Foreground detection of group-housed pigs based on the combination of Mixture of Gaussians using prediction mechanism and threshold segmentation

• Foreground detection method for overhead views of group-housed pigs.
• Mixture of Gaussians using prediction mechanism and threshold segmentation.
• Deals with complex scenes having light changes, slow movements and varying colours.

In this paper, a foreground detection method to obtain the foreground objects of pigs in overhead views of group-housed environments is proposed. The method is based on the combination of Mixture of Gaussians (MoG) using prediction mechanism (PM) and threshold segmentation algorithm. First, the “valid region” is manually set according to a priori knowledge. Second, the foreground objects of pigs are detected using the PM-MoG algorithm. The algorithm uses the detected binary image of the previous frame to predict the current frame in the valid region for pixels that fulfil background updating conditions. Different update strategies are used to update the background for different circumstances. Third, the maximum entropy threshold segmentation algorithm is used according to the colour information of foreground objects. Finally, the results of the two previous steps of foreground detection are fused. The experimental results show that the method is effective and can extract relatively complete foreground objects of pigs in complex scenes. These complex scenes include light changes, the influence of ground urine stains, water stains, manure, and other sundries, pigs' slow movement patterns, and varying colours of foreground objects. The average foreground detection rate is approximately 92%. The experimental results set the foundation for further exploration of individual identification of group-housed pigs, their behaviour analysis, and other objectives.