Scholarlink

<p>Early detection of lambing is essential for improving animal welfare and farm management, as it enables timely intervention and reduces complications. Wearable inertial sensors have been applied to sheep monitoring, with frequent transitions between standing and lying identified as key behavioral indicators of lambing. However, unlike in larger livestock, no accelerometry-based system currently provides real-time detection for small ruminants, and existing studies remain limited to preliminary approaches. This study monitored 61 ewes using accelerometers sampling at 20 Hz, while lambing was simultaneously recorded on video to establish precise birth times for 113 events. Video analysis also documented litter size and the need for assistance. Data were organized per ewe, supplemented with information such as birth year, previous lambing records, and ultrasound results. A video of one birth was included to illustrate behavior during the process. The dataset provides a valuable foundation for developing algorithms capable of classifying birth-related behaviors, thereby supporting future automated lambing detection systems.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Tjotta accelerometer monitored lambing dataset

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年2月11日

查看全文
  • Pedro Goncalves,
  • Shelemia Nyamuryekung’e,
  • Gustavo Corrente,
  • Grete Helen Meisfjord Jørgensen

摘要

Early detection of lambing is essential for improving animal welfare and farm management, as it enables timely intervention and reduces complications. Wearable inertial sensors have been applied to sheep monitoring, with frequent transitions between standing and lying identified as key behavioral indicators of lambing. However, unlike in larger livestock, no accelerometry-based system currently provides real-time detection for small ruminants, and existing studies remain limited to preliminary approaches. This study monitored 61 ewes using accelerometers sampling at 20 Hz, while lambing was simultaneously recorded on video to establish precise birth times for 113 events. Video analysis also documented litter size and the need for assistance. Data were organized per ewe, supplemented with information such as birth year, previous lambing records, and ultrasound results. A video of one birth was included to illustrate behavior during the process. The dataset provides a valuable foundation for developing algorithms capable of classifying birth-related behaviors, thereby supporting future automated lambing detection systems.