25 July 2006 A weighty issue for sheep – Do you want your photo taken?

 

 

Traditionally, farmers have either used the “look and guess” method or have relied on weighing their sheep on scales, an expensive and tedious method that ties up staff.

 

A group from the Agriculture and Life Sciences Division at Lincoln University have reviewed the feasibility of using a new method to estimate the live weight of sheep.

 

The group included Jonathon Burke, a research student, Associate Professor Peter Nuthall and Professor Alan McKinnon.

 

The method reviewed the idea of using an automated video image processing system to estimate the live weight of sheep. This involves taking video images of sheep which are then analysed by computer to provide physical data about the animals. It has the potential to be a relatively inexpensive, non-invasive, practical method monitoring the weight of a flock of sheep over a period of time.

 

An important feature of this proposal is that is has the potential to be carried out remotely, without the need to touch or disturb the animal and allows a large quantity of useful data to be collected that would aid farmers in the successful management of their sheep.

 

The review found that this technique has already been trialled in pigs and salmon.

 

One big challenge, however, that needs to be overcome before this method can be used in sheep is the fact that sheep are covered with wool for most of the year. This obscures the body of the sheep and prevents the capture of true body conformation (although the image of shorn sheep could be taken). Another issue to be overcome is that the physical characteristics of wool are not the same all year round and also growth rate changes throughout the year.

 

If a way to estimate the depth of the wool is found then it may be possible to ‘undress’ the sheep and evaluate its weight using suitable parameters such as sheep breed, initial wool depth or time since last shearing as well as standard measurements taken from databases, such as heart girth, length and condition score to estimate live weight from the video image because it is already known that there is a strong relationship between wool growth and live weight.

 

“While there are, of course, quite a number of practical issues that need to be worked through before these ideas are able to be put into practice our review was valuable because it showed us that a lot of individual work had been done at both the sheep and the computer end. If further research could be funded it would be exciting to watch as researchers work to close the gaps and make this idea come to practical fruition,” said Dr Nuthall.

 

In practice, digital cameras could be set up at sites where sheep congregate, for example, water troughs, or aerial photos could be taken by a camera attached to a helium balloon, or perhaps even a satellite circling the earth, to take a series of photographs for analysis. Computer programmes would then use mathematical models to derive the live weight of the individual sheep.

 

One way of using this technology would involve farmers assessing the condition score of four or five sheep in a mob and then image processing could be used to calculate a sheep’s length and heart girth, and the sheep’s age and condition score could be added into the calculation to estimate the weight more accurately.

 

Another approach that might be worth exploring is using infra-red emissions from the sheep to estimate body weight. Emission levels are likely to be proportional to body weight. Again, however, the insulating quality of the wool would need to be allowed for.

 

“There is nothing surer than as the years move on agriculture production will increasingly use technology to improve efficiency,” said Dr Nuthall.