| 摘要 |
A method for creating a text label 3 for an image 4 comprises selecting a group of pixels 20 which contain the image and searching for a text string associated with the image. The text string is a candidate text string based on at least one first criteria associated with the said text string. The first criteria may include the number of characters in the strings or the proportion of alphabetical characters. For example, a text string may be rejected where it comprises only a single character or where it comprises only numerals such as a telephone number. Upon determining the text string is a candidate text string, the method creates a text label for the image based upon the candidate text string. In this way the most appropriate label for an image can be determined from amongst a group of potential labels. The text string may be recognised using OCR. The method is preferably used for labelling features of a GUI for automatically creating a test script to test the GUI. |
