There is provided a tomographic image forming apparatus which divides light output from a light source inside the apparatus into measurement light and reference light, and which generates a cross-sectional image of an imaging target, based on light intensity of interference light obtained from reflected light obtained by emitting the measurement light to the imaging target and the reference light. A second image is generated by converting a first image in which line data generated based on the light intensity and having information in a direction of a first axis which serves as a depth direction of the imaging target is arranged in a direction of a second axis, into a frequency domain. An artifact is removed or reduced by performing filtering on the second image. A third image is generated by inversely converting the processed second image into a spatial domain.