Carbon fiber reinforced plastics (CFRPs) structures are very susceptible to invisible damage induced by the low velocity impact. The impact area localization can be useful information for detecting this damage. In this paper, the low velocity impact area localization system using fiber Bragg grating (FBG) sensors and area localization algorithm based on extreme learning machine (ELM) were investigated. The FBG sensors were used to detect the impact signal. Fourier transform, ReliefF algorithm, and principal component analysis technology were used to extract the impact signal characteristic. The ELM technology was used to realize the impact area localization. Finally, the impact area localization system was established and verified on a CFRP plate with 240-mm × 240-mm experiment area. The experimental results showed that the proposed system made accurate identification 35 times for 36 times experiments. The area localization accuracy was 96.9%. In addition, the precision of the area localization system was 40-mm × 40-mm area. This paper provided a reliable method for CFRP low velocity impact area localization.Index Terms-Fiber Bragg grating, low velocity impact area localization, signal characteristic, carbon fiber reinforced plastics, extreme learning machine.