An image acquisition device 1 uses as as an image pickup element 6 a two-dimensional image pickup element which is adapted to perform rolling readout while having a plurality of pixel columns 31. By utilizing a delay in image data readout timings among the pixel columns 31 in rolling readout, movement of a predetermined part Sa of a sample S within a field V of an objective lens 25 is synchronized with the rolling readout such that each pixel column 31 of the image pickup element 6 is exposed to an optical image of the predetermined part Sa in the sample S, while changing a focal position of the objective lens 25. As a consequence, image data from each pixel column 31 includes contrast information obtained when the focal position of the objective lens 25 is changed in the same part of the sample S, whereby focus information can be calculated rapidly and accurately according to the contrast information.
