For traditional structural system identification method, wire or wireless sensors and data acquisition equipments are usually needed. To improve the efficiency of structural system identification in structural health monitoring (SHM), a non-contact measurement method without artificial targets based on the computer vision technology is proposed. Consumer-grade cameras and feature optical technology based on point feature matching are adopted to detect the motions of objectives and acquiring the time history of objectives. The accuracy of the proposed method is verified by a shaking table test, in which the results of consumer-grade cameras are compared with those from linear variable differential transformer (LVDT). The research results indicate that the proposed measurement method based on computer vision technology is a closable approach for structural health monitoring.