While　three-dimensional　tactile　sensor　arrays　featuring　traditional　3D　measurement　structures　have　been　extensively　studied,　their　limited　spatial　resolution　hinders　their　ability　to　perceive　small　objects　during　dexterous　manipulation.　Moreover,　existing　researches　on　tactile　super-resolution　still　suffer　from　lack　of　contour　perception,　mapping　accuracy　and　difficult　access　to　super-resolution　datasets.　Therefore,　a　prominent　two-stage　super-resolution　algorithm　with　digital-twin-driven　enhancement　is　proposed　for　sensor　arrays　with　classical　3D　measurement　structures.　Firstly,　based　on　the　intrinsic　features　of　classical　3D　measurement　structures,　digital-twin-driven　observation　of　low/high　resolution　tactile　sensor　arrays　is　introduced　for　reliable　and　accurate　tactile　super-resolution　datasets.　Secondly,　a　novel　deconstructive　interpolative　up-sampling　is　then　proposed　for　digital-twin-driven　enhancement　and　super-resolution　multiplier　augmentation.　Furtherly,　a　interpolation-convolution　two-stage　super-resolution　network　is　designed,　where　the　deconstructive　interpolative　up-sampling　is　proposed　as　the　first-stage　network　to　effectively　and　accurately　increase　the　scale　of　tactile　information,　and　a　convolutional　neural　network　with　channel　features　learning　is　further　constructed　as　the　second-stage　for　high-quality　quadruple　super-resolution.　Comprehensive　experiments　demonstrate　the　reliability　of　the　digital-twin-dataset　and　the　effectiveness　of　our　approach,　with　the　digital-twin-trained　network　achieving　quadruple　super-resolution　(PSNR:31.12,　SSIM:0.965)　for　the　self-made　sensor,　enabling　clear　recognition　of　complex　Braille　letters　and　higher　positioning　resolution　(0.625mm)　for　small　objects　(connector　of　RF　antennas　e.g.).　Our　super-resolution　tactile　sensing　framework　realizes　the　assembly　of　RF　antennas　and　holds　promise　for　enhancing　the　dexterity　of　robotic　manipulation.　IEEE