High-resolution palmprint recognition is a challenging problem due to deficiencies in images, such as poor quality, skin distortion, and unallocated images. Considering the importance of high-resolution palmprints in forensic applications, this study proposes a novel multimodal palmprint scheme that combines the left and right palmprints using feature-level fusion by exploiting the similarity between the left and right palmprints using high-resolution images. The proposed system accepts as input palmprints that were captured at 500 ppi, which is the standard in forensic applications. The system is implemented by employing a statistical gray-level co-occurrence matrix (GLCM) as the texture feature extraction algorithm. Then, the features are ranked based on their probability distribution functions (PDFs) to select the most significant features. Finally, an enhanced probabilistic neural network (PNN) is used to estimate the recognition system. The benchmark THUPALMLAB database is used to conduct experiments, the results of which demonstrate that the proposed method can yield satisfactory results.