Sub-window inference: A novel approach for improved sheep counting in high-density aerial images

• Published in: Computers and electronics in agriculture 2024-10, Vol.225, p.109271, Article 109271
• Main Authors: Biggs, David R., Theart, Rensu P., Schreve, Kristiaan
• Format: Article
• Language: English
• Subjects: Automated sheep counting; High-density object counting; Instance segmentation for livestock; Machine learning for animal monitoring; Small object detection; UAV/drone imagery
• ISSN: 0168-1699
• DOI: 10.1016/j.compag.2024.109271

Traditional sheep counting methods are labour-intensive, time-consuming, and potentially disruptive to sheep behaviour. Unmanned aerial vehicles (UAVs) and machine learning (ML) techniques have emerged to address these challenges by automating the process. However, these solutions face difficulties due to low object-to-image-pixel ratios and high object densities in images captured for sheep counting, which can compromise detection and counting accuracy. In this study, we introduce and evaluate a novel approach, sub-window inference, designed to increase the object-to-image-pixel ratios, thereby enhancing the performance of existing object detection and segmentation models. Our method was compared to four other object counting techniques, demonstrating superior performance in terms of a reduced mean absolute error (MAE) of 3.21 sheep and a mean absolute percentage error (MAPE) of 1.27%. Furthermore, our findings indicate that incorporating random cropping data augmentation during model training significantly enhances both detection and counting accuracy. It is important to note that a limitation of sub-window inference is that it does not facilitate real-time count predictions. Overall, our proposed method of sub-window inference reduces the MAE in automated sheep counting techniques involving UAVs and ML, presenting a highly effective solution that benefits sheep farmers. •Low object-to-image-pixel ratios have adverse effects on detection performance.•Smaller objects tend to have lower object-to-image-pixel ratios.•Sub-cropping input images improves object-to-image-pixel ratios and detection results.•Random cropping data augmentation during training improves detection and counting accuracy.