The present disclosure provides a system for stitching image Chinese characters into a word. In the system, a scanning module scans a plurality of images, wherein a physical space of each image is mapped with one logic space on which a colliding unit is formed; when a collision event between the colliding unit of the image and a colliding unit of its neighboring image occurs in the logic space, an associating module associates the Chinese character corresponding to the colliding unit with the Chinese character corresponding to the colliding unit of its neighboring image to generate a word, and when the word was recorded in a pre-stored vocabulary, a displaying module displays the word; in this way, the system enables Chinese beginners to learn words such as compounds, idioms, and proverbs, which improves the efficiency of Chinese beginners to learn Chinese, promotes their learning enthusiasm, and enhances user experience. Another object of the present disclosure is to provide a mobile terminal that employs the system for stitching image Chinese characters into a word.