Human tactile perception involves the response of skin tactile receptors to external stimuli and the processing of this information by the central nervous system. However, pressure sensors based on a single sensing mechanism face challenges when applied as tactile biomimetic skins for simultaneous mechanical stimulus perception and object recognition. Here, it is propose that a multifunctional ionic fiber membrane, modified with 1-ethyl-3-methylimidazolium dicyandiamide ([EMIM][DCA]), can achieve high-performance ionic capacitive sensing and triboelectric generation. [EMIM][DCA] is highly doped into thermoplastic polyurethanes through unique hydrogen bonds, resulting in a capacitive tactile sensor with ultra-high sensitivity (184.3 kPa−1) and ultra-low detection limit (1.9 Pa). Dicyandiamide ions introduce electron-donating groups to the positive triboelectric layer, increasing the output performance of triboelectric nanogenerators (TENG) by 2.47 times, with excellent stability exceeding 20 000 cycles. By integrating these excellent tactile sensing and triboelectric properties, an intelligent glove is developed to capture subtle gripping motions and identify different materials. Expanding the array of TENG and integrating machine learning into an automated sorting gripper can further enhance the variety of recognized objects and improve material recognition accuracy to 99.17%. This strategy offers a more comprehensive approach to imparting artificial tactile perception to robots, demonstrating significant potential in human-machine interaction.