In recent years, optical coherence tomography (OCT) has been played an increasingly important role in the diagnosis of eye diseases, rapid automatic OCT film layer extraction can provide the basis for the initial diagnosis, to some extent, the workload of ophthalmologists is reduced. In this paper, the film layer extraction and preliminary diagnosis of the OCT image are completed by means of region segmentation. Firstly, ILM and the boundary of IS/OS are obtained by using threshold segmentation and reducing region, then, the thickness of the retina, the holes between film layers and the reflected signals of the vitreous are extracted, as a basis for disease diagnosis, finally, the extraction of intermediate three film layers are completed by means of subdivision region methods.