The Robot Fusion Mapping Method Applied in Indoor Poultry Farming Environment

Temperature and humidity fluctuations in indoor poultry settings significantly impact poultry survival, influencing farm efficiency. Employing robots for inspections equipped with specialized tools can enhance inspection efficiency by replacing manual labor. Yet, indoor environments pose challenges like odors, cleanliness issues, and inspection complexities. To address these, we propose a novel method-Visual-Assisted G2 Laser Robot Mapping (VAGL)-tailored for indoor poultry farming. VAGL integrates laser and visual mapping to cover structured and unstructured areas effectively. Laser mapping handles structured zones, overcoming limitations of single-line LiDAR by converting depth images into laser data for detecting obstacles at varying levels. For unstructured areas, like corners with railings, we utilize the visual ORB- SLAM mapping method. Employing Bayesian fusion enhances mapping accuracy and completeness by integrating laser and visual mapping data. Simulation tests using MATLAB's image correction tools reveal significantly improved mapping accuracy and efficiency compared to using only laser or visual mapping methods. This study presents a promising technical solution for large-scale indoor poultry farming, potentially boosting environmental monitoring and management precision and efficiency in farming practices.