A target detection method for passive terahertz images is put forward. Original 0.22 THz images are captured using pre-developed passive scan imaging system. The original images are de-noised by spatial filtering. Then the point spread function of the imaging system is measured through pinhole images analysis, and is applied to reconstruct the images with the Lucy-Richardson iterative algorithm. Finally, gray level transform and edge detection are adopted to improve the images′ contrast and resolution capability. Experimental results show that the proposed method is able to effectively improve passive terahertz images′ quality and enhance the imaging system′s ability of detecting hidden suspicious objects.