Deciphering Distress: Stress Detection in White Leghorn Birds Through Vocalization Analysis
摘要
The poultry industry plays a crucial role in global food production, and poultry birds’ well-being is paramount. Stress in poultry birds can lead to reduced productivity, compromised health, and economic losses. This research paper explores using sound analysis techniques to detect and classify stress in White Leghorn (WL) poultry birds. Around 12.5 h of vocalization data from healthy WL birds’ data has been collected in normal and stressed conditions. The approach of this paper employs AI and ML techniques for non-intrusive stress assessment, achieving 98.86% accuracy with a hybrid feature set. This paper explores alternative acoustic features such as signal time energy (STE), zero crossing rate (ZCR), and pitch intonation profiles. This study showcases the feasibility of automated stress detection and its potential impact on poultry welfare and management.