An image processing apparatus for synthesizing continuously captured images includes a positioning unit configured to position images to be synthesized by using a motion vector indicating a positional deviation of the images between the images, and a synthesis unit configured to synthesize the images positioned by the positioning unit. The image processing apparatus further includes a group classification unit configured to classify the images into groups based on an order of imaging, and a setting unit configured to set a reference image in each group. The positioning unit includes a first positioning unit configured to position the images within the groups with respect to the respective reference images, and a second positioning unit configured to position the reference images between the groups.