An intelligent terminal based man-machine interaction method, comprising: recognizing a reference object; recognizing a target object and outputting a recognition result, wherein the recognition result comprises at least one of target object information and visible image-text information recorded by the target object; acquiring position information about the target object with respect to the reference object; determining the intention of a user according to the position information; and according to the intention of the user, processing the recognition result and then outputting same to the user. According to the intelligent terminal based man-machine interaction method, a target object can be recognized and a recognition result is output. Also the intention of a user can be acquired according to the position information about the target object with respect to a reference object, and the recognition result is processed and is then output to the user according to the intention of the user. During an interaction process, a user only needs to place a target object at a corresponding position, and thus the effects of simple operation and better user experience are achieved. The invention also relates to an intelligent terminal based man-machine interaction system.