Estrus Detection Using Background Image Subtraction Technique in Tie-Stalled Cows.

Area of Science:

• Animal Science
• Veterinary Medicine
• Agricultural Technology

Background:

• Estrus detection is crucial for efficient cattle reproduction.
• Traditional methods can be labor-intensive and stressful for animals.
• Non-contact monitoring offers a promising alternative for dairy cows.

Purpose of the Study:

• To evaluate the background image subtraction technique for estrus detection in tie-stalled cows.
• To assess the influence of camera viewing angles on detection accuracy.
• To establish a non-contact method for identifying estrus in confined cattle.

Main Methods:

• Webcams captured front, top, and rear views of six cows over ten estrous cycles.
• Background subtraction identified moving objects, with pixel counts analyzed.
• The pixel ratio (recent to past 24-hour movement) was calculated hourly.
• Estrus and ovulation times were confirmed using physical observation and transrectal ultrasonography.

Main Results:

• The mean pixel ratio increased significantly around 48 hours before ovulation, peaking during estrus.
• Detection accuracy was independent of camera shooting direction.
• Front-view video analysis achieved 90% sensitivity and 50% precision in detecting estrus.

Conclusions:

• Background image subtraction is a viable non-contact method for estrus detection in tie-stalled cows.
• The technique offers a practical tool for improving reproductive management in dairy herds.
• Further refinement could enhance precision and applicability in various housing systems.