Automated rating of welfare indicators for pigs in the slaughterhouse – a pilot study

In Germany farmers are legally obligated to monitor animal welfare indicators, and based on that, improve their conditions. Besides the difficulties in defining indicators, the comparable assessment is another challenge. Therefore in the poultry industry, camera systems have been installed in slaughterhouses to evaluate the foot pad condition with a software. The results are open to farmers as feedback and partly to veterinary departments. Such a system has not been built up in pig slaughterhouses yet, mainly due to the more complex anatomy and the lack of suitable defined indicators.

Therefore, the aim of this study is to develop a commercial camera system for the assessment of animal welfare indicators for pigs in the slaughterhouse.

In this pilot study the focus was on external indicators, which can be recorded directly in the slaughterhouse with a camera system (CLK GmbH, Münster, Germany). The system was installed in a German slaughterhouse slaughtering 3,500 – 5,000 pigs per day. It was located after the cleaning of the pigs. The system included five cameras and a special one for the 3D pictures, lights, triggers, a blue background. It took eight pictures of every pig (head, back, hind quarter, both lateral views, three 3D pictures of the joints of the hind legs).

These pictures were interpreted by a veterinarian, discussed in a consortium, and used by the manufacturer to develop a program for an automated evaluation. In the next step, the automated results are controlled by an individual observer.

Ear necrosis, tail damages, wounds from chasing and tattooing, and swellings of the joints of front and hind limbs were defined as indicators in this project. These indicators are in agreement with the German Quality Assurance Initiative (QS GmbH, Bonn, Germany).

Between the 11-09-15 and 05-01-16 approximately 750,000 pictures were taken. Until June, 2016 presumably 4,400,000 new pictures will be taken. These pictures are analysed and categorized critically.

Hundreds of pictures of indicators and differential diagnosis were found yet. Based on these, a software is currently developed. In cooperation of veterinarians and the manufacturer, the pictures were optimized. The correct classification of every pig was guaranteed. Moreover, pictures were taken manually in the stables to compare the results of the camera and the original condition.

First steps for an automated rating system of animal welfare indicators for pigs in the slaughterhouse have been made. Resulting in a powerful tool to support the objective monitoring of animal welfare, also in large animal numbers, it will lead to a better, comparable feedback both to the farmer and the veterinarian.

Disclosure of Interest: None Declared.