The detection of delaminations in composite structures due to impacts is a critical issue for any structural health monitoring (SHM) programme. In this work, an experimental campaign is carried out to investigate the presence of delamination induced in carbon-fibre-reinforced-plastic (CFRP) plates by low velocity impacts (LVI). A reduced number of piezoelectric devices for actuation and sensing purposes are employed. The proposed study starts with the choice of the proper positions of the piezoelectric sensor and actuator through the analysis of the numerical curvature mode shapes of the composite laminated plate. A wavelet packet transform (WPT)-based algorithm is applied on the vibrating response of the plates to extract wavelet-based damage sensitive features. Linear Discriminant Analysis (LDA) is then applied on the extracted damage sensitive features to enhance the performance of the proposed delamination identification procedure. Results show the effectiveness of the developed SHM routine when a reduced number of sensors is either desirable or mandatory.@en