In close range photogrammetry and vision metrology, several images which are taken at different stations are required for high accuracy. Before camera calibration and 3D reconstruction, the targets in the images must be recognized and located with high accuracy firstly. Furthermore, in order to monitor the deformation of the surface, real-time and on-line photogrammetry system is needed, in which high speed is necessary. So, the image processing method and speed will affect the accuracy and speed of the photogrammetry system. This paper describes a fast target location method for the photogrammetry system. Experimental results show that the target edge pixels preserve the important geometric information for subpixel centroid, which can reach accuracies to 2-3% of the pixel size. The process time of an image with 3008x2000 pixels is about 0.1S, much higher than other similar methods.© (2010) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.