A technique based on a multi-dimensional signal processing approach is here described for performance enhancement of distributed optical fibre sensors. In particular, the main features of linear and nonlinear image denoising techniques are described for signal-to-noise ratio enhancement in Brillouin optical time-domain analysers. Experimental results demonstrate the possibility to enhance the performance of distributed Brillouin sensors by more than 13 dB using a nonlinear image denoising approach, while more than 20 dB enhancement can be obtained with video denoising.