Multi-Pig Part Detection and Association with a Fully-Convolutional Network
Computer vision systems have the potential to provide automated, non-invasive monitoring of livestock animals, however, the lack of public datasets with well-defined targets and evaluation metrics presents a significant challenge for researchers. Consequently, existing solutions often focus on achie...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2019-02-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/19/4/852 |
| _version_ | 1798039181473087488 |
|---|---|
| author | Eric T. Psota Mateusz Mittek Lance C. Pérez Ty Schmidt Benny Mote |
| author_facet | Eric T. Psota Mateusz Mittek Lance C. Pérez Ty Schmidt Benny Mote |
| author_sort | Eric T. Psota |
| collection | DOAJ |
| description | Computer vision systems have the potential to provide automated, non-invasive monitoring of livestock animals, however, the lack of public datasets with well-defined targets and evaluation metrics presents a significant challenge for researchers. Consequently, existing solutions often focus on achieving task-specific objectives using relatively small, private datasets. This work introduces a new dataset and method for instance-level detection of multiple pigs in group-housed environments. The method uses a single fully-convolutional neural network to detect the location and orientation of each animal, where both body part locations and pairwise associations are represented in the image space. Accompanying this method is a new dataset containing 2000 annotated images with 24,842 individually annotated pigs from 17 different locations. The proposed method achieves over 99% precision and over 96% recall when detecting pigs in environments previously seen by the network during training. To evaluate the robustness of the trained network, it is also tested on environments and lighting conditions unseen in the training set, where it achieves 91% precision and 67% recall. The dataset is publicly available for download. |
| first_indexed | 2024-04-11T21:50:21Z |
| format | Article |
| id | doaj.art-426c05e82de4461d8d9bfa34c6399cd8 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-04-11T21:50:21Z |
| publishDate | 2019-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-426c05e82de4461d8d9bfa34c6399cd82022-12-22T04:01:16ZengMDPI AGSensors1424-82202019-02-0119485210.3390/s19040852s19040852Multi-Pig Part Detection and Association with a Fully-Convolutional NetworkEric T. Psota0Mateusz Mittek1Lance C. Pérez2Ty Schmidt3Benny Mote4Department of Electrical and Computer Engineering, University of Nebraska–Lincoln, Lincoln, NE 68505, USADepartment of Electrical and Computer Engineering, University of Nebraska–Lincoln, Lincoln, NE 68505, USADepartment of Electrical and Computer Engineering, University of Nebraska–Lincoln, Lincoln, NE 68505, USADepartment of Animal Science, University of Nebraska–Lincoln, Lincoln, NE 68588, USADepartment of Animal Science, University of Nebraska–Lincoln, Lincoln, NE 68588, USAComputer vision systems have the potential to provide automated, non-invasive monitoring of livestock animals, however, the lack of public datasets with well-defined targets and evaluation metrics presents a significant challenge for researchers. Consequently, existing solutions often focus on achieving task-specific objectives using relatively small, private datasets. This work introduces a new dataset and method for instance-level detection of multiple pigs in group-housed environments. The method uses a single fully-convolutional neural network to detect the location and orientation of each animal, where both body part locations and pairwise associations are represented in the image space. Accompanying this method is a new dataset containing 2000 annotated images with 24,842 individually annotated pigs from 17 different locations. The proposed method achieves over 99% precision and over 96% recall when detecting pigs in environments previously seen by the network during training. To evaluate the robustness of the trained network, it is also tested on environments and lighting conditions unseen in the training set, where it achieves 91% precision and 67% recall. The dataset is publicly available for download.https://www.mdpi.com/1424-8220/19/4/852computer visiondeep learningimage processingpose estimationanimal detectionprecision livestock |
| spellingShingle | Eric T. Psota Mateusz Mittek Lance C. Pérez Ty Schmidt Benny Mote Multi-Pig Part Detection and Association with a Fully-Convolutional Network Sensors computer vision deep learning image processing pose estimation animal detection precision livestock |
| title | Multi-Pig Part Detection and Association with a Fully-Convolutional Network |
| title_full | Multi-Pig Part Detection and Association with a Fully-Convolutional Network |
| title_fullStr | Multi-Pig Part Detection and Association with a Fully-Convolutional Network |
| title_full_unstemmed | Multi-Pig Part Detection and Association with a Fully-Convolutional Network |
| title_short | Multi-Pig Part Detection and Association with a Fully-Convolutional Network |
| title_sort | multi pig part detection and association with a fully convolutional network |
| topic | computer vision deep learning image processing pose estimation animal detection precision livestock |
| url | https://www.mdpi.com/1424-8220/19/4/852 |
| work_keys_str_mv | AT erictpsota multipigpartdetectionandassociationwithafullyconvolutionalnetwork AT mateuszmittek multipigpartdetectionandassociationwithafullyconvolutionalnetwork AT lancecperez multipigpartdetectionandassociationwithafullyconvolutionalnetwork AT tyschmidt multipigpartdetectionandassociationwithafullyconvolutionalnetwork AT bennymote multipigpartdetectionandassociationwithafullyconvolutionalnetwork |