Dataset associated with Figures 2-4 in the Manuscript entitled: "Memristor-Based Adaptive Neuromorphic Perception in Unstructured Environments". More in details, the dataset "Fig2" contains the original data corresponding to the following subplots: Fig.2c) The U-I test of memristors. In this test, a 500mV peak-to-peak sine wave is applied to the memristor and the 10kΩ fixed-value resistor. Fig.2d) Electrical characterisation data of memristors using positive pulses. Fig.2e) Electrical characterisation data of memristors using negative pulses. Fig.2f) Values of resistance in the piezoelectric film and in the memristor, force values, stimuli and response values, corresponding to the amplification processing of hazardous stimuli. Fig.2h) Values of resistance in the piezoelectric film and in the memristor, force values, stimuli and response values, corresponding to the adaptation processing of mild stimuli. Fig.2i) Sensitization processing achieved through further differentiation. Fig.2k&l Realization of adaptation at different speeds and the adaptive rate. The dataset "Fig.3" contains the original data corresponding to the following subplots: Fig.3b) resistance values of the piezoelectric film and the memristor when learning the properties of sharp objects and ultimately achieving safe grasping. Fig.3c) resistance values of the piezoelectric film and the memristor when learning the properties of slippery objects and ultimately achieving stable grasping. The dataset "Fig.4" contains the original data corresponding to the subplot Fig.4d) light intensity, differential voltage and memristor state value in front vehicle detection.