<p>Muzzle print recognition, an essential aspect of biometric identification for animals, has gained significant attention in recent years. This study introduces a novel method for enhancing the precision and robustness of muzzle print identification through the utilization of a Two-Stream Convolutional Neural Network (CNN) integrated with Gabor Filters. The proposed Two-Stream CNN architecture comprises of two distinct streams, the Spatial Stream, which processes raw muzzle images, and the Texture Stream, dedicated for extracting texture features through Gabor Filters. By integrating feature maps from both streams, the model effectively utilizes the synergistic advantages of spatial and texture information, thereby augmenting the discriminative capabilities of the recognition system. The Gabor Filters applied in the Texture Stream enable the extraction of intricate texture details, making the model highly sensitive to variations in muzzle patterns. By conducting training and fine-tuning procedures on animal muzzle prints in the Beef Cattle Muzzle/Nose print database for individual identification, a recognition accuracy of 98.99% was achieved. Additional performance metrics include a recall of 98.36%, precision of 98.39%, specificity of 100% and F1-Score of 98.18%. The ability of the proposed model to accurately differentiate intricate variations in muzzle prints presents significant prospects for diverse fields of application, such as wildlife preservation, animal monitoring, and security measures.</p>