Search

The potential of using computer vision to estimate the weight of pigs has been demonstrated in several studies. In this paper, we propose a Vision Analysis and Prediction (VAP) system, which consisted of two main parts: (1) computer vision – a vision algorithm was developed to segment the pigs and identify which pig area could be used to estimate the corresponding weight; (2) statistical analysis – a statistical method for predicting pig weight using the area was presented, including BLUP model and linear regression as well as a prediction function. The results showed that our approach has successfully estimated the weight of growing pigs (approximate ranges from 20 to 105 kg) with an accuracy of weight estimation of 97.76% on average for the predicted mean weight at pen level. The proposed system does not require modifications to the building and shows a strong potential to be utilized in pig weight estimation in slaughter pens. • A VAP system has great potential for pig weight estimation in slaughter pens. • The VAP includes two parts: a computer vision algorithm and a predicting model. • The VAP can accurately predict pig weights from 20 to 105 kg in a slaughter pen. • The VAP can help for optimal decision making to improve animal productivity. [ABSTRACT FROM AUTHOR]

Abstract: The aim of the study was to investigate the relationship between the ear skin temperature and the behaviour of pigs. Fifty-four pigs weighing 75±5kg were used in three replications (18 pigs per replication) and housed in pens (six pigs per pen) in a controlled climate facility. The room temperature was changed by 2°C from 18°C down to 10°C and up again to 22°C. The ear skin temperature (EST) was continuously recorded and the activity, lying posture, location and contact with pen mates were scored by 12min scan sampling for 24h at the set point temperatures 18°C, 10°C and 22°C. A diurnal rhythm in the EST, the posture and the lying behaviour was found. The EST was highest at night and lowest in the afternoon. During night the pigs had more physical contact to pen mates than during day time. For all three set point temperatures the predominant lying position during the night was the fully recumbent position. The room temperature affected the lying behaviour and the EST. With decreasing room temperature the pigs increased their contact to pen mates and fewer pigs were observed lying in the fully recumbent position. The EST decreased with decreasing room temperature, and the range in the EST''s at the three set point temperatures was larger during day than night (4°C versus 2°C). The results indicate that pigs adjust their behaviour to a higher EST when resting than when they are active, and they use behavioural adjustment (e.g. increased/decreased contact to pen mates) to bring their skin temperature into a preferred interval. [Copyright &y& Elsevier]

Abstract: This paper analyzes the problem faced by managers of pig fattening units when the pigs are subject to common epidemic diseases, such as respiratory diseases or Swine Influenza. The problem is defined as a simultaneous optimization of the delivery policy and control strategies for disease control. As some disease controls, such as vaccination, are implemented at the batch level and last through the lifetime of the vaccinated pigs, the problem requires that decisions are addressed at multiple timescales. Using the framework of multi-level hierarchic Markov processes a general framework was defined. The elements defining the framework were the decisions at the tactical (batch) level and the operational (daily) level. Decisions at the strategic level, e.g., decisions lasting longer than the length of a batch, were considered as constraints to the lower levels of the decision problem and not optimized. The major components of the general framework were a stochastic growth model, an epidemic model to handle within-batch spread of disease, a transition model to handle the between-batch spread of disease, and link between the growth model and the disease model and a general set of control measures. The model was illustrated using an example with Swine Influenza in an all-in-all-out production system. Results show that the optimal delivery policy and control strategies for disease interact. This suggests that decision support must adopt an integrated approach to modelling the decision complex in pig production in order to represent the full complexity of the system. [Copyright &y& Elsevier]

Rooting behaviour appears to be an important part of the behavioural repertoire in pigs. It is suggested that it is in fact a behavioural need. The present study investigated whether nose-ringed gilts, which are unable to root, can satisfactorily substitute rooting with other exploratory activities. The behaviour of 24 gilts in groups of three, half of them nose-ringed, was studied on undisturbed days as well as on days of exposure to four different tests: (1) delayed feeding; (2) double feeding; (3) removal to an unknown area; (4) removal to an unknown area supplied with a trough with stones. The gilts’ posture and their behaviour were recorded by 1-min scan samplings, whereas pawing, aggression, and comfort behaviour were recorded by all occurrence sampling. Observations took place for 2.5 h in the morning and 2.5 h in the afternoon. On undisturbed days no significant differences were found between ringed and unringed gilts as regards the percentage of scans in which the gilts were active and foraging or exploring. However, the unringed gilts rooted significantly more than the ringed gilts (P<0.01). The kind of foraging and exploratory behaviours that the ringed gilts chose as substitutes for rooting depended on the replicate in question, maybe due to more or less nutrition in the grass. On an undisturbed day, there was no significant difference as regards the number of times ringed and unringed gilts were observed performing any behaviour that could be considered a sign of frustration. During delayed and double feeding, ringed and unringed gilts reacted similarly. When exposed to an unknown area, ringed gilts increased the levels of alertness and comfort behaviour, which might indicate frustration. We conclude that in a stable environment ringed gilts were able to substitute other exploratory behaviours for rooting in a satisfactory way. However, it was clear that the gilts that were able to root did so for a large part of their active time, indicating that the experience obtained by rooting was rewarding. [Copyright &y& Elsevier]

Abstract: The most critical period for the survival of piglets is during the first 2 days of life. Causes of early piglet mortality include reduced vitality due to hypoxia during parturition, hypothermia and lack of adequate colostrum intake. Besides, low vitality piglets may run a larger risk of being crushed by the sow. In the light of the thermoregulatory challenges facing newborn piglets and the possible sow-piglet interactions under loose housing, we investigated whether floor heating around parturition affected early piglet vitality and behaviour related to survival. Twenty-three Landrace×Yorkshire sows of 2nd parity were housed individually in 7.5m2 pens in a climate controlled facility. HEAT sows (n =12) were exposed to pen floor heating (33.5°C) from 12h after onset of nest building and until 48h after birth of 1st piglet, whereas CONT sows (n =11) received no floor heating (21.2°C). The concentration of lactate in umbilical cord blood at birth—an indicator of hypoxia—increased with the birth duration (P <0.001) and with declining piglet weight (P <0.001), with no significant effect of floor heating. After the initial drop in body temperature at birth, floor heating resulted in an earlier recovery of piglet temperatures (P <0.001), i.e. the piglet''s period for experiencing hypothermia after birth was reduced. HEAT piglets also suckled sooner than CONT piglets after the 1st hour post partum (with ratios between HEAT/CONT hazard functions being 2.9–6.4 for latency to suckle in the period 1–3h after birth). Moreover, fewer live-born piglets died during the first 3 days (P =0.047), as well as during the first week in the floor heated litters (mean (S.E.) HEAT 8.7 (2.8)% versus CONT 15.5 (5.2)%, P =0.014). We did not find any effect of floor heating on duration of parturition, inter-birth intervals, litter size, early piglet weight gains, blood glucose and lactate concentrations at birth, heart rate at birth, piglet activity from birth to first suckling or amount of parvovirus antibodies transferred from sow to piglets. In conclusion, floor heating around parturition had no evident effect on the innate piglet vitality, but it had favourable effects on the early recovery of piglet body temperature, latency to first suckle and survival of piglets in the loose house system. [Copyright &y& Elsevier]

Abstract: The aim of the present experiment was to measure pigs’ preferences for four different rooting materials using a set-up where pigs were given the opportunity to consider the relative attractiveness of two different materials simultaneously. Ten pigs had access to work for 50min daily in an operant conditioning experiment using concurrent schedules of reinforcement. During the 50min two different rooting materials were available after completing a fixed number of presses on two different operant panels inside a test pen. Long straw was always used as reinforcement on one panel, while one of four alternative rooting materials (chopped straw, fir branches, peat or long straw) was used as reinforcement on the other panel. The cost of access to the resources varied on both panels according to five pairs of Fixed Ratio schedules (FR8/FR40, FR16/FR32, FR24/FR24, FR32/FR16, FR40/FR8). For each material all pigs went through three test runs of the five FR pairs. Measures of preferences were based both on the generalized matching law (bias) and on the cross-points between the demand curves for the reference material and for the alternative material. Both the bias and the cross-point was significantly affected by type of alternative material in that the cross-point for peat (FR9.0) was significantly lower (P &#60;0.05) and bias (−0.34) significantly higher (P &#60;0.05) than for all other alternative materials (cross-point: branches FR18.3, chopped straw FR23.8, long straw FR29.2; log(bias): branches −0.27, chopped straw −0.01, long straw 0.40). The intercept of the demand curves for the alternative materials were significantly affected (P &#60;0.05) by type of material in that it was higher for peat (1.73±0.14)(lsmeans±std. err.) than for all other materials. The intercept for branches (1.31±0.14) was higher than for chopped straw (0.92±0.14) and long straw (0.54±0.17). The slopes of the alternative demand curves were unaffected by type of material as were the intercepts and slopes of the reference curves. The number of times of pressing the alternative panel in succession before shifting to the reference panel was significantly higher (P &#60;0.001) across all FR''s for peat (11±0.7 at FR24) and branches (9.4±0.7 at FR24) than for chopped straw (8.2±0.6 at FR24) and long straw (7.0±0.6 at FR24). The pigs achieved a new reward sooner following delivery of the previous one (P &#60;0.05) when the alternative material was either peat (1.3±0.13min) or branches (1.3±0.13min) compared to long straw (1.8±0.19min) and chopped straw (1.7±0.13min). All three alternatives were able to substitute completely for long straw, in that the slope of the demand curve of the reference material was not influenced by type of alternative material. In conclusion, both measures of cross-points and bias indicated a preference for peat and branches over chopped straw and long straw. [Copyright &y& Elsevier]