An image processing device includes an input unit and an alignment unit. The input unit inputs a cell shape image and a fluorescence image. The cell shape image shows a shape of a cell in a tissue section. The fluorescence image shows expression of a specific protein as a fluorescent bright point in a region same as a region in the tissue section. The alignment unit aligns the cell shape image and the fluorescence image based on an information source detected in both the cell shape image and the fluorescence image.