In order to conduct an in-depth study on the deterioration mechanism of marine solidified soil, Python+OpenCV image processing software is used to process and extract data from SEM images of marine solidified soil with different curing types and erosion time. On the basis of considering the influence of the difference in picture brightness, the fractal theory is introduced to quantitatively analyze the variation of microscopic particles of solidified soil. At the same time, combined with the test results of unconfined compressive strength, the relationship between the microscopic particle structure change of marine solidified soil and the macroscopic mechanical properties is established. The results show that: (1) the difference of brightness of different SEM images will have an impact on the results of the study. In the binarization process, the error can be reduced by controlling the whole amount of pixel (2) the microscopic particle variation of solidified soil eroded by seawater has fractal characteristics. For cement solidified soil (CS) and steel slag powder + cement cured soil (A-SSP-CS) mixed with activating agent, the fractal dimension of microscopic particles decreases gradually with the extension of erosion time, while the fractal dimension of microscopic particles increases with the extension of erosion time for steel slag powder + cement solidified soil (SSP-CS). (3) for CS and A-SSP-CS, the unconfined compressive strength (UCS) decreases gradually with the extension of seawater erosion time, and the fractal dimension of microscopic particles decreases gradually. For SSP-CS, its UCS will gradually increase with the extension of seawater erosion time, and the fractal dimension of particles will also increase gradually.