In order to avoid the problems in multi-threshold-based segmentation method, an image segmentation method based on principal component analysis (PCA) was introduced.Through reconstructing an image by Singular Value Decomposition, the gray value of the background pixels can be transformed to near zero.With this method, only two thresholds are employed to segment the objects belonging to different gray value ranges in the image.The vadility of the method was confirmed by segmentation experiments of several images.The experimental results show that this method is insensitive to uneven illumination and noise, and only two thresholds need to be optimized, which greatly reduces the complexity and time consumption of threshold optimization in the multi-threshold-based segmentation method.