviour of dairy cows is driven by oestrus and is also related to their welfare. To improve the accuracy and speed of the detection of mounting of dairy cows in dense scenes, a detection model for the mounting behaviour of dairy cows was developed in this study based on the feature enhancement module. Firstly, an improved attention module (C3GC-3) based on global context information and convolution was proposed to capture long-distance dependence. Then, a feature enhancement module based on atrous spatial pyramid pooling and the C3GC-3 was designed to realise the multiscale fusion of high-level semantic information and improve the feature extraction ability of the model for dense dairy cows images. Finally, the K-means clustering algorithm was used to obtain new anchors for the mounting behaviour of dairy cows, and CIoU was used to optimise the loss function. A camera was installed on a farm containing 200 dairy cows for data collection. The mosaic method was used to expand the 2668 images of the mounting behaviour of dairy cows to train the model, and the remaining 675 images were used to test the model. The experimental results showed that the proposed model had a high inference speed of 61 fps, and the detection time for each image was 16.3 ms, which met the real-time performance requirements for detecting mounting in dairy cows. The mean average precision of the model was 94.3%, which was 5.9% higher than that of YOLOv5l. The proposed model showed promising results regarding the mounting of dairy cows and might be used in any weather. • A mounting behaviour dataset of group-raised dairy cows in dense scenes was proposed.. • ASPP can improve the accuracy of multi-scale object detection. • Attention mechanism can improve the feature extraction ability of the model. • Mean average precision model was 94.3%, which is 5.9% higher than that of YOLOv5l.

Estrus Detection in Dairy Cows from Location and Acceleration Data using Machine Learning

Accurate and timely detection of estrus, in cows, in dairy farms is very important for reproduction, health and milk production. Traditional estrus detection methods like manual observation and chin head chalking are outdated and not suitable for the dairy farm because of large number of animals. A lot of automated estrus detection methods have been proposed like milk yield fluctuation, milk progesterone detection etc., but they are either too complex to implement or have low detection rate. Whereas, the proposed estrus detection method can be easily implemented, cheaply and accurately. This method uses features extracted from 3D acceleration data, obtained using accelerometer attached to cow’s neck. The data is then clustered using k-means into 3 clusters. Categories are assigned based on data variance. As a result, the three clusters are categorized as: low activity, medium activity, or high activity. Based on this information, activity index is calculated and then it is used for the estrus detection. Machine learning classifiers including SVM, and D-trees are used for the activity recognition. SVM and D-tree demonstrate an accuracy of 96% and 86% respectively.

Potential of an activity index combining acceleration and location for automated estrus detection in dairy cows

The objective of this study was to assess the effectiveness and practicability of an activity index combining acceleration and location data for automated estrus detection in dairy cows. By using a wearable neck tag, measurements of acceleration and location were gathered from 22 multiparous cows monitored incessantly for 6 days to derive activity records of each cow. The maximum-minimum distance clustering (MMDC) method was used to divide hourly activity data into low, medium, high, and intensity level groups. The weighted sum of the proportions of the low, medium, high, and intensity activities in an hour constituted the activity level. The activity index was defined as the ratio of the variation in hourly activity level compared to the same time period during the previous three days. Furthermore, whether the cow was in estrus was judged above a set threshold. The study showed that the power consumption and communication effects of the neck tags were acceptable for indoor-housing conditions. For the two consecutive time periods, the activity-index-based detection algorithm achieved 90.91% for accuracy, 100% for precision, 100% for specificity, 83.33% for recall, 90.91% for F1 score, and 0.82 for Kappa coefficient. On the basis of these results, it can be concluded that the combination of acceleration and location in the activity index can promote estrus detection in dairy cows.

Detection of cow mounting behavior using region geometry and optical flow characteristics

Computer vision technology may improve the accuracy and robustness of estrus detection in dairy cows. In this study, we proposed a method for detecting mounting behavior in dairy cattle using the geometric and optical flow characteristics of identified image regions in videos taken on dairy farms. Videos captured on a farm often have a complex background, which can interfere with target detection. In this study, we used masking technology to remove the unrelated background, converted the RGB color space to HSV color space, and adjusted the summation coefficients of the HSV channels to improve the contrast between the cows and background images. Subsequently, the proposed Background Subtraction with Color and Texture Features (BSCTF) algorithm was used to detect cow regions. Then, to perform inter-frame differential processing on detection regions, the geometric and optical flow characteristics of the regions were extracted, and seven optimized features were used to construct regional feature vectors. Finally, a support vector machine (SVM) classifier was trained to classify the detected regions into mounting regions and non-mounting regions, which allowed the identification of mounting behavior. We obtained accuracy and omission rates of this detection method of 98.3% and 6.4%, respectively. The average recognition accuracy and false positive rates of the SVM classifier were 90.9% and 4.2%, respectively. These results demonstrate that the proposed method is effective for detecting the mounting behavior of dairy cows and is convenient for further judging whether cows are in estrus.

Automated oestrus detection using multimetric behaviour recognition in seasonal-calving dairy cattle on pasture

AIM To evaluate the performance of a novel accelerometer-based oestrus detection system (ODS) for dairy cows on pasture, in comparison with measurement of concentrations of progesterone in milk, ultrasonographic examination of ovaries and farmer observations. METHODS Mixed-breed lactating dairy cows (n=109) in a commercial, seasonal-calving herd managed at pasture under typical farming conditions in Ireland, were fitted with oestrus detection collars 3 weeks prior to mating start date. The ODS performed multimetric analysis of eight different motion patterns to generate oestrus alerts. Data were collected during the artificial insemination period of 66 days, commencing on 16 April 2015. Transrectal ultrasonographic examinations of the reproductive tract and measurements of concentrations of progesterone in milk were used to confirm oestrus events. Visual observations by the farmer and the number of theoretically expected oestrus events were used to evaluate the number of false negative ODS alerts. The percentage of eligible cows that were detected in oestrus at least once (and were confirmed true positives) was calculated for the first 21, 42 and 63 days of the insemination period. RESULTS During the insemination period, the ODS generated 194 oestrus alerts and 140 (72.2%) were confirmed as true positives. Six confirmed oestrus events recognised by the farmer did not generate ODS alerts. The positive predictive value of the ODS was 72.2 (95% CI=65.3–78.4)%. To account for oestrus events not identified by the ODS or the farmer, four theoretical missed oestrus events were added to the false negatives. Estimated sensitivity of the automated ODS was 93.3 (95% CI=88.1–96.8)%. The proportion of eligible cows that were detected in oestrus during the first 21 days of the insemination period was 92/106 (86.8%), and during the first 42 and 63 days of the insemination period was 103/106 (97.2%) and 105/106 (99.1%), respectively. CONCLUSIONS and CLINICAL RELEVANCE The ODS under investigation was suitable for oestrus detection in dairy cows on pasture and showed a high sensitivity of oestrus detection. Multimetric analysis of behavioural data seems to be the superior approach to developing and improving ODS for dairy cows on pasture. Due to a high proportion of false positive alerts, its use as a stand-alone system for oestrus detection cannot be recommended. As it is the first time the system was investigated, testing on other farms would be necessary for further validation.

245 Standardizing infrared thermography (IRT) and micro-behavioral biometrics for estrus detection in dairy cows

245 Standardizing infrared thermography (IRT) and micro-behavioral biometrics for estrus detection in dairy cows

Accuracy and intensity of estrus detection with activity monitoring systems for lactating dairy cows

Activity monitors are becoming a widely used tool for estrus detection in dairy cows. The objective of this study was to assess the intensity and accuracy of estrus detection with automated activity monitoring systems.

Aquagrams of raw milk for oestrus detection in dairy cows.

The purpose of this research was to develop rapid and cost-effective method for oestrus detection in dairy cows by means of near infrared spectroscopy and aquaphotomics, using raw milk from individual cows. We found that aquaphotomics approach showed consistent specific water spectral pattern of milk at the oestrus periods of the investigated Holstein cows. Characteristic changes were detected especially in foremilk collected at morning milking. They were reflected in calculated aquagrams of milk spectra where distinctive spectral pattern of oestrus showed increased light absorbance of strongly hydrogen-bonded water. Results showed that monitoring of raw milk near infrared spectra provides an opportunity for analysing hormone levels indirectly, through the changes of water spectral pattern caused by complex physiological changes related to fertile periods.

Estrus Detection in Dairy Cows from Acceleration Data using Self-learning Classification Models

Automatic estrus detection techniques in dairy cows have been present by different traits. Pedometers and accelerators are the most common sensor equipment. Most of the detection methods are associated with the supervised classification technique, which the training set becomes a crucial reference. The training set obtained by visual observation is subjective and time consuming. Another limitation of this approach is that it usually does not consider the factors affecting successful alerts, such as the discriminative figure, activity type of cows, the location and direction of the sensor node placed on the neck collar of a cow. This paper presents a novel estrus detection method that uses k-means clustering algorithm to create the training set online for each cow. And the training set is finally used to build an activity classification model by SVM. The activity index counted by the classification results in each sampling period can measure cow’s activity variation for assessing the onset of estrus. The experimental results indicate that the peak of estrus time are higher than that of non-estrus time at least twice in the activity index curve, and it can enhance the sensitivity and significantly reduce the error rate.

Dairy cows continuously-housed in tie-stalls failed to manifest activity changes during estrus

Felton, C. A., Colazo, M. G., Ponce-Barajas, P., Bench, C. J. and Ambrose, D. J. 2012. Dairy cows continuously-housed in tie-stalls failed to manifest activity changes during estrus. Can. J. Anim. Sci. 92: 189–196. While pedometers are effective for estrus detection in free-stall herds, their efficacy in tie-stalls is not established. The objectives were to determine if pedometers equipped with a behavior tag for monitoring lying activity (AfiMilk Pedometer Plus Tag system) can accurately record cow activity, and if they can facilitate estrus detection in dairy cows continuously housed in tie-stalls. Nineteen lactating Holstein cows were subjected to estrus induction with prostaglandin F2α, and 18 pregnant cows served as controls. Activity was monitored for 10 d using pedometers and 24-h video recording. Pedometer and video records were compared for 6 d surrounding estrus; stepping (rc=0.85; P<0.01), lying time (rc=0.94; P<0.01), and lying bouts (rc=0.89; P<0.01) were highly correlated. Progesterone concentrations confirmed that 17 of the 19 cows responded to estrus induction; 16 of those 17 cows ovulated. Changes in mean daily activity did not differ (P>0.05) between estrus-induced cows and controls, or in the estrus-induced cows in the 48 to 72 h preceding the day of ovulation. In conclusion, while the pedometers accurately recorded cow activity in lactating cows housed in tie-stalls, they could not facilitate estrus detection because cow activity did not significantly change during estrus.

Dairy cows continuously-housed in tie-stalls failed to manifest activity changes during estrus

Felton, C. A., Colazo, M. G., Ponce-Barajas, P., Bench, C. J. and Ambrose, D. J. 2012. Dairy cows continuously-housed in tie-stalls failed to manifest activity changes during estrus. Can. J. Anim. Sci. 92: 189-196. While pedometers are effective for estrus detection in free-stall herds, their efficacy in tie-stalls is not established. The objectives were to determine if pedometers equipped with a behavior tag for monitoring lying activity (AfiMilk Pedometer Plus Tag system) can accurately record cow activity, and if they can facilitate estrus detection in dairy cows continuously housed in tie-stalls. Nineteen lactating Holstein cows were subjected to estrus induction with prostaglandin F2a, and 18 pregnant cows served as controls. Activity was monitored for 10 d using pedometers and 24-h video recording. Pedometer and video records were compared for 6 d surrounding estrus; stepping (rc=0.85; P<0.01), lying time (rc=0.94; P<0.01), and lying bouts (rc=0.89; P<0.01) were highly correlated. Progesterone conce...

Effects of management and health on the use of activity monitoring for estrus detection in dairy cows

The aim was to investigate 1) the relationship between the physical activity index created for each cow by activity monitoring systems and the identification of the preovulatory follicular phase, and 2) the influence of various production, health, and cow factors on the relationship between physical activity and estrous behavior. Eighty-nine spring calving cows, on pasture, were monitored during the breeding season using the neck-mounted estrous activity monitor Heatime (SCR Engineers Ltd., Netanya, Israel). Milk samples were collected twice weekly for progesterone assay to characterize resumption of reproductive activity. Reproductive tract health was assessed weekly by ultrasonography and vaginal mucus scoring. Body condition score and milk yield were assessed every 2 wk. Heatime identified 72% of preovulatory follicular phases from which 145 inseminations resulted in 69 conceptions; 32% of activity clusters were associated with high-progesterone states (i.e., false positives). By inclusion of a 6 to 8-h duration threshold and maintaining the borderline peak activity threshold, this was improved to 87.5% with 21.3% false positives. Mean (± standard error of the mean) peak activity and cluster duration (19.3±0.53 and 10.8±0.38, respectively) were highest for the second or subsequent preovulatory follicular phases followed, in descending order, by those during first preovulatory follicular phases (14.8±2.13 and 8.4±1.4, respectively) and high progesterone states (8.0±0.47 and 3.0±0.42, respectively). The odds of an activity cluster being in a preovulatory follicular phase rather than a high-progesterone phase improved by 29% for every 1-unit increase in peak activity and by 91% for every 2-h increase in duration. The probability of an activity cluster detecting a preovulatory follicular phase was improved if it was a second or subsequent follicular phase, if body condition score was 0.25 units higher, if milk yield was 10kg lower, and uterine infection was absent. Conception rate was influenced by insemination on the same day (52%) or day after a cluster (34.3%); inseminations were carried out using the a.m.-p.m. rule. Advances in the development of more accurate automatic monitoring of the preovulatory follicular phase will aid the timing of insemination and, thus, improve conception rates.

Oestrus detection in dairy cows from activity and lying data using on-line individual models

Automated monitoring and detection of oestrus in dairy cows is attractive for reasons of economy in dairy farming. While high performance detection has been shown possible using high-priced progesterone measurements, detection results were less reliable when only low-cost sensor data were available. Aiming at improving detection scheme reliability with the use of low-cost sensor data, this study combines information from step count and leg tilt sensors. Introducing a lying balance for the individual animal, a novel change detection scheme is derived from observed distributions of the step count data and the lying balance. Detection and hypothesis testing are based on generalised likelihood ratio optimisation combined with time-wise joint probability windowing based on the duration of oestrus and oestrus intervals. It is shown to be essential that cow-specific parameters and test statistics are derived on-line from data to cope with behaviours of individuals. Performance is validated on 18 sequences of data where definite proof of prior oestrus was available in form of subsequent pregnancy. These data were extracted from data sequences from 44 dairy cows over an 8 months period. The results show sensitivity 88.9% and error rate 5.9.%, which is very satisfactory when only cheap sensor data are used.

The Use of Pedometry for Estrus Detection in Dairy Cows in Israel

The objectives of this review are to describe the use of pedometry on commercial dairy farms in Israel, to evaluate its efficiency in heat detection and to describe a clinical trial comparing between pedometry and the Ovsynch technique. Pedometry is the major tool of heat detection on most farms in Israel today. On many farms automated electronic pedometry is the sole mean of heat detection. Production and reproduction parameters are monitored by Hachaklait Veterinary Services Ltd. Results are compared with the farm history and with national means and goals. The average herd rate of undetected heat in more than 120 herds recorded between 2004 and 2008 has increased from 30.3 to 38.9% in primiparous cows and from 33.9 to 43.9% in multiparous cows respectively. The average duration of the waiting period has dropped from 106.2 to 93.4 days and from 99.9 to 87.3 days in primiparous and multiparous cows respectively. The average annual rate of cows shown on heat by pedometry and not inseminated by the A.I. technicians in recent years was 13% and remained steady. The mean herd rate of normal length heat cycles (18-24 days) in multiparous cows in recent years has been fairly steady; 57.4 to 58.4% of all cycles detected in 2004 and 2008 respectively. Herd rate of double cycles is also used to estimate the sensitivity of pedometry systems. Average rate of double cycles (37-60 days) per herd per annum in multiparous cows has dropped steadily from 22.6 to 20.1% between 2004 and 2008 respectively. Mean herd rate of short cycles in multiparous cows in 2008 was 7.4% Pregnancy checks are performed by the herds' veterinarians by transrectal palpation from 40 days post A.I. onward. In 2008 the average herd rate of negative pregnancy checks in cows was 27.5% (9.5-53.4%). The wide range indicates a large variability of pedometry system in tracking non-conceived cycling cows. Beside activity and heat detection pedometry systems record other individual cow parameters. A controlled field trial comparing between pedometry, observation and Ovsynch did not yield significant differences in reproductive measures.

On the use of physical activity monitoring for estrus detection in dairy cows

Detection of estrus in dairy cattle is effectively aided by electronic activity tags or pedometers. Characterization of estrus intensity and duration is also possible from activity data. This study aimed to develop an algorithm to detect and characterize behavioral estrus from hourly recorded activity data and to apply the algorithm to activity data from an experimental herd. The herd comprised of Holstein (n=211), Jersey (n=126), and Red Dane (n=178) cattle, with virgin heifers (n=132) and lactating cows in the first 4 parities; n=895 cow-parities, with a total of 3,674 activity episodes. The algorithm was based on deviations from exponentially smoothed hourly activity counts and was used to identify onset, duration, and intensity of estrus. Learning data included 461 successful inseminations with activity records over a 2-wk period before and after the artificial insemination. Rates of estrus detection and error rate depended on the chosen threshold level. At a threshold giving 74.6% detection rate, daily error rate was 1.3%. When applied to a subset of the complete data where milk progesterone was also available, concordance of days to first activity-detected estrus with the similar trait based on progesterone was also dependent on the chosen threshold so that, with stricter thresholds, the agreement was closer. A single-trait mixed model was used to determine the effects of systematic factors on the estrus activity traits. In general, an activity episode lasted 9.24h in heifers and 8.12h in cows, with the average strength of 1.03 ln units (equivalent to a 2.8-fold increase) in both age groups. Red Danes had significantly fewer days to first episode of high activity than Holsteins and Jerseys (29.4, 33.1, and 33.9 d, respectively). However, Jerseys had significantly shorter duration and less strength of estrus than both Red Danes and Holsteins of comparable age. The random effect of cow affected days to first episode of high activity and strength as well as estrus duration. Days from calving to first episode of high activity correlated negatively with body condition scores in early lactation. The results suggest that data from activity monitors could supply valuable information about fertility traits and could thereby be helpful in management of herd fertility. To establish the complementarities or interdependence between progesterone and activity measurements, further studies with more information from different sources of measuring estrus are needed.

Oestrus Detection in Dairy Cows Using Likelihood Ratio Tests

This paper addresses detection of oestrus in dairy cows using methods from statistical change detection. The activity of the cows was measured by a necklace attached sensor. Statistical properties of the activity measure were investigated. Using data sets from 17 cows, diurnal activity variations were identified for the ensemble and for the individual cows. A diurnal filter was adapted to remove the daily variation of the individual. Change detection algorithms were designed for the actual probability densities, which were Rayleigh distributed with individual parameters for each cow. A generalized likelihood ratio algorithm was derived for the compensated activity signal and detection algorithm was tested on 2323 days of activity, which contained 42 oestruses on 12 cows in total. The application of statistical change detection methods is a new approach for detecting oestrus in dairy cows and the results are shown to outperform earlier approaches in respect to combined statistics of false alarms and missed detections.