Counting cattle in UAV images: dealing with clustered animals and animal/background contrast changes.
Abstract: The management of livestock in extensive production systems may be challenging, especially in large areas. Using Unmanned Aerial Vehicles (UAVs) to collect images from the area of interest is quickly becoming a viable alternative, but suitable algorithms for extraction of relevant information from the images are still rare. This article proposes a method for counting cattle which combines a deep learning model for rough animal location, color space manipulation to increase contrast between animals and background, mathematical morphology to isolate the animals and infer the number of individuals in clustered groups, and image matching to take into account image overlap. Using Nelore and Canchim breeds as a case study, the proposed approach yields accuracies over 90% under a wide variety of conditions and backgrounds.
| Main Authors: | , , , |
|---|---|
| Other Authors: | |
| Format: | Artigo de periódico biblioteca |
| Language: | English eng |
| Published: |
2020-04-15
|
| Subjects: | Redes neurais, Rede neural convolucional, Veículo aéreo não tripulado, Canchim breed, Nelore breed, Convolutional neural networks, Mathematical morphology, Deep learning mode, Gado de Corte, Gado Nelore, Gado Canchim, Unmanned aerial vehicles, Neural networks, |
| Online Access: | http://www.alice.cnptia.embrapa.br/alice/handle/doc/1121664 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Abstract: The management of livestock in extensive production systems may be challenging, especially in large areas. Using Unmanned Aerial Vehicles (UAVs) to collect images from the area of interest is quickly becoming a viable alternative, but suitable algorithms for extraction of relevant information from the images are still rare. This article proposes a method for counting cattle which combines a deep learning model for rough animal location, color space manipulation to increase contrast between animals and background, mathematical morphology to isolate the animals and infer the number of individuals in clustered groups, and image matching to take into account image overlap. Using Nelore and Canchim breeds as a case study, the proposed approach yields accuracies over 90% under a wide variety of conditions and backgrounds. |
|---|