In the recent years, various plate-like structures have been widely applied in advanced engineering structures. However, unavoidable defects or damage, such as delamination and impact-induced damage, can occur during their manufacturing process and service life, which can heavily threaten their structural integrity. This study proposes an intelligent damage detection method based on principal component analysis (PCA) is proposed, which allows to decompose pseudo-defect-mode (PDM) from low-frequency vibration test and visualize defects in plate-like structures without any requirement of baseline data. In this study, a novel framework for decomposing pseudo-defect-modes from low-frequency vibration is established to isolate and filter the damage information. In addition, an intelligent strategy, namely pseudo-frequency response function, for identification of pseudo-defect-modes is proposed. Finally, a baseline-free damage index is developed to detect various defects or damage in plate-like structures. The explorations on numerical simulations and experimental tests have been conducted by using PCA-based damage detection method to effectively obtain the PDMs and damage characteristics from aluminium plate with impact-induced damage and composite laminates with delamination. Results validate the effectivity of proposed method, and indicate that it can serve as a reliable and robust vibration-based damage detection method. Compared with other researches, this work elaborates the possibility to observe defect modes in low-frequency range, which paves a new way to solve the sensitivity challenge in low-frequency based damage detection. Moreover, it plays an essential role in practical engineering application, and provides an intelligent strategy for the promotion of non-destructive testing.