Abstract: In order to realize stereo vision localization for mobile robot with high efficiency and low cost in the automated factory, a stereo vision localization method for mobile robot based on auxiliary target was proposed. Firstly, background subtraction based on Gaussian mixture model and morphological methods were used to detect motion areas of mobile robot from video images. Secondly, feature points in motion areas of mobile robot were extracted by the FAST algorithm. On this basis, by analyzing three-channel color characteristics of neighborhood pixels on each feature point, four vertices of the square auxiliary target fixed on the mobile robot were screened from extracted feature points. Finally, the 3D coordinate for localization of mobile robot was calculated with intrinsic and extrinsic parameters of binocular cameras calibrated. Experiment results show the feasibility and validity of the proposed method, which has the advantage of small algorithm complexity and high real-time.