An image capturing method and device, for use in resolving the problem in the prior art of low recognition rate caused when a target photographed object is recognized by using a recognition method based on a position error value and an area error value. The image capturing method comprises: a terminal device obtains an image comprising a to-be-photographed target object, and extracts feature information of the to-be-photographed target object in the image comprising the to-be-photographed target object; the terminal device determines a target region in a preview frame presented on a display interface of the terminal device; and the terminal device triggers photographing when it is determined that feature information of an image in the target region comprises the feature information of the to-be-photographed target object.