Since each material has a unique ability to lose or obtain electrons, specific triboelectric signals are produced when triboelectric materials are in contact with different objects. Triboelectric nanogenerator (TENG) devices show great potential for use as tactile sensors; nevertheless, analyzing the structure− function relationship of functionalized triboelectric sensing interfaces under environmental conditions and improving the sensing stability and accuracy through the design of hydrophobic structure on the triboelectric material surface remain major challenges in the development of intelligent sensing networks. Compared with the traditional rigid micronanostructure, the elastic micronanostructure strategy is applied to achieve both hydrophobicity and stability of triboelectric materials based on the template method in this work. The corresponding surface roughness and contact angle are 89.9 nm and 117.9°, respectively. As expected, the output voltage and charge density are enhanced by almost 65.8 and 33.4%, respectively, with the establishment of an elastic micronanostructure on the triboelectric material surface. More importantly, the triboelectric signal waveforms also present acceptable durability for subsequent recognition after immersion in water or ethanol for 12 days and metal impact for 12 000 cycles. Hence, combined with deep machine learning and triboelectric effect, a material perception system integrated with a moistureresistant TENG-based sensor after fatigue testing, data processing, and display modules is also developed for real-time monitoring with approximately 100% (mask), 76% (plank), 93% (plastic), and 89% (rubber) identification accuracies in the natural environment. Finally, the proposed hydrophobic and elastic triboelectric materials show broad potential for application in the field of human−computer interaction.