The high-resolution en face tomographic images of normal and cancerous human liver tissues were obtained by using a full field optical coherence tomography (FFOCT) system built in our laboratory. To solve the problem that it is difficult to distinguish between the normal and early cancerous liver tissues with unaided human eye from the FFOCT images, the gray-scale features of the two kinds of the FFOCT images were extracted and statistical analysis was made to identify the differences between them in the values of parameters characterizing their features. The results showed that four of the five characteristic parameters, including mean, variance and kurtosis etc., showed significant differences between normal and cancerous liver tissues, which can be used to identify whether the liver tissues were cancerous. The results of this study lay a foundation for further study on how to use the feature parameters extracted from the FFOCT images as the pathological indicators of cancerous tissues in the actual clinical diagnosis.