Scholarlink

In the agricultural environment, hawks and other birds of prey that are hunting livestock (e.g., chickens) are critical issues. To address the problems with pasture-raised livestock, the detection mechanism of predators is essential. We focus on hawks as major predators, but the availability of hawks’ dataset is very limited. In this paper, we attempt to generate more datasets using data augmentation techniques with several parameters. We also design deep-learning networks that can effectively detect airborne predators to protect livestock. The network has been optimized to improve its performance. We analyze the experimental results and provide implications. The detection scheme can be utilized in the chicken industry to allow corporations to increase animal welfare by raising chickens in an outdoor pasture environment.

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

Implications for Designing Hawks Detection with Data Augmentation and Network Optimizations

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Adam Smith,
  • Judah Small,
  • Byeong Kil Lee

摘要

In the agricultural environment, hawks and other birds of prey that are hunting livestock (e.g., chickens) are critical issues. To address the problems with pasture-raised livestock, the detection mechanism of predators is essential. We focus on hawks as major predators, but the availability of hawks’ dataset is very limited. In this paper, we attempt to generate more datasets using data augmentation techniques with several parameters. We also design deep-learning networks that can effectively detect airborne predators to protect livestock. The network has been optimized to improve its performance. We analyze the experimental results and provide implications. The detection scheme can be utilized in the chicken industry to allow corporations to increase animal welfare by raising chickens in an outdoor pasture environment.