Processamento de imagens: metodologia para o estudo da distribuição espacial de aves poedeiras.

Abstract

Automatized systems for animal behaviour evaluation in scientific experiments are an interesting challenge for the areas of Computer Vision and animal science precision. The extracted information of the images can be valuable for understanding of determined events. The processing and image analysis of confined animals use techniques of segmentation to study the spatial distribution in interest regions. The fact that the animal is in a definitive place can be reflect of the search for a comfortable environment. This work analyzed the distribution of Laying hens (Hy-Line W36) in stress and comfort conditions at a climatic chamber in the region of the water through. It is a simple application of the techniques and the central objective is to develop a methodology for established robust systems of image analysis. The technique of direct observation of the drink behaviour (quantitative) and the analysis of image that evaluates instanteneal data of amount of birds in the region of water through were compared. It was verified that there is relation between the number of birds in the region of the water through and the drinking behaviour, for the comfort condition. However, when it deals with stress condition, this relation is not found, therefore becomes greater the frequency in this region.