Camera-based detection, measurement and classification
of pigs in a slaughterhouse

In a slaughterhouse, images of slaughtered pigs are to be recorded and archived. For recordings at different positions in the process, we have developed customized recording units and installed the solutions in protective housings, suitable for the harsh environment of a slaughterhouse.

First, the pigs lying on a belt are passing under a recording unit and the camera counts the animals reliably, even if they are very different in position or touching each other. As an expansion option, gender or pig species can also be distinguished by means of AI (artificial intelligence).

Subsequently, the pigs, suspended by their hind legs and passing a wall at constant distances, are examined with another recording system. Based on contrast differences in previously defined image areas, the camera detects when an animal is in front of the background and the system captures an image of each passing pig, which is then stored with a time stamp. The camera continuously sends images to a compact PC in the server room, which evaluates the permanent image stream.

The camera measures the width, thickness and height of the slaughtered animals, and AI-based differentiation is also made according to gender characteristics. If higher accuracies are required, a laser triangulation camera is used, which accurately measures the surface of the pig in 3D and creates a topography.

An extension of the solution could be the automated recording and documentation of animal welfare indicators. For example, body part changes such as ear rim or tail can be recorded and compared to specifications.

Camera-based examinations of individual organs and, of course, the recording of other slaughter animal species can also be implemented according to your individual requirements.