Abstract This paper investigates on the entire finger dorsal surface for human identity that can be extremely beneficial for forensics applications and its related fields. Further, this paper formulates a novel approach to achieve improved performance by simultaneous extraction and integration of finger knuckle geometric and texture features by score level fusion. The geometric features are derived through Angular Geometric Analysis Method (AGAM) which extracts angular-based feature information for unique identification. Similarly, Texture Feature Extraction Methods (TFEM) viz., Completed Local Ternary Pattern (CLTP) generation method, 2D Log Gabor Filter (2DLGF) method and Fourier – Scale Invariant Feature Transform (F-SIFT) method are incorporated to derive the local texture features of an acquired finger back knuckle surface. The experimental results indicate that integration of geometric and local texture features of finger knuckle regions shows decrease in error rate by 27% (in average) when compared to the existing benchmark system taken for comparison.