一种禽类异常识别方法及装置
By extracting image and audio information from the poultry farm monitoring system, abnormal poultry behavior can be automatically identified, solving the problem of high manpower consumption in manual identification in existing technologies and achieving efficient and accurate monitoring of poultry anomalies.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- NANJING FORESTRY UNIV
- Filing Date
- 2018-12-27
- Publication Date
- 2026-07-17
AI Technical Summary
Existing poultry farm monitoring systems cannot automatically identify abnormal poultry behavior, requiring a significant amount of manpower.
By extracting image information from surveillance videos, selecting reference images, calculating the area difference between the image area not covered by feed and the background image area in the feed trough, determining the time of maximum value as the end time of feed addition, and combining audio information for machine learning classification, it is possible to determine whether poultry feeding is abnormal.
It has automated the identification of abnormal poultry behavior, improved identification efficiency and accuracy, reduced human intervention, and can distinguish between non-urgent and urgent abnormal behaviors.
Smart Images
Figure CN109711346B_ABST