Accurate cattle age estimation is critical for effective livestock management, impacting breeding, nutrition, and productivity. Traditional methods like tooth inspection and body weight analysis are subjective, require specialized skills, and are often costly. Although DNA testing offers precision, it is impractical for routine use due to high costs and complexity. Hence, developing a cost-effective, non-invasive solution for cattle age estimation is crucial, particularly in rural areas where access to technology is limited. Despite advancements in computer vision and machine learning for livestock health and breed identification, their use in age estimation through dentition images remains underexplored. This study presents a novel approach by leveraging high-resolution images of cattle teeth and canal regions. The YOLOv9 model is used to detect and segment critical regions, while Local Binary Patterns (LBP) extract textural features to improve prediction accuracy. A custom Feature Engineering Pipeline (FEP) further refines the extracted features, reducing them from 27 to 9. This not only enhances model efficiency but also ensures compatibility with mobile devices, making it suitable for resource-constrained environments. The final model, utilizing LBP features, achieved an R2 of 0.99, with the Root Mean Square Error (RMSE) reduced from 52 to 0.84 and a Mean Absolute Error (MAE) of 0.22. This research introduces a new dentition image dataset and provides a robust, low-cost solution that bridges traditional livestock management and modern AI-driven techniques for accurate cattle age estimation.