An image data retrieving method and an image data retrieving device are provided. The image data retrieving method includes: receiving an image including a plurality of data from a communication interface; obtaining a plurality of regions of interest from the image, wherein each of the regions of interest is a data image including at least one of the data; dividing the regions of interest into a plurality of groups, wherein at least one of the data included in the regions of interest of each of the groups has a same type; combining the regions of interest of each of the groups into a to-be-recognized image; and performing an optical character recognition to the to-be-recognized image corresponding to each of the groups respectively to obtain the data corresponding to the regions of interest of each of the groups.