In this thesis, we propose a system of saliency-based detection and recognition for static waymarks. First the users have to capture the video and in the video select the static waymarks which have the three physical properties: 1) strong contrast, 2) simple texture, and 3) regular shape. Based on these properties, we design the saliency detection which can fast find the candidate region in the image. In the region, the directional filter banks and HSV color space will analyze the shape and the color respectively for recognition. With our system, we consider the following challenges:1) lighting change, 2) scale and 3) rotation of objects, 4) different shapes of waymarks, and 5) imbalanced training data.