In an embodiment according to the present invention, an image processing apparatus (70) includes an uplift shape estimator (751), a subsidence shape estimator (752), and an abnormality candidate detector (755). The uplift shape estimator (751) and the subsidence shape estimator (752) model gradient variations of pixel values in a lumen in-vivo image according to the pixel values of the lumen in-vivo image. The abnormality candidate detector (755) detects an abnormality candidate area, which is shown in the lumen in-vivo image, according to a difference between the pixel value of each pixel constituting the lumen in-vivo lumen and an estimated pixel value of each pixel that is determined according to the modeled gradient variations of the pixel values.
