Towards the Estimation of Body Weight in Sheep Using Metaheuristic Algorithms from Biometric Parameters in...

Area of Science:

• Agricultural Science
• Computer Vision
• Mathematical Modeling

Background:

• Sheep body weight (BW) is crucial for genetic management, nutrition, and health monitoring.
• Accurate BW estimation aids producers in optimizing livestock management practices.

Purpose of the Study:

• To develop and validate a polynomial model for estimating sheep BW using biometric parameters.
• To compare the performance of Genetic Algorithm (GA) and Cuckoo Search Algorithm (CSA) in optimizing the model.

Main Methods:

• Utilized computer vision to measure sheep biometric parameters with <5% error.
• Developed a polynomial model with adjustable degree and metaheuristic algorithms (GA, CSA) to estimate BW.
• Evaluated model performance using Root-Mean-Squared Error (RMSE).

Main Results:

• The Cuckoo Search Algorithm (CSA) achieved a lower RMSE (7.55%) compared to the Genetic Algorithm (GA) (7.68%).
• The polynomial model demonstrated feasibility for accurate sheep BW estimation.
• Biometric parameter estimation and the mathematical model are adaptable for embedded microsystems.

Conclusions:

• The proposed polynomial model, optimized by metaheuristic algorithms, provides a viable method for estimating sheep body weight.
• Computer vision-based biometric measurements offer a practical approach for livestock monitoring.
• The system's adaptability to embedded microsystems suggests potential for real-time applications in animal agriculture.