Subsurface defects directly affect the imaging quality and laser damage threshold of fused silica optical components. In comparison with the two-dimensional cross-section size and depth of the defect, the quantitative detection results of the three-dimensional contour of the subsurface defect and its volume can facilitate a more accurate evaluation of the processing quality of the fused silica optical component. In this work, a tomography scanning experiment of the fused silica sample is performed using a confocal microscope based on the principle of confocal microscope. Moreover, a three-dimensional reconstruction algorithm for the subsurface defects of fused silica components is proposed based on the analysis of the characteristics of subsurface defect images. The proposed algorithm outperforms other reconstruction methods in terms of efficiency and accuracy. Based on the statistical results of the reconstructed defects, the complete three-dimensional information of the subsurface defects of the fused silica samples is quantitatively obtained.