In order to take advantage of global navigation map for robot self-localization and solve kidnap problem, a robot vision localization system is presented based on graphic content matching. It can make good use of the different objects and their layout in different rooms or corridors to fix robot position, which cannot be disturbed by similar objects. This vision localization system is composed of frames overlap region extraction and overlap region re-building through sub-blocks matching, and the interference caused by points on the wall and mismatching sub-blocks can be deleted. The image distortion can be adjusted to the same before matching. In the experiment, this graphic matching method can match the real-time robot vision with keyframes global map effectively, and find out the most similar keyframe for each vision image and fix robot position exactly. More than 95% robot vi-sion can be matched and position RMSE<0.5 m. Robot can also localize itself effectively when it is kidnapped.rebuilding